Advanced Topics

Stanford Mathematics Tournament 2000

1. How many different ways are there to paint the sides of a tetrahedron with exactly 4 colors? Each side
gets its own color, and two colorings are the same if one can be rotated to get the other.
√ 6 √ 6
2. Simplify ( −1+i
2
3
) + ( −1−i
2
3
) to the form a + bi.

∞ 1
3. Evaluate n=1 n2 +2n .
4. Five positive integers from 1 to 15 are chosen without replacement. What is the probability that their
sum is divisible by 3?
5. Find all 3-digit numbers which are the sums of the cubes of their digits.
6. 6 people each have a hat. If they shuffle their hats and redistribute them, what is the probability that
exactly one person gets their own hat back?
√ √
7. Assume that a, b, c, d are positive integers, and ac = db = 34 , a2 + c2 − b2 + d2 = 15. Find ac + bd −
ad − bc.
8. How many non-isomorphic graphs with 9 vertices, with each vertex connected to exactly 6 other
vertices, are there? (Two graphs are isomorphic if one can relabel the vertices of one graph to make
all edges be exactly the same.)

9. The Cincinnati Reals are playing the Houston Alphas in the last game of the Swirled Series. The
Alphas are leading by 1 run in the bottom of the 9th (last) inning, and the Reals are at bat. Each
batter has a 13 chance of hitting a single and a 23 chance of making an out. If the Reals hit 5 or more
singles before they make 3 outs, they will win. If the Reals hit exactly 4 singles before making 3 outs,
they will tie the game and send it into extra innings, and they will have a 35 chance of eventually
winning the game (since they have the added momentum of coming from behind). If the Reals hit
fewer than 4 singles, they will LOSE! What is the probability that the Alphas hold off the Reals and
win, sending the packed Alphadome into a frenzy? Express the answer as a fraction.
10. I call two people A and B and think of a natural number n. Then I give the number n to A and the
number n + 1 to B. I tell them that they have both been given natural numbers, and further that they
are consecutive natural numbers. However, I don’t tell A what B’s number is and vice versa. I start
by asking A if he knows B’s number. He says “no”. Then I ask B if he knows A’s number, and he says
“no” too. I go back to A and ask, and so on. A and B can both hear each other’s responses. Do I ever
get a “yes” in response? If so, who responds first with “yes” and how many times does he say “no”
before this? Assume that both A and B are very intelligent and logical. You may need to consider
multiple cases.
 Algebra
Stanford Mathematics Tournament 2000

1. How many integers x satisfy |x| + 5 < 7 and |x − 3| > 2?

2. Evaluate 20003 − 1999 · 20002 − 19992 · 2000 + 19993.
3. Five students take a test on which any integer score from 0 to 100 inclusive is possible. What is the
largest possible difference between the median and the mean of the scores?
4. What is the fewest number of multiplications required to reach x2000 from x, using only previously
generated powers of x? For example, x → x2 → x4 → x8 → x16 → x32 → x64 → x128 → x256 →
x512 → x1024 → x1536 → x1792 → x1920 → x1984 → x2000 uses 15 multiplications.
5. A jacket was originally priced $100. The price was reduced by 10% three times and increased by 10%
four times in some order. To the nearest cent, what was the final price?
6. Barbara, Edward, Abhinav, and Alex took turns writing this test. Working alone, they could finish it
in 10, 9, 11, and 12 days, respectively. If only one person works on the test per day, and nobody works
on it unless everyone else has spent at least as many days working on it, how many days (an integer)
did it take to write this test?
7. A number n is called multiplicatively perfect if the product of all the positive divisors of n is n2 .
Determine the number of positive multiplicatively perfect numbers less than 100.
8. A man has three daughters. The product of their ages is 168, and he remembers that the sum of their
ages is the number of trees in his yard. He counts the trees but cannot determine any of their ages.
What are all possible ages of his oldest daughter?
√ √
9. ac = db = 34 , a2 + c2 − b2 + d2 = 15. Find ac + bd − ad − bc.
10. Find the smallest positive integer a such that x4 + a2 is not prime for any integer x.
 Calculus
Stanford Mathematics Tournament 2000

1. Find the slope of the tangent at the point of inflection of y = x3 − 3x2 + 6x + 2000.
2. Karen is attempting to climb a rope that is not securely fastened. If she pulls herself up x feet at once,
then the rope slips x3 feet down. How many feet at a time must she pull herself up to climb with as
few pulls as possible?
3. A rectangle of length 14 π and height 4 is bisected by the x-axis and is in the first and fourth quadrants,
with the leftmost edge on the y-axis. The graph of y = sin(x) + C divides the area of the square in
half. What is C?
π
4. For what value of x (0 < x < 2) does tan x + cot x achieve its minimum?

xi
5. For −1 < x < 1, let f (x) = ∞ i=1 i . Find a closed form expression (a closed form expression is one
not involving summation) for f .
√
6. A hallway of width 6 feet meets a hallway of width 6 5 feet at right angles. Find the length of the
longest pipe that can be carried horizontally around this corner.
7. An envelope of a set of lines is a curve tangent to all of them. What is the envelope of the family of
lines y = x20 + x(1 − x12 ), with x0 ranging over the positive real numbers?
0

 π
2
8. Find 0 ln sin θdθ.
    √
9. Let f (x) = x+ 0+ x + 0 + x + . . .. If f (a) = 4, then find f  (a).
√
10. A mirror is constructed in the shape of y equals ± x for 0 ≤ x ≤ 1, and ±1 for 1 < x < 9. A ray of
light enters at (10,1) with slope 1. How many times does it bounce before leaving?
 General
Stanford Mathematics Tournament 2000

1. If a = 2b + c, b = 2c + d, 2c = d + a − 1, d = a − c, what is b?
2. The temperatures f ◦ F and c◦ C are equal when f = 95 c + 32. What temperature is the same in both
◦
F and ◦ C?
3. A twelve foot tree casts a five foot shadow. How long is Henry’s shadow (at the same time of day) if
he is five and a half feet tall?
4. Tickets for the football game are $10 for students and $15 for non-students. If 3000 fans attend and
pay $36250, how many students went?
5. Find the interior angle between two sides of a regular octagon (degrees).
6. Three cards, only one of which is an ace, are placed face down on a table. You select one, but do not
look at it. The dealer turns over one of the other cards, which is not the ace (if neither are, he picks
one of them randomly to turn over). You get a chance to change your choice and pick either of the
remaining two face-down cards. If you selected the cards so as to maximize the chance of finding the
ace on the second try, what is the probability that you selected it on the
(a) first try?
(b) second try?
√
7. Find [ 19992000] where [x] is the greatest integer less than or equal to x.
8. Bobo the clown was juggling his spherical cows again when he realized that when he drops a cow is
related to how many cows he started off juggling. If he juggles 1, he drops it after 64 seconds. When
juggling 2, he drops one after 55 seconds, and the other 55 seconds later. In fact, he was able to create
the following table:
cows started juggling 1 2 3 4 5 6 7 8 9 10 11
seconds he drops after 64 55 47 40 33 27 22 18 14 13 12
cows started juggling 12 13 14 15 16 17 18 19 20 21 22
seconds he drops after 11 10 9 8 7 6 5 4 3 2 1
He can only juggle up to 22 cows. To juggle the cows the longest, what number of cows should he start
off juggling? How long (in minutes) can he juggle for?
9. Edward’s formula for the stock market predicts correctly that the price of HMMT is directly propor-
tional to a secret quantity x and inversely proportional to y, the number of hours he slept the night
before. If the price of HMMT is $12 when x = 8 and y = 4, how many dollars does it cost when x = 4
and y = 8?
10. Bob has a 12 foot by 20 foot garden. He wants to put fencing around it to keep out the neighbor’s dog.
Normal fenceposts cost $2 each while strong ones cost $3 each. If he needs one fencepost for every 2
feet and has $70 to spend on the fenceposts, what is the largest number of strong fenceposts he can
buy?
a+b
11. If a@b = a−b , find n such that 3@n = 3.
12. In 2020, the United States admits North Mathematica as the 51st state. It consists of 5 islands joined
by bridges as shown. Is it possible to cross all the bridges without doubling over? If so, what is the
difference (positive) between the number of the start island and the number of the end island?
 1 2 3

4 5

13. How many permutations of 123456 have exactly one number in the correct place?
14. The author of this question was born on April 24, 1977. What day of the week was that?
(53 ) (35 )
15. Which is greater: (35 ) or (53 ) ?

16. Joe bikes x miles East at 20 mph to his friend’s house. He then turns South and bikes x miles at 20
mph to the store. Then, Joe turns East again and goes to his grandma’s house at 14 mph. On this last
leg, he has to carry flour he bought for her at the store. Her house is 2 more miles from the store than
Joe’s friend’s house is from the store. Joe spends a total of 1 hour on the bike to get to his grandma’s
house. If Joe then rides straight home in his grandma’s helicopter at 78 mph, how many minutes does
it take Joe to get home from his grandma’s house?
17. In how many distinct ways can the letters of STANTON be arranged?
18. You use a lock with four dials, each of which is set to a number between 0 and 9 (inclusive). You
can never remember your code, so normally you just leave the lock with each dial one higher than the
correct value. Unfortunately, last night someone changed all the values to 5. All you remember about
your code is that none of the digits are prime, 0, or 1, and that the average value of the digits is 5.
How many combinations will you have to try?
19. Eleven pirates find a treasure chest. When they split up the coins in it, they find that there are 5 coins
left. They throw one pirate overboard and split the coins again, only to find that there are 3 coins left
over. So, they throw another pirate over and try again. This time, the coins split evenly. What is the
least number of coins there could have been?
20. Given: AC has length 5, semicircle AB has radius 1, semicircle BC has diameter 3. What percent of
the the big circle is shaded?

A B C

21. Find the area of the six-pointed star if all edges are of length s, all acute angles are 60◦ and all obtuse
angles are 240◦.

22. An equilateral triangle with sides of length 4 has an isosceles triangle with the same base and half the
height cut out of it. Find the remaining area.
 77
23. What are the last two digits of 77 ?
24. Peter is randomly filling boxes with candy. If he has 10 pieces of candy and 5 boxes in a row labeled A,
B, C, D, and E, how many ways can he distribute the candy so that no two adjacent boxes are empty?
25. How many points does one have to place on a unit square to guarantee that two of them are strictly
less than 1/2 unit apart?
√ √
26. Janet√is trying to find Tim in a Cartesian
√ forest. Janet is 5√ 2 miles from (0,0), 41 miles
√ from (1,0),
and 61 miles from (0,1). Tim is 65 miles from (0,0), 2 13 miles from (1,0), and 58 miles from
(0,1). How many miles apart are Janet and Tim?
 Geometry
Stanford Mathematics Tournament 2000

1. How many rectangles are there on an 8x8 checkerboard?

2. In a triangle the sum of squares of the sides is 96. What is the maximum possible value of the sum of
the medians?
3. Find P B, given that P A = 15, P C = 20, P D = 7, and ABCD is a square.
A B
C #
C #
#
C #
C#
aP

aa

aa
D C

4. Find the total area of the non-triangle regions in the figure below (the shaded area).
a/4

b/6
b/3

b/3

a/3
a

5. Side AB = 3. ABF is an equilateral triangle. Side DE = AB = AF = GE.  F ED = 60◦ . F G = 1.

Calculate the area of ABCDE.
A

G
F C B

E D

√
6. What is the area of the largest circle contained in an equilateral triangle of area 8 3?
 x2 y2
7. Let ABC be a triangle inscribed in the ellipse 4 + 9 = 1. If its centroid is the origin (0,0), find its
area.
√
8. A sphere is inscribed inside a pyramid with a square as a base whose height is 215 times the length
of one edge of the base. A cube is inscribed inside the sphere. What is the ratio of the volume of the
pyramid to the volume of the cube?
9. How many hexagons are in the figure below with vertices on the given vertices? (Note that a hexagon
need not be convex, and edges may cross!)

√
10. Let C1 and C2 be two concentric reflective hollow metal spheres of radius R and R 3 respectively.
From a point P on the surface of C2 , a ray of light is emitted inward at 30◦ from the radial direction.
The ray eventually returns to P. How many total reflections off of C1 and C2 does it take?
 Team Test
Stanford Mathematics Tournament 2000

1. You are given a number, and round it to the nearest thousandth, round this result to nearest hundredth,
and round this result to the nearest tenth. If the final result is .7, what is the smallest number you
could have been given? As is customary, 5’s are always rounded up. Give the answer as a decimal.
2. The price of a gold ring in a certain universe is proportional to the square of its purity and the cube
of its diameter. The purity is inversely proportional to the square of the depth of the gold mine
and directly proportional to the square of the price, while the diameter is determined so that it is
proportional to the cube root of the price and also directly proportional to the depth of the mine. How
does the price vary solely in terms of the depth of the gold mine?
3. Find the sum of all integers from 1 to 1000 inclusive which contain at least one 7 in their digits, i.e.
find 7 + 17 + ... + 979 + 987 + 997.
4. All arrangements of letters VNNWHTAAIE are listed in lexicographic (dictionary) order. If AAEHIN-
NTVW is the first entry, what entry number is VANNAWHITE?
1
5. Given cos (α + β) + sin (α − β) = 0, tan β = 2000 , find tan α.
6. If α is a root of x3 −x−1 = 0, compute the value of α10 +2α8 −α7 −3α6 −3α5 +4α4 +2α3 −4α4 −6α−17.
7. 8712 is an integral multiple of its reversal, 2178, as 8712=4*2178. Find another 4-digit number which
is a non-trivial integral multiple of its reversal.
8. A woman has $1.58 in pennies, nickels, dimes, quarters, half-dollars and silver dollars. If she has a
different number of coins of each denomination, how many coins does she have?

9. Find all positive primes of the form 4x4 + 1, for x an integer.

10. How many times per day do at least two of the three hands on a clock coincide?
3
11. Find all polynomials f (x) with integer coefficients such that the coefficients of both f (x) and [f (x)]
lie in the set {0, 1, −1}
12. At a dance, Abhinav starts from point (a, 0) and moves along the negative x direction with speed va ,
while Pei-Hsin starts from (0, b) and glides in the negative y-direction with speed vb . What is the
distance of closest approach between the two?
13. Let P1 , P2 , . . . , Pn be a convex n-gon. If all lines Pi Pj are joined, what is the maximum possible number
of intersections in terms of n obtained from strictly inside the polygon?
1+xn
14. Define a sequence < xn> of real numbers by specifying an initial x0 and by the recurrence xn+1 = 1−xn .
Find xn as a function of x0 and n, in closed form. There may be multiple cases.

15. limn→∞ nr 2 1 − cos 2π
n =?
 Advanced Topics Solutions
Stanford Mathematics Tournament 2000

1. Assume we have 4 colors - 1, 2, 3, and 4. Fix the bottom as color 1. On the remaining sides you can
have colors 2, 3, 4 (in that order), or 2, 4, 3, which are not rotationally identical. So, there are 2 ways
to color it.
√
6
2. Since cos 2π 1
3 = − 2 and sin 2π 3 2π 2π
3 = 2 , we can write the first term as (cos 3 + i sin 3 ) . Since cos 3 =
4π
√
1 4π 3 4π 4π 6
− 2 and sin 3 = − 2 , we can write the second term as (cos 3 + i sin 3 ) . Now, we apply DeMoivre’s
Theorem to simplify the first expression to (cos 6 · 2π 2π
3 + sin 6 · 3 ) = (cos 4π + sin 4π) = 1 + 0 = 1.
Similarly, we simplify the second expression to (cos 6 · 3 + sin 6 · 4π
4π
3 ) = (cos 8π + sin 8π) = 1 + 0 = 1.
Thus, the total sum is 1 + 1 = 2.
1 1
1 1 − 1 1
3. We know that n2 +2n = n(n+2) = n 2n+2 . So, if we sum this from 1 to ∞, all terms except for 1
2 + 2
2
will cancel out (a ”telescoping” series). Therefore, the sum will be 34 .
4. The possibilities for the numbers are:

• all five are divisible by 3

• three are divisible by 3, one is ≡ 1 (mod 3) and one is ≡ 2 (mod 3)
• two are divisible by 3, and the other three are either ≡ 1 (mod 3) or ≡ 2 (mod 3)
• one is divisible by 3, two are ≡ 1 (mod 3) and two are ≡ 2 (mod 3)
• four are ≡ 1 (mod 3) and one is ≡ 2 (mod 3)
• four are ≡ 2 (mod 3) and one is ≡ 1 (mod 3)


This gives us 1001 possible combinations out of 15
5 or 3003. So, the probability is
1001
3003 = 13 .
5. 153,370,371,407
6. There are 6 people that could get their hat back, so we must multiply 6 by the number of ways that
the other 5 people can arrange their hats such that no one gets his/her hat back. So, the number of
ways this will happen is (6 · derangement of 5), or 6 ∗ 44 = 264. Since there are 6! = 720 possible
11
arrangements of hats, the probability of exactly one person getting their hat back is 264
720 = 30 .

7. We can view these conditions as a geometry diagram as seen below. So, we know that fe = 34 (since
 √ √
e = a − b = 34 c − 34 d = 34 f and we know that e2 + f 2 = 15 (since this is a2 + c2 − b2 + d2 ). Also,
note that ac + bd − ad − bc = (a − b)(c − d) = ef . So, solving for e and f , we find that e2 + f 2 = 225,
2 2 2 2
so 16e2 + 16f 2 = 3600, so (4e) + (4f ) = 3600, so (3f ) + (4f ) = 3600, so f 2 (32 + 42 ) = 3600, so
2 2 3
25f = 3600, so f = 144 and f = 12. Thus, e = 4 · 12 = 9. Therefore, ef = 9 ∗ 12 = 108.

e
f
a
b

c
 8. It suffices to consider the complements of the graphs, so we are looking for graphs with 9 vertices,
where each vertex is connected to 2 others. There are 4 different graphs - see below.

3
9. The probability of the Reals hitting 0 singles is ( 23 ) . The probability of the Reals hitting exactly 1

 3
single is 32 ·( 23 ) · 13 , since there are 3 spots to put the two outs (the last spot must be an out, since the

 3 3
inning has to end on an out). The probability of the Reals hitting exactly 2 singles is 42 ·( 23 ) · ( 13 ) .

5 2 3 1 3
The probability of the Reals hitting exactly 3 singles is 2 ·( 3 ) · ( 3 ) . If any of these happen, the
Alphas win right away. Adding these gives us a 656 729 chance of this happening. If exactly 4 singles occur

6 2 3 1 4
(with probability 2 ·( 3 ) · ( 3 ) ), then there is a 25 chance that the Alphas win. The probability of this
happening is 25 · 72940
. , the total probability of the Alphas winning is the sum of these two probabilities,
656 16 224
or 729 + 729 = 243 .
10. A will say yes when B says no to n − 1 or n, as A will then know B’s number is one greater than A’s
number. Thus, A responds first, after n−1
2
“no” responses if n is odd, after n 2
“no” responses
if n is even.
 Algebra Solutions
Stanford Mathematics Tournament 2000

1. The only integers that satisfy |x| + 5 < 7 are the ones that satisfy |x| < 2 - namely, −1, 0, 1. The
integers that satisfy |x − 3| > 2 are 6, 7, 8, . . . and 0, −1, −2, . . .. So, the integers that satisfy both are
0, −1, and there are 2 of them.
2. 20003 −1999·20002−19992 ·2000+19993 can be factored into (2000 − 1999)20002 +19992(−2000 + 1999),
which reduces to 20002 − 19992. This factors into (2000 + 1999)(2000 − 1999), which is equal to 3999.
3. Let the scores be a,b,c,d,e, where 0 ≤ a ≤ b ≤ c ≤ d ≤ e ≤ 100. So, the mean is 15 (a + b + c + d + e),
and the median is c. So, we want to maximize 15 (a + b + c + d + e) − c. To do this, we must maximize
d and e and minimize or maximize c. One way to do this is to let a = b = c = 0 and d = e = 100, so
the difference between the mean and the median is 15 (0 + 0 + 0 + 100 + 100) − 0) = 200 5 = 40. If we
maximize c, then c = d = e = 100, and then the mean is 15 (0 + 0 + 100 + 100 + 100) = 60, and the
median is 60, with a difference of 40 as well.
4. A shortest path is x → x2 → x4 → x8 → x12 → x24 → x25 → x50 → x100 → x200 → x400 → x800 →
x1600 → x2000 , using 13 multiplications.
5. The price starts at $100. Clearly, the order of price changes does not matter. It is reduced by 10%
three times ( $100 → $90 → $81 → $72.90), and the new price is $72.90. It is increased by 10% four
times ($72.90 → $80.19 → $88.209 → $97.0299 → $106.73289) , and the new price is $106.73289.
Rounded to the nearest cent, this is $106.73.
6. Every day Edward works, he gets 19 of the test done. Similarly, every day Barbara works, she gets 10
1
1
of the test done, every day Abhinav works, he gets 11 of the test done, and every day Alex works, he
1
gets 12 of the test done. So, after 4 days (after everyone has worked on the test one day, they have
completed 19 + 10 1 1
+ 11 1
+ 12 = 38.535% of the test. After 8 days, they have completed twice that,
or 77.0707% of the test. After Edward, Barbara, and Abhinav each work one more day, the test will
be complete in the minimum amount of time, so the test will take 11 days to complete. If the least
efficient workers work after the 8th day, the test still takes 11 days to complete.
7. All multiplicatively perfect numbers have exactly 4 distinct positive divisors, or 1. So, we must look
for numbers that are either
• 1
• a product of two distinct primes
• a cube of a prime
Numbers satisfying one of these conditions less than 100 are: 1, 6, 8, 10, 14, 15, 21, 22, 26, 27, 33, 34,
35, 38, 39, 46, 51, 55, 57, 58, 62, 65, 69, 74, 77, 82, 85, 86, 87, 91, 93, 94, 95. There are 33 of these.
8. 168 = 23 · 3 · 7. There are only 2 combinations of these whose sums allow indistinguishability of the
ages. If there are 27 trees, 2, 4, 21 and 1, 12, 14 years are possible. If there are 21 trees, 2, 7, 12 and
3, 4, 14 are possible. So, the possible ages of the oldest daughter are 12, 14, 21.
 9. We can view these conditions as a geometry diagram as seen below. So, we know that fe = 34 (since
 √ √
e = a − b = 34 c − 34 d = 34 f and we know that e2 + f 2 = 15 (since this is a2 + c2 − b2 + d2 ). Also,
note that ac + bd − ad − bc = (a − b)(c − d) = ef . So, solving for e and f , we find that e2 + f 2 = 225,
2 2 2 2
so 16e2 + 16f 2 = 3600, so (4e) + (4f ) = 3600, so (3f ) + (4f ) = 3600, so f 2 (32 + 42 ) = 3600, so
2 2 3
25f = 3600, so f = 144 and f = 12. Thus, e = 4 12 = 9. Therefore, ef = 9 ∗ 12 = 108.

e
f
a
b

10. a = 1 clearly does not work, since if x = 1, then x4 + a2 = 2, which is prime. a = 2 clearly does not
work, since if x = 1, then x4 + a2 = 5, which is also prime. Here is a table for a’s and values of x that
show they do not work:
a x a4 + x2
3 10 10009
4 1 17
5 2 41
6 1 37
7 20 160049
So, consider a = 8; i.e. the sum x4 +64. This is the same as (x2 +8)2 −16x2 =(x2 + 4x + 8)(x2 − 4x + 8)
by the difference of squares. This is clearly not prime for any integer x. So, the answer is a = 8.
 Calculus Solutions
Stanford Mathematics Tournament 2000

1. y = x3 − 3x2 + 6x + 2000, so y  = 3x2 − 6x + 6 and y  = 6x − 6, so the point of inflection is the solution

2
to 6x − 6 = 0, or x = 1. At x = 1, the slope is f  |x=1 = 3(1) − 6(1) + 6 = 3.
2. The change in Karen’s position is x − x3 . The optimal length to √
climb is at a critical point. The only
realistic critical point is at the solution to 1 − 3x2 = 0 or x = 33 .
 1
4π
π
3. 0 sin x+C = 0, from the statement of the problem. So, [− cos x + Cx]|04 = 0. Thus, cos π4 + π4 C +1 =
√ √
2 2 2−4
0. So, − 2 + π4 C + 1 = 0, and solving for C, we find that C = π
.

4. Let y = tan x. So, we want to find the minimum of y + y1 , where 0 ≤ y ≤ ∞. Taking the derivative and
minimizing, we find that the minimum occurs at y = 1, so the minimum of the given function occurs
at arctan 1 = π4 .
∞ xi
∞ 1
5. f (x) = i=1 i . So, f  (x) = i=0 xi = 1−x . Thus, f (x) = − ln(1 − x).
6. Assume the pipe barely fits around the corner √ (i.e. it is in contact with the corner). The lower corner
is at (0, 0) and the upper corner is at (6, 6 5). Call x0 the point on the lower wall it hits at the tightest
√
spot. Given  an x0 ,√the longest a pipe could be with one end at x0 and leaning against the (6, 6 5)
√ 2
corner is x20 + (6 5 + 36 5
x0 −6 ) . We want the minimum of all of these ”longest pipes”, because the
pipe needs to fit at all angles around the corner. Taking the derivative (without the square root for
simplicity) and setting it equal to 0, we need to solve x30 − 6x20 + 36x0√ − 1296 = 0. We can quickly find
that x0 = 12 is the only good solution, so the maximum length is 12 6.
7. For each value of x, we want to find the minimum (maximum) of y for the range of x0 . Therefore, take
dy
dx0 , treating x as a constant. Set this equal to 0, and solve for x0 relative to x. Plug this in for x0 in
the given family to obtain the envelope y = x + x1 , x = 0.
 π
8. Let I denote the given integral. Under the transformation θ → π
2 − θ, I transforms to 2
0
ln(cos(θ))dθ.
So,
π
2I = 02 ln(sin θ cos θ)dθ
π
= 0 (ln(sin 2θ)  − ln 2)d(2θ)/2
= − 2 ln
π
2 + 1 π
2 0 πsin(α)dα giving I = − π2 ln 2.
= − π2 · ln 2 + 02 sin(α)dα
= − π2 · ln 2 + I
2 2
9. Note first that ([f (x)]2 − x) = f (x), so if f (a) = 4, then (16 − a) = 4, so a = 14. Now, f  (x) =
f
(x)
1+
2([f (x)]2 −x) 4
2f (x) , so f  (14) = 31 .

10. Solution: Throughout this solution we will use the fact that when light bounces off a mirror, the angle
of incidence is equal to the angle of reflection. First the beam hits the point (8,-1), then (6,1), (4,-1),
(2,1), and then is travelling along the line y = x − 1. Thus the beam hits the parabola at the point
√
1− 5 1− 5
√ √ 2 2
√ √
500
(1 + 2 , 2 ). To estimate 5, notice that 22 = 484 and 23 = 529, so 5 = 10 = 2.2 . . .. Thus
√
1− 5
2 = −.6 . . ., so the light hits the parabola at approximately (.4,-.6). The slope of the tangent to
the parabola at this point is −1 2 (.4)
−1/2
, which is about -.8, so we need to find the slope of the beam
after it reflects off of this tangent. For purposes of finding this slope, change coordinates so that the
point of intersection is the origin. The beam is coming in along y = x, and y = 1.2x is perpendicular
to the tangent. The diagram below should clarify the setup.
We will find the new path of the light by finding the reflection about the line y = 1.2x of a point on
its incoming path. We know the point (1,1.2) is on the line y = 1.2x, so a perpendicular through this
point is y − 1.2 = −.8(x − 1), which intersects y = x at the point (1.1,1.1). Thus the new path goes
through the point (.9,1.3), so it has slope 1.4 (all values rounded to one decimal place). Going back
 perpendicular to
tangent
reflected
beam

y=x

tangent

to our original coordinate system, the light is now travelling along the line y + .6 = 1.4(x − .4), so it
next hits the mirror at (1.5,1). After that the x coordinate increases by 2/1.4 = 1.4 between bounces,
so it hits (2.9,-1), (4.3,1), (5.7,-1), (7.1,1), (8.5,-1), and finally (9.9,1). A closer examination of the
approximations made (e.g. by refining them to two decimal places) reveals that the last bounce is
actually further to the left (at (9.21,1), to be more precise), so indeed the light does bounce 12 times.
 General Solutions
Stanford Mathematics Tournament 2000

1. Since d = a − c, substitute into the equation b = 2c + d to get b = 2c + a − c = a + c. Also, substitute

into 2c = d + a − 1 to get 2c = a − c + a − 1, or 3c = 2a − 1. Now, since b = a + c, we can substitute into
a = 2b + c to get a = 2a + 2c + c, or a = −3c. Since we know 3c = 2a − 1 from above, we substitute in
to get 3c = −6c − 1, or c = − 91 . Thus, we find that a = 13 , d = 49 , and b = 29 .
9
2. Let x be the temperature we are looking for, so x = 5 + 32. So, − 45 x = 32, so x = − 45 · 32 =−40.
5.5 x
3. Let x be the length of Henry’s shadow in feet. Using similar triangles, we find that 12 = 5, so
55
x = 5.5 11
12 · 5 = 24 · 5 = 24 .

4. Let x be the number of students and y be the number of non-students. We then have the equations
x + y = 3000 and 10x + 15y = 36250. Substituting, we find that 10(3000 − y) + 15y = 36250, or
30000 − 10y + 15y = 36250, so 30000 + 5y = 36250. So, 5y = 6250, and y = 1250, so x =1750.
5. The total number of degrees in an octagon is (8 − 2) · 180 = 1080. Since the degrees are evenly
distributed among the angles, the measure of one interior angle is 1080 ◦
8 =135 .

6. Pick any card first, then pick the other face-down card.
1
(a) 3
2
(b) 3
√
7. 19992000 ≈ 4471.241, and [4471.241] =4471.
8. The time that Bobo can juggle is the number of cows times seconds. So, we get the following table:
cows started juggling 1 2 3 4 5 6 7 8 9 10 11
total time 64 110 141 160 165 162 154 144 126 130 132
cows started juggling 12 13 14 15 16 17 18 19 20 21 22
total time 132 130 126 120 112 102 90 76 60 42 22
Thus, we see that the maximum occurs with 5 cows, and the total time is 165 seconds = 2 34 minutes.
9. Let p be the price of HMMT. So p = k · xy , where k is a constant to be determined. We know that
when x = 8 and y = 4 that p = 12, so, solving for k, we find that k = 6. So, when x = 4 and y = 8,
we find that p = 6 · 48 =3.
10. On the 12 foot sides, he needs 7 posts, and on the 20 foot sides, he needs 9 posts, so he needs
7 + 9 + 7 + 9 = 32 total posts. Let x be the number of normal fenceposts and y be the number of
strong fenceposts, so x + y = 32. To spend $70, we have the equation 2x + 3y = 70. Substituting, we
find that 2(32 − y) + 3y = 70, so 64 − 2y + 3y = 70, and 64 + y = 70, so y =6.
11. Substituting, we find that 3+n
3−n = 3, so 3 + n = 9 − 3n, thus 4n = 6. So, n = 32 .
12. Yes - one possible path is 4 → 1 → 2 → 4 → 5 → 2 → 3 → 5, so the difference between the start island
and end island is 1.
13. The number of rearrangements keeping 1 number in its spot and rearranging the other 5 such that
none are in the right spot is 44. There are 6 numbers to fix, this gives us an answer of 44 · 6 =264.
14. Feburary 26, 2000 is a Saturday. April 24, 2000 is 23 · 365 + 1 + 1 + 1 + 1 + 1 + 1 = 8401 days away
from April 24, 1977 (including the leap years). So, Feburary 26, 2000 is 8401 − 3 − 31 − 24 = 8343 days
after April 24, 1977. Now, 8343
7 = 1191 with a remainder of 6. So, 6 days before Saturday is Sunday.

15. Notice that 36 = 729 while 54 = 625, and since ln 5 > ln 3, it follows that 54 ln 3 < 36 ln 5, so
53 ∗ 5 ln 3 < 35 ∗ 3 ln 5, and so by laws of logarithms, 53 ln 35 < 35 ln 53 . Again applying laws of
3 5
logarithms, it follows that ln (35 )(5 ) < ln (53 )(3 ) . So, since ln x is an increasing function, it follows
(53 ) (35 ) (35 )
that (35 ) < (53 ) , and so it follows that (53 ) is greater. One can also use the laws of
(54 ) (36 )
exponents to reduce the values to 3 and 5 . The second is clearly larger.
 x x
16. Expressing the total time Joe has biked in hours leads to the equation 20 + 20 + x+2
14 = 1. So, x = 5.
1
Thus, we can construct the diagram below, and find the total time that it takes to get back: 13∗ 78 = 16
hours, or 10 minutes.

5 mi. Friend’s house

5 mi.
13 mi.

shore 7 mi. Gma’s house

12 mi.

17. There are 7! ways to arrange those letters. However, for every distinct arrangement, there are 2!∗2! = 4
total arrangements of the 2 T’s and 2 N’s. Therefore, the total number of distinct ways to arrange the
7!
letters is 2!2! = 1260.
18. The only digits possible are 4, 6, 8, and 9. The only groups of numbers allowed keeping the average at
5 are 8444 are 6464. There are 4 ways to arrange 8444 and 6 ways to arrange 6464 so there are only
10 combinations to try.
19. Let y be the number of coins in the chest. From the problem, we know that y ≡ 5 (mod 11), y ≡ 3
(mod 10), and y ≡ 0 (mod 9). Combining these gives us that y ≡ 423 (mod 990), so the answer is
423.
2
20. The area of the big circle is ( 52 ) π = 25
4 π. The area of the circle with diameter AB is π, and the
area of the circle with diameter BC is 94 π. Thus, the percentage of the big circle that is shaded is
25 9
4 π−π− 4 π 25−4−9 12
25
π
= 25 = 25
= 48%.
4

21. It is clear that we can split the figure into 12 equilateral triangles, all of which have side length s. So,
2
√ √
since the area of one of these triangles is s 4 3 , the total area is 3s2 3.
2
√ √ √
22. Aiso = 2eq so ∆A = Aeq − Aiso = 12 Aeq . Also, Aeq = 4 ·4 3 = 4 3, so ∆A = 2 3. Thus, the area is
A
√
2 3.
23. Notice that 72k+1 ≡ 3 (mod 4) for k ∈ Z (k is an integer). Also, 71 ≡ 7 (mod 100), 72 ≡ 49
(mod 100), 73 ≡ 43 (mod 100), 74 ≡ 1 (mod 100), with the cycle repeating afterwards. So, clearly
77 77
77 ≡ 3 (mod 4), and since the cycle has period 4 (mod 100), we can conclude that 77 ≡ 43
(mod 100).
24. If there are no empty boxes, there are 126 ways of distributing the identical candy. If there is one
empty box, there are 84 ways of distributing the candy and 5 ways of choosing the empty box, so there
are 84 · 5 = 420 ways. If there are two empty boxes, there are 36 ways of distributing the candy and
6 combinations of assigning the empty boxes so that they are not adjacent, so there are 36 · 6 = 216
ways. If there are three empty boxes, there are 9 ways of distributing the candy, and only 1 way of
arranging the empty boxes. Having four or five empty boxes is impossible, since some two would have
to be adjacent. So, the total number of ways is 126 + 420 + 216 + 9 = 771.
25. We can place 9 points as shown, all at least 12 unit apart, but the next point must be less than 12 , so
10 points must be placed. There is no arrangement of 10 points with distance at least 12 . The proof
of this is a simple application of the Pigeonhole Principle.
 √
26. Janet is at (5, −5) and Tim is at (7, 4). They are 85 miles apart.
 Geometry Solutions
Stanford Mathematics Tournament 2000

1. Consider the board labeled as below, with labels for columns and rows. To choose any rectangle on
the board, it is sufficient to choose some number (1-8) of adjacent columns, and some number (1-8) of
adjacent rows, since the rectangle can be created by forming the intersection of the columns and rows.
For instance, the intersection of columns 2,3 and rows 3,4,5 is the rectangle shaded below. So, there
are 8 ways to choose 1 adjacent column, 7 ways to choose 2 adjacent columns, . . ., 1 way to choose 8
adjacent columns, so there are 8 + 7 + 6 + 5 + 4 + 3 + 2 + 1 = 36 total ways to choose the columns,
and 36 ways to choose the rows. Thus, the total number of ways to choose a rectangle (i.e. the total
number of rectangles) is 362 = 1296.

1 2 3 4 5 6 7 8

2. The maximum occurs in an equilateral triangle, in which case the sides a = b − c are given by a2 +
2 2 2
√ √
a 3
√
b + c = 3a = 96, so a = 4 2. Thus, the medians are 3 · ( 2 ) =6 6.
2 2 2 2
3. Since ABCD is a square, we can write (P A) + (P C) = (P B) + (P C) . So, we can substitute in for
P A,P C, and P D to get that P B =24.
4. Notice that in general, when there is a rectangle of side length x and y, the area of the non-triangle
regions (created by drawing a line connecting the midpoint of two opposite lines and a line connecting
two opposite corners - see diagram for examples) is simply 34 of the original area of the box, since the
area of the excluded triangles are 12 · y2 · x2 + 12 · y2 · x2 = 2 · xy xy
2 = 4 , so the desired area is simply
3
ab − 4 = 4 ab, as desired. So, if we consider the four rectangles making up the large rectangle that
ab

are divided this way (the one with dimensions a4 x 3b , etc.), we can say that the total shaded area is 34
of the total area of the rectangle - that is, 34 ab.

5. The area of ABF = 12 bh = 12 · 3h = 32 (3 sin 60◦ ) = 92 sin 60◦ . The area of F CDE = area of ABF
- area of F GC = 92 sin 60◦ − 12 sin 60◦ = 4 sin 60◦ . Therefore, area of ABCDE = area of ABF +
√
9 17 17 3
area of F CDE = 2 sin 60◦ + 4 sin 60◦ = 2 sin 60◦ = 4
.

h1
G
h2
F
b C2
B

E D
 √ √
6. The area of an equilateral triangle with side length s is 43 s2 . Therefore, the side length is 4 2 and
√
height h = 2 6. Now, if r is the radius of the inscribed circle, then r = h3 , since we have an equilateral
triangle. Thus the area is πr2 = 8π
3
.

7. Let’s “resize” the coordinates to be x = x, y  = 2y

3 . This keeps the origin at (0,0), but turns our ellipse
√
into a circle of radius 2. Thus now, the triangle is equilateral, and we can see it now has area√ 3 3.
Once we expand back, we can see we are just multiplying the area by 32 and so the answer is 9 2 3 .
sin A
8. By standard formula, we have that the radius of the inscibed circle, r, is r = ab2a+b (isoceles triangle
that is formed by cutting 
the pyramid vertically in half (cuts the base into 2 equal rectangles)).
2 b2
h2 + ( 2b ) = a2 gives a = h2 + 4 . Also sin A = h
a. Therefore, r =  bh
2
. Note that the
2 h2 + b4 +b
diameter of the cube is the
√ diameter of the 2r
sphere. Let l be the length of the side of the cube, so the
diameter of the cube is l 3 = 2r, so l = √ 3
. So, the volume of the pyramid is 13 b2 h and the cube
√
25 3
volume is l3 . So, the ratio is 6
.

C
a h a
r
A A

b=base

9. A hexagon can be formed by removing any vertex, removing all vertices connected to that vertex, and
then removing any edges connected to any of the removed vertices, and these are the only hexagons in
the diagram. Thus, since there are 10 vertices, there are 10 hexagons in the figure.
10. Consider the flattened version of the situation. Then let O be the center fo the spheres, A be the first
reflection
√ point, B be the point of C1 such that OB is perpendicular to OP . Then since OB = 1,OP =
3,  P BO = 60◦ and P, B, A are collinear,  BAO = 60◦ implies that  P OA = 30◦ . Therefore, each
1
reflection takes the ray 12 around the circle, so there are 11 reflections.
 Team Test Solutions
Stanford Mathematics Tournament 2000

1. .6445 rounds to .645 to .65 to .7. Otherwise .6444... rounds to .644. So the smallest number is .6445.
2. Let c =price, p=purity, d =diameter, h=depth of gold mine, ki =constant. We are given c = k1 p2 d3 ,
2 √ 4 1
p = k2 hc 2 , and d = k3 3 ch. So, c = k1 k22 hc 4 k33 ch3 = k4 c5 h1 . Thus, k4 c4 = h, and c = k5 h 4 . Thus, p
1
varies as h 4 .
3. The sum of the numbers from 700 to 799 is 799·800
2 − 699·700
2 = 74950. The sum of the numbers from 70 to
79·80 69·70
79 is 2 − 2 = 745. So, all numbers that end from 70 to 79 (excluding those starting with 7, since
we counted those from 700 to 799) is 745 · 9 + 10(100 + 200 + . . . + 600 + 800 + 900) = 44705. The sum
of all numbers ending in 7 is 9(7+17+27+37+47+57+67+87+97)+9(100+200+. . .+600+800+900) =
38187. So, the total sum of numbers containing a 7 is 74950 + 44705 + 38186 = 157842.
4. 738,826. This can be arrived at by stepping down, starting with finding how many combinations are
there that begin with a letter other than V or W, and so forth. The answer is 8·9! 4·7!
2·2 + 2 + 4 · 6! + 4 ·
4! + 3! + 2! + 2! = 738826.
5. 0 = cos (α + β) + sin (α − β) = cos α · cos β − sin α · sin β + sin α · cos β − sin β cos α = (cos α + sin α) ·
(cos β − sin β). So cos α + sin α = 0 or cos β − sin β = 0. Then tan α = −1 or tan β = 1. Since tan β is
1
given as 2000 , tan α =−1.
6. Since α3 − α − 1 = 0, then α10 = α8 + α7 . So, we can reduce our expression to 3α8 − 3α6 − 3α5 + 4α4 +
2α3 − 4α2 − 6α − 17. Also, 3α8 − 3α6 − 3α5 = 0, so our expression reduces to 4α4 + 2α3 − 4α2 − 6α − 17.
Also, 4α4 − 4α2 − 4α0 , so our expression reduces to 2α3 − 2α − 17. Now, 2α3 − 2α − 2 = 0, so our
expression reduces to -15, which is our answer.
7. Another 4-digit number that satisfies this property is 9801, since 9801=9*1089.
8. If she has a silver dollar, then she would have too many other coins, as 0 half dollars, 2 quarters, 3
dimes, etc. would be greater than the total. So she has no silver dollars, and at least one of every
other denomination. Continuing, it turns out the only feasible solution is 0 silver dollars, 1 half dollar,
2 quarters, 3 dimes, 4 nickels, 8 pennies, for a total of 18 coins.
4 4
9. It suffices to consider x ≥ 1, since 4(−x) + 1 = 4(x) + 1, and 4(0) + 1 = 1 is not prime. So,
2 2
4x4 + 1 = (4x4 + 4x2 + 1) − 4x2 = (2x2 + 1) − (2x) = (2x2 + 1 − 2x)(2x2 + 1 + 2x). For integers x,
both 2x − 2x + 1 and 2x + 2x + 1 are integers, so this factors 4x4 + 1 unless 2x2 − 2x + 1 = ±1
2 2

or 2x2 + 2x + 1 = ±1. Since x > 0, then 2x2 + 2x + 1 > 1, so we must have 2x2 − 2x + 1 = ±1.
4
2x2 − 2x + 1 = −1 is absurd (4x4 + 1, 2x2 + 2x + 1 > 0, so 2x2 − 2x + 1 = 2x4x +1
2 +2x+1 > 0), so we solve

2x2 − 2x + 1 = 1, or 2x2 − 2x = 0, so x(x − 1) = 0, and x = 0 or x = 1. We have already rejected

4
x = 0, so the only case left is x = 1, or 4(1) + 1 = 5.
10. The second hand crosses the minute hand 59 times an hour. The second hand crosses the hour hand
60 times an hour, except for 2 of the hours, due to the movement of the hour hand. The minute hand
and the hour hand cross 22 times total, because the hour hand completes 2 rotations in a day, and
the minute hand completes 24. The second, hour, and minute hand all coincide only at noon and
midnight, but we’ve counted each of these 12:00’s 3 times instead of once. Therefore, the answer is
59 · 60 + 60 · 24 − 2 + 22 − 2 · 2, giving us 2872 crossings.
11. f (x) is either 0 or something of the form ±xm , where m ≥ 0.
12. A’sposition is (a − Va t, 0) and P ’s position is (0, b − Vb t). So, at time t, the distance between them
is (a − Va t)2 + (b − Vb t)2 . Notice that this distance is the same as the distance between the point
(a, b) and the line (Va t, Vb t), which is the same as the line Vb x − Va y = 0. The distance from a line
Ax + By + C = 0 and (x0 , y0 ) is |Ax√0A +By0 +C|
2 +B 2
, so the answer is |bV
√ a −aV
2
b|
2
Va +Vb

13. Given any
4vertices, there is exactly one intersection of all the diagonals connecting them. So, the
answer is n
4
.
14. x0 if n ≡ 0 (mod 4), 1+x 1
1−x0 if n ≡ 1 (mod 4), − x0 if n ≡ 2 (mod 4),
0 x0 −1
x0 +1 if n ≡ 3 (mod 4).
15. Consider a regular n-gon with radius r. Let x be the side length of the n-gon. So, since the central
angle is 2π
n (see diagram below), use the Law of Cosines to find that x2 = r2 + r2 − 2r ∗ r cos 2π n ,
2 2 2π
√ 2π
so x = 2r (1 − cos n ). Thus, x = r 2 1 − cos n . So, the total perimeter of the n-gon is nx =
√ 
nr 2 1 − cos 2π n . Now, if we take limn→∞ of the perimeter, the result will be 2πn, since the n-gon
√   √
approaches a cirle, so limn→∞ nr 2 1 − cos 2πn = 2πr, and so limn→∞ nr 1 − cos 2π
n
= πr 2.

r r
2π
n
 2001 Stanford Math Tournament
Algebra
1. Find the result of adding seven to the result of forty divided by one-half.
2. Each valve A, B, and C, when open, releases water into a tank at its own constant rate. With all three
valves open, the tank fills in 1 hour, with only valves A and C open it takes 1.5 hours, and with only
valves B and C open it takes 2 hours. How many hours will it take to fill the tank with only valves A
and B open?
3. Julie has a 12 foot by 20 foot garden. She wants to put fencing around it to keep out the neighbor’s
dog. Normal fenceposts cost $2 each while strong ones cost $3 each. If Julie needs one fencepost for
every 2 feet and has $70 to spend on fenceposts, what is the greatest number of strong fenceposts she
can buy?
4. p(x) is a real polynomial of degree at most 3. Suppose there are four distinct solutions to the equation
p(x) = 7. What is p(0)?


N N
5. Let f : → be defined by f (x) =
2, x=0
(f (x − 1))2 , x = 0
What is log2 f (11)?

6. If for three distinct positive numbers x, y, and z,

y x+y x
= = ,
x−z z y

then find the numerical value of x/y.

7. If logA B + logB A = 3 and A < B, find logB A.
1
8. Determine the value of 1 + 2+ 1 .
1+ 1
2+ 1
1+···

9. Find all solutions to (x − 3)(x − 1)(x + 3)(x + 5) = 13.

10. Suppose x, y, z satisfy

x+y+z = 3
2 2 2
x +y +z = 5
x3 + y 3 + z 3 = 7

Find x4 + y 4 + z 4 .
 2001 Stanford Math Tournament
Advanced Topics
ex −e−x
1. Suppose sinh(x) = 2 . What is the inverse function of sinh(x)?
7017 !
2. Write 2010 170 8!
as a decimal in base 6. The subscript indicates the base in which the number is written
(i.e., 2010 is 20 base 10.)
3. There are 36 penguins in a row, and Barbara Manatee is standing in front of them. In general, a
penguin rotation of penguins p1 , p2 , . . . , pn is a rearrangement of them such that p1 moves to where
p2 was standing, and in general pi moves to where pi+1 was standing, and pn moves to where p1
was standing. So, after a penguin rotation, the new order of these penguins is pn , p1 , p2 , . . . , pn−1 .
Whenever Barbara Manatee blows her whistle, the 2-4 penguins go through a penguin rotation, the
5-9 penguins go through a penguin rotation, the 10-16 penguins go through a penguin rotation, the
17-25 penguins go through a penguin rotation, and the 26-36 penguins go through a penguin rotation.
What is the least positive number of whistle blows such that the penguins all return to their original
position?
4. Eleven pirates find a treasure chest. When they split up the coins in it, evenly among all the pirates,
they find that there are 5 coins left. They throw one pirate overboard and split the coins again, only
to find that there are 3 coins left over. So, they throw another pirate over and try again. This time,
the coins split perfectly. What is the least number of coins there could have been?
5. Evaluate 1 · 2−1 + 3 · 2−2 + 5 · 2−3 + 7 · 2−4 + . . .
6. How many subsets of {n | n > 0 and n is a multiple of 3 less than 100} are also subsets of {n | n > 0
and n is a multiple of 4 less than 100}?
7. There are 2000 dots spaced evenly around a circle. If 4 distinct dots A, B, C, and D are picked
randomly, what is the probability that AB intersects CD?
8. Ashley, Bob, Carol, and Doug are rescued from a desert island by a pirate who forces them to play a
game. Each of the four, in alphabetical order by their first names, is forced to roll two dice. If the total
on the two dice is either 8 or 9, the person rolling the dice is forced to walk the plank. The players go
in order until one player loses: A, B, C, D, A, B, .... What is the probability that Doug survives?
9. There are 19 men numbered 1 though 19 and 20 women numbered 1 through 20 entered in a computer
dating service. The computer wants to match every man to a compatible woman, and each man is
only compatible with women who have a number that is greater than equal to his (i.e. man 19 is only
compatible with women 19 and 20, man 18 is only compatible with women 18, 19, 20, etc.). If each
women is matched with at most one man, let n be the number of ways that the computer can match
them. What is the prime factorization of n?
10. David is playing with Legos with velcro attached to the ends. He has green Legos of length 1, blue
Legos of length 2, and red Legos of length 3, and wants to combine them (by attaching them at the
ends) to make a “super-lego” of length 10. If any different ordering of colors is considered a distinct
“super-lego”, how many ways can he make this “super-lego”?
 2001 Stanford Math Tournament
Calculus
d2959
1. What is dx2959 sin x?
2. If f (x) = [x] is the greatest integer function, what is f  (3.7)?
3. Suppose that f is a monotonically increasing continuous function defined on the real numbers. We
2
know that f (0) = 0 and f (2) = 3. Let S be the set of all possible values of 0 f (x) dx. What is the
least upper bound of S?
 5 x2
4. Evaluate −4 |x| dx.
∞ 1
5. 1 x2 +3x =?
2 2
6. Given a point (p, q) on the ellipse xa2 + yb2 = 1 where p = 0, find the x-intercept of the tangent line at
(p, q) in terms of p, a, and b. (Note that a, b = 0.)
7. Find the number of real solutions to sin(6πx) = x, where x is in radians.
8. If a, b, and c are positive real numbers such that a + b + c = 16 and a2 + b2 + c2 = 160, what is the
maximum possible value of abc?
9. For a given k > 0, n ≥ 2k > 0, consider the square R in the plane consisting of all points (x, y) with
k ≤ x + y, and blue otherwise. Find the area of the gray
0 ≤ x, y ≤ n. Color each point in R gray if xy
region in terms of n and k.
 n
10. Let f (x) = x2 − 1 , where n is a positive integer. Determine, in terms of n, (a, b, c), where a,b,
and c are the number of distinct roots of f (n) (x) in the intervals (−∞, −1), (−1, 1), and (1, +∞),
respectively.
 2001 Stanford Math Tournament
Geometry
1. Find the coordinates of the points of intersection of the graphs of the equations y = |2x| − 2 and
y = −|2x| + 2.
2. Jacques is building an igloo for his dog. The igloo’s inside and outside are both perfectly hemispherical.
254
The interior height at the center is 2 feet. The igloo has no door yet and contains 2187 π cubic yards
of hand-packed snow. What is the circumference of the igloo at its base in feet?
3. Find the area of the convex quadrilateral whose vertices are (0, 0), (4, 5), (9, 21), (−3, 7).
4. E is a point in the interior of rectangle ABCD. AB = 6, triangle ABE has area 6, and triangle CDE
has area 12. Find (EA)2 − (EB)2 + (EC)2 − (ED)2 .
5. Two identical cones, each 2 inches in height, are held one directly above another with the pointed end
facing down. The upper cone is completely filled with water. A small hole is punctured in the bottom
of the upper cone so that the water trickles down into the bottom cone. When the water reaches a
depth of 1 inch in the bottom cone, what is its depth in the upper cone?

6. Find the radius of a circle inscribed in the triangle determined by the lines 4x + 3y = 24, 56x − 33y =
−264, and 3x − 4y = 18.
7. In the figure, AB is tangent at A to the circle with center O; point D is interior to the circle; and DB
intersects the circle at C. If BC = DC = 3, OD = 2 and AB = 6, then find the radius of the circle.

O
C

A B

8. Let S be the solid tetrahedron with boundary points (0, 0, 0), (2, 4, 0), (5, 1, 0), (3, 2, 10). Let z1 =
max{q | ( 12 23 9 5
5 , 10 , q) ∈ S} and let z2 = max{r | ( 2 , 4 , r) ∈ S}. Find z1 − z2 .

9. Circles A and B are tangent and have radii 1 and 2, respectively. A tangent to circle A from the
point B intersects circle A at C. D is chosen on circle B so that AC is parallel to BD and the two
segments BC and AD do not intersect. Segment AD intersects circle A at E. The line through B and
E intersects circle A through another point F . Find EF .
10. E is a point inside square ABCD such that ∠ECD = ∠EDC = 15◦ . Find ∠AEB.
 2001 Stanford Math Tournament
General Test
1. If a  b is defined as 2a − ba , what value is associated with 3  2?
2. Find the coordinates of the points of intersection of the graphs of the equations y = |2x| − 2 and
y = −|2x| + 2.
3. ABCD is a 4-digit number. What is the largest number that can be formed with AB a prime 2-digit
number and C and D each a prime 1-digit number?
4. Let A be the set of all non-composite positive integers. Let B be the set of all squares of integers. Let
C be the set of all multiples of 3. What is the size of A ∩ (B ∪ C)?
5. What is i2959 ?
6. The points Q = (9, 14) and R = (a, b) are symmetric with respect to the point P = (5, 3). What are
the coordinates of point R?
7. If F (x) = 3x3 − 2x2 + x − 3, find F (1 + i).
8. Express the absolute value of the difference between 0.36 and 0.36 as a common fraction.
9. Jonathan chooses 10 cards without replacement from a standard 52-card deck of cards (without jokers).
What is the probability that he does not draw the 3 of clubs and he does not draw the King of spades?
10. Julie has a 12 foot by 20 foot garden. She wants to put fencing around it to keep out the neighbor’s
dog. Normal fenceposts cost $2 each while strong ones cost $3 each. If Julie needs one fencepost for
every 2 feet and has $70 to spend on fenceposts, what is the largest number of strong fenceposts she
can buy?
11. Anne has a cube that she wants to paint. If she decides to paint each face a different color and she has
6 colors, how many distinct ways can she paint the cube? (If one painted cube can be rotated to get
another, they are the same.)
12. Find, in degrees, the sum of angles 1, 2, 3, 4, and 5 in the star-shaped figure shown.

2
3

5 4

13. Each valve A, B, and C, when open, releases water into a tank at its own constant rate. With all three
valves open, the tank fills in 1 hour, with only valves A and C open it takes 1.5 hours, and with only
valves B and C open it takes 2 hours. How many hours will it take to fill the tank with only valves A
and B open?
14. Lattice paths are paths consisting of one unit steps in the positive horizontal or vertical directions.
How many distinct lattice paths are there from the origin to the point (5, 4)?
15. Three circles, each of area 4π, are all externally tangent. Their centers form a triangle. What is the
area of the triangle?
16. Two equilateral triangles sharing an edge have a combined area of π. What is the square of the length
of their shared edge?
17. Frogger wants to cross a stream. He starts on one bank and jumps onto the first passing log, which is
traveling at 9 feet per second. The following logs are at speeds such that the second is twice is fast as
the first, the third is twice as fast as the second, the fourth is one-third as fast as the third, and the
last (fifth) is one-third as fast as the fourth (all of the logs travel in the same direction). If there is a
2 second interval between jumps, then how far does he travel down river from his original point, once
he reaches the opposite bank?
18. Rice tuition was $7200 one semester. Each student takes a minimum of 12 credits and a maximum of
20 credits. For each credit, a class meets for 3 hours a week. A semester is 15 weeks with no breaks.
What is the difference in cost in dollars per hour of class that a student taking the maximum load and
a student taking the minimum load pays, to the nearest penny?
19. Suppose a1 , a2 , a3 , . . . is a sequence of numbers such that a1 = 1 and an+1 = an + (2n + 1) for all
positive integers n. Find a20 .
20. My spouse and I have 9 kids. Each child gets married and has exactly 9 children of their own. By the
time I have 100,000 descendants (including every generation) what is the longest title that applies to
me? (Assuming no married blood relatives.)

a. grandparent
b. great grandparent
c. great great grandparent
d. great great great grandparent
e. great great great great grandparent
f. great great great great great grandparent
g. great great great great great great grandparent

21. Find the greatest integer x for which 320 > 32x .
22. If x + y = xy, with x and y real, what value can x not have?
23. Instead of using two standard cubical dice in a board game, three standard cubical dice are used so
that the game goes more quickly. In the regular game, doubles are needed to get out of the “pit”. In
the revised game, doubles or triples will get you out. How many times as likely is it for a player to get
out of the “pit” on one toss under the new rules as compared to the old rules?
24. A circle is drawn with center at the origin and radius 2.5. Find the coordinates of all intersections
of the circle with an origin-centered square of side length 4 whose sides are parallel to the coordinate
axes.
(7017 )!
25. Write 2010 (170 8 )!
as a decimal in base 6. The subscript indicates the base in which the number is
written (i.e., 156 is 15 base 6, so 156 = 1110 )
26. Jacques is building an igloo for his dog. The igloo’s inside and outside are both perfectly hemispherical.
254
The interior height at the center is 2 feet. The igloo has no door yet and contains 2187 π cubic yards
of hand-packed snow. What is the circumference of the igloo at its base in feet?
27. How many solutions are there to x7 + y 7 = z 7 with x, y, z real?
28. Find all prime factors of 318 − 218 .
√
29. If a = 1 and a 10000a = 10a , find a. The subscript indicates the base in which the number is written.
30. Suppose 100 students have at least one of math, applied math, or statistics as one of their majors.
There are 24 statistics majors, 46 applied math majors, and 55 pure math majors. There are 14 who
are at least pure & applied math, 10 at least applied math & statistics and 7 at least statistics & pure
math. How many triple majors (people with all three majors) are there, if any?
31. The sum of 3 real numbers is 0. If the sum of their cubes is π e , what is their product?

x+3
32. Given a real-valued function f (x) = |2x+6| , what is the least integer value which lies in the domain
of the function?
33. E is a point in the interior of rectangle ABCD. AB = 6, triangle ABE has area 6, and triangle CDE
has area 12. Find (EA)2 − (EB)2 + (EC)2 − (ED)2 .
34. What is the radius of the incircle of a right triangle with legs of lengths 7 and 24?
35. If logA B + logB A = 3 and A < B, find logB A.
36. ABC is an equilateral triangle with edge of length 256. Let n be the maximum number of non-
overlapping equilateral triangles of edge length 12 that can be fit into ABC. What is the prime
factorization of n?
 
37. If x2 + x12 = 7, find x3 + x13 .
38. Eleven pirates find a treasure chest. When they split up the coins in it evenly among all the pirates,
they find that there are 5 coins left. They throw one pirate overboard and split the coins again, only
to find that there are 3 coins left over. So, they throw another pirate over and try again. This time,
the coins split perfectly. What is the least number of coins there could have been?
 2001 Stanford Math Tournament
Team Test
1. ABCD is a square with sides of unit length. Points E and F are taken on sides AB and AD respectively
so that AE = AF and the quadrilateral CDF E has maximum area. What is this maximum area?
2. How many positive integers between 1 and 400 (inclusive) have exactly 15 positive integer factors?
3. Find the 2000th positive integer that is not the difference between any two integer squares.
4. For what values of a does the system of equations

x2 = y 2 , (x − a)2 + y 2 = 1
have exactly 2 solutions?
5. What quadratic polynomial whose coefficient of x2 is 1 has roots which are the complex conjugates of
the solutions of x2 − 6x + 11 = 2xi − 10i? (Note that the complex conjugate of a + bi is a − bi, where
a and b are real numbers.)
6. Find the least n such that any subset of {1, 2, . . . , 100} with n elements has 2 elements with a difference
of 9.
7. The median to a 10 cm side of a triangle has length 9 cm and is perpendicular to a second median of
the triangle. Find the exact value in centimeters of the length of the third median.
8. Janet and Donald agree to meet for lunch between 11:30 and 12:30. They each arrive at a random
time in that interval. If Janet has to wait more than 15 minutes for Donald, she gets bored and leaves.
Donald is busier so will only wait 5 minutes for Janet. What is the probability that the two will eat
together? Express your answer as a fraction.
9. What is the minimum number of straight cuts needed to cut a cake in 100 pieces? The pieces do not
need to be the same size or shape but cannot be rearranged between cuts. You may assume that the
cake is a large cube and may be cut from any direction.
10. You know that the binary function  takes in two non-negative integers and has the following properties:
• 0a=1
• aa=0
• If a < b, then a  b = (b − a)[(a − 1)  (b − 1)].
Find a general formula for x  y, assuming that y ≥ x > 0.
11. Christopher and Robin are playing a game in which they take turns tossing a circular token of diameter
1 inch onto an infinite checkerboard whose squares have sides of 2 inches. If the token lands entirely
in a square, the player who tossed the token gets 1 point; otherwise, the other player gets 1 point. A
player wins as soon as he gets two more points than the other player. If Christopher tosses first, what
is the probability that he will win? Express your answer as a fraction.
12. A binary string is a string consisting of only 0’s and 1’s (for instance, 001010, 101, etc.). What is
the probability that a randomly chosen binary string of length 10 has 2 consecutive 0’s? Express your
answer as a fraction.
13. You have 2 six-sided dice. One is a normal fair die, while the other has 2 ones, 2 threes, and 2 fives.
You pick a die and roll it. Because of some secret magnetic attraction of the unfair die, you have a
75% chance of picking the unfair die and a 25% chance of picking the fair die. If you roll a three, what
is the probability that you chose the fair die?

14. Find the prime factorization of 1≤i<j≤100 ij.
15. Let ABC be an isosceles triangle with ∠ABC = ∠ACB = 80◦ . Let D be a point on AB such that
∠DCB = 60◦ and E be a point on AC such that ∠ABE = 30◦ . Find ∠CDE in degrees.
 2001 Stanford Math Tournament
Advanced Topics
ex −e−x
1. Suppose sinh(x) = 2 . What is the inverse function of sinh(x)?
√
Solution: ln(x + x2 + 1)
ex −e−x y −y
y= 2 Switch x and y to find the inverse, so x = e −e
. 2 , or 2x = ey − e−y , so 2xey = e2y − 1,
√
2 √
and e − 2xey − 1 = 0. By using the quadratic formula, we see that ey = 2x± 24x +4 = x ± x2 + 1,
2y
√ √ √
so y = ln(x + x2 + 1) (since y cannot be ln(x − x2 + 1), since x − x2 + 1 is always negative, and
ln is not defined for negative numbers).
7017 !
2. Write 2010 170 8!
as a decimal in base 6. The subscript indicates the base in which the number is written
(i.e., 2010 is 20 base 10.)
Solution: 0.16 or 0.1
The easiest way to solve this problem is to convert everything into base 10.

7017 = 7 × 171 + 0 × 170 = 119

1708 = 1 × 82 + 7 × 81 + 0 × 80 = 120
7017 ! 119! 1
2010 = 20 =
1708 ! 120! 6

1
Converting to base 6, 6 = 0.16 .
3. There are 36 penguins in a row, and Barbara Manatee is standing in front of them. In general, a
penguin rotation of penguins p1 , p2 , . . . , pn is a rearrangement of them such that p1 moves to where
p2 was standing, and in general pi moves to where pi+1 was standing, and pn moves to where p1
was standing. So, after a penguin rotation, the new order of these penguins is pn , p1 , p2 , . . . , pn−1 .
Whenever Barbara Manatee blows her whistle, the 2-4 penguins go through a penguin rotation, the
5-9 penguins go through a penguin rotation, the 10-16 penguins go through a penguin rotation, the
17-25 penguins go through a penguin rotation, and the 26-36 penguins go through a penguin rotation.
What is the least positive number of whistle blows such that the penguins all return to their original
position?
Solution: 3465
Since we have groups of 3,5,7,9,11 penguins rotating, for them all to be in the right place requires
the number of whistle blows to be a multiple of the least common multiple of 3,5,7,9,11 (denoted
[3, 5, 7, 9, 11]). So, we calculate [3, 5, 7, 9, 11] = [5, 7, 9, 11] = 5 · 7 · 9 · 11 = 3465.
4. Eleven pirates find a treasure chest. When they split up the coins in it, evenly among all the pirates,
they find that there are 5 coins left. They throw one pirate overboard and split the coins again, only
to find that there are 3 coins left over. So, they throw another pirate over and try again. This time,
the coins split perfectly. What is the least number of coins there could have been?
Solution: 423
Let n be the number of coins in the chest. Then, n ≡ 5 (mod 11) (this means that the remainder of
n when divided by 11 is 5), n ≡ 3 (mod 10), and n ≡ 0 (mod 9). The Chinese Remainder Theorem
gives an elegant solution to this problem, or you can solve it as follows. Since n ≡ 3 (mod 10),
we know that the last digit of n must be 3. Also, since n ≡ 0 (mod 9), n is divisible by 9, so
n = 9q for some q. Since n ends in a 3, q must end in a 7. That means that q looks like 10p + 7, so
n = 9(10p + 7) = 90p + 63. Now, you can just try p = 1, 2, . . ., and you’ll find that p = 4 works.
5. Evaluate 1 · 2−1 + 3 · 2−2 + 5 · 2−3 + 7 · 2−4 + . . .
Solution: 3
 We will express this as a sum of geometric series as follows:

∞


1 · 2−1 + 3 · 2−2 + 5 · 2−3 + 7 · 2−4 + . . . = (2n − 1)2−n
n=1

∞ ∞


= (2n)2−n − 2−n
n=1 n=1

∞
1/2
= n2−n+1 −
n=1
1 − 1/2

∞
n
= 2−n+1 − 1
n=1 j=1

∞
∞
= 2−n+1 − 1
j=1 n=j

∞
2−j+1
= −1
j=1
1 − 1/2
∞


= 2−j+2 − 1
j=1
2
= −1
1 − 1/2
= 3

6. How many subsets of {n | n > 0 and n is a multiple of 3 less than 100} are also subsets of {n | n > 0
and n is a multiple of 4 less than 100}?
Solution: 256
Let A = {n | n > 0 and n is a multiple of 3 less than 100} and B = {n | n > 0 and n is a multiple of
4 less than 100}. Then, A ∩ B consists of all multiples of 12 less than 100; there are 8 of these. The
subsets of A that are also subsets of B are exactly the subsets of A ∩ B; since A ∩ B has 8 elements,
it has 28 = 256 subsets.

7. There are 2000 dots spaced evenly around a circle. If 4 distinct dots A, B, C, and D are picked
randomly, what is the probability that AB intersects CD?
1
Solution: 3
Randomly pick the 4 dots A, B, C, and D. Now, if we start from A and go clockwise, there are 6
equally likely possibilities for the order that the other 3 dots will appear:

B, C, D C, B, D D, B, C
B, D, C C, D, B D, C, B

AB and CD intersect if and only if B is the second of these three dots; this happens with probability
1
3.

8. Ashley, Bob, Carol, and Doug are rescued from a desert island by a pirate who forces them to play a
game. Each of the four, in alphabetical order by their first names, is forced to roll two dice. If the total
on the two dice is either 8 or 9, the person rolling the dice is forced to walk the plank. The players go
in order until one player loses: A, B, C, D, A, B, .... What is the probability that Doug survives?
Solution: 148/175
 We will instead find the probability that Doug doesn’t survive. Notice that the only time Doug doesn’t
survive is if all four survive the first n rounds (where a round is once through all four), and then Alice,
Bob, and Carol survive the n + 1-st round but Doug does not. The probability of rolling an 8 or 9 is
5 4 1
 4
36 + 36 = 4 , so the probability that all four survive a round is 34 = 256 81
. So, all four survive the
 81 n
first n rounds with probability 256 . The probability that Alice, Bob, and Carol survive the n + 1-st
 3
round but Doug does not is 34 14 = 256 27
. So, the probability that Doug has to walk the plank in the
27
 
81 n
n + 1-st round is 256 256 . Thus, the probability that Doug doesn’t survive is

∞  n
27 81 27 1 27
= 81 = .
n=0
256 256 256 1 − 256 175

148
So, the probability that Doug survives is 175 .

9. There are 19 men numbered 1 though 19 and 20 women numbered 1 through 20 entered in a computer
dating service. The computer wants to match every man to a compatible woman, and each man is
only compatible with women who have a number that is greater than equal to his (i.e. man 19 is only
compatible with women 19 and 20, man 18 is only compatible with women 18, 19, 20, etc.). If each
women is matched with at most one man, let n be the number of ways that the computer can match
them. What is the prime factorization of n?
Solution: 219
There are 2 ways to match man 19, with either woman 19 or woman 20. Given that, there are 2 ways
to match man 18, with either woman 18 or with the one that man 19 didn’t choose. This continues all
the way back to man 1, so there are 2 choices for each of the 19 men, so the total number of matchings
is 219 .
10. David is playing with Legos with velcro attached to the ends. He has green Legos of length 1, blue
Legos of length 2, and red Legos of length 3, and wants to combine them (by attaching them at the
ends) to make a “super-lego” of length 10. If any different ordering of colors is considered a distinct
“super-lego”, how many ways can he make this “super-lego”?
Solution: 274
Let an be the number of ways to make a super-lego of length n. Then a1 is of course 1, a2 = 2 (we
can have either a lego of length 2 or 2 legos of length 1), and a3 = 4. Now, observe that to get a
super-lego of length n, we must have either taken a super-lego of length n − 3 and attached a lego of
length 3, a super-lego of length n − 2 and attached a lego of length 2, or a super-lego of length n − 1
and attached a lego of length 1. So, an = an−3 + an−2 + an−1 . Using this relationship, it’s easy to find
that a10 = 274.
 2001 Stanford Math Tournament
Algebra
1. Find the result of adding seven to the result of forty divided by one-half.
Solution: 87
Since we are dividing forty by one-half, the result is 80 (not 20), and adding 7 gives 87.
2. Each valve A, B, and C, when open, releases water into a tank at its own constant rate. With all three
valves open, the tank fills in 1 hour, with only valves A and C open it takes 1.5 hours, and with only
valves B and C open it takes 2 hours. How many hours will it take to fill the tank with only valves A
and B open?
6
Solution: 1.2 = 5
Let rA , rB , rC be the rates of valves A, B, and C in tanks per hour, respectively. Then, 1(rA +rB +rC ) =
1, 1.5(rA + rC ) = 1, and 2(rB + rC ) = 1. So, rA + rB + rC = 1, rA + rC = 23 , and rB + rC = 12 . So,
rA + rB = 2(rA + rB + rC ) − (rA + rC ) − (rB + rC ) = 2 − 23 − 12 = 56 . Thus, it takes 65 = 1.2 hours to
fill the tank with only valves A and B open.
3. Julie has a 12 foot by 20 foot garden. She wants to put fencing around it to keep out the neighbor’s
dog. Normal fenceposts cost $2 each while strong ones cost $3 each. If Julie needs one fencepost for
every 2 feet and has $70 to spend on fenceposts, what is the greatest number of strong fenceposts she
can buy?
Solution: 6
Let x be the number of normal fenceposts and y be the number of strong fenceposts. The perimeter
of Julie’s garden is 64 feet, and she needs one fencepost every 2 feet, so x + y = 32. The total cost of
the fenceposts is 2x + 3y, and we want 2x + 3y ≤ 70. Since 2x + 2y = 64, we find that y ≤ 6.
4. p(x) is a real polynomial of degree at most 3. Suppose there are four distinct solutions to the equation
p(x) = 7. What is p(0)?
Solution: 7
Since the degree of p(x) − 7 is at most 3, it has at most 3 roots. But we have 4 solutions, so p(x) − 7
must be constant and equal to zero. So, p(x) = 7, and thus p(0) = 7.

N N
5. Let f : → be defined by f (x) =
2, x=0
(f (x − 1))2 , x = 0
What is log2 f (11)?

Solution: 2048
For x = 0, log2 f (x) = 2 log2 f (x − 1). For x = 0, log2 f (x) = 1. So, in general, log2 f (x) = 2x . Thus,
log2 f (11) = 211 = 2048.
6. If for three distinct positive numbers x, y, and z,
y x+y x
= = ,
x−z z y
then find the numerical value of x/y.
Solution: 2
2 2 2
z , yz = x + xy − xz − yz. Since x−z = y , y = x − xz. Finally, since
y
Since x−z = x+y y x x+y
z = x
y,
2 2
xz = yx + y = yx + x − xz = yz + yz = 2yz. So, x/y = 2.
7. If logA B + logB A = 3 and A < B, find logB A.
√
3− 5
Solution: 2
1
Let x = logB A. Then B x = A, so B = A x . Thus, logA B = x1 . So, x + x1 = 3. Multiplying through
√
by x, x2 + 1 = 3x, so x2 − 3x + 1 = 0. By the quadratic formula, x = 3±2 5 Since A < B, logB A < 1,
√
so x = 3−2 5
 1
8. Determine the value of 1 + 2+ 1 .
1+ 1
2+ 1
1+···
√
1+ 3
Solution: 2
1 1 x 3x+1
Let x = 1 + 2+ 1 . Then, observe that x = 1 + 1
2+ x
, so x = 1 + 2x+1 = 2x+1 . Therefore,
1+ 1
2+ 1
1+···
√ √ √
2± 4+8 2±2 3 1± 3
2x2 + x = 3x + 1, so 2x2 − 2x − 1 = 0. By the quadratic formula, x = 4 = 4 = 2 . Now,
√
1 + 2+ 1 1 is clearly positive, so it must be 1+2 3 .
1+ 1
2+ 1
1+···

9. Find all solutions to (x − 3)(x − 1)(x + 3)(x + 5) = 13.

√ √
Solution: ± 3 − 1, ± 17 − 1
Let y = x+1, so x = y−1, and our new equation is (y−4)(y−2)(y+2)(y+4) = 13, or (y 2 −16)(y 2
√ −4) √=
4 2 2 2
13, and by√expanding,
√ we get that y − 20y + 51 = 0, so (y − 3)(y − 17) = 0, so y = ± 17, ± 3,
and x = ± 3 − 1, ± 17 − 1.
10. Suppose x, y, z satisfy

x+y+z = 3
x2 + y 2 + z 2 = 5
x3 + y 3 + z 3 = 7

Find x4 + y 4 + z 4 .
Solution: 9
We have

9 = (x + y + z)2 = x2 + y 2 + z 2 + 2(xy + yz + xz) = 5 + 2(xy + yz + xz)

so xy + yz + xz = 2. Similarly,

15 = (x2 + y 2 + z 2 )(x + y + z)
= x3 + y 3 + z 3 + x2 y + y 2 x + y 2 z + z 2 y + z 2 x + x2 z
= 7 + xy(x + y) + yz(y + z) + zx(z + x)
= 7 + xy(3 − z) + yz(3 − x) + zx(3 − y) [since x + y + z = 3]
= 7 + 3(xy + yz + zx) − 3xyz
= 13 − 3xyz, since xy + yz + zx = 2
−2
So, xyz = 3 . Finally,

21 = (x3 + y 3 + z 3 )(x + y + z)
= x4 + y 4 + z 4 + x3 y + y 3 x + y 3 z + z 3 y + z 3 x + x3 z
= x4 + y 4 + z 4 + xy(x2 + y 2 ) + yz(y 2 + z 2 ) + zx(z 2 + x2 )
= x4 + y 4 + z 4 + xy(5 − z 2 ) + yz(5 − x2 ) + zx(5 − y 2 )
= x4 + y 4 + z 4 + 5(xy + yz + zx) − xyz(x + y + z)
= x4 + y 4 + z 4 + 12

So, x4 + y 4 + z 4 = 9.
 2001 Stanford Math Tournament
Calculus
d2959
1. What is dx2959 sin x?
Solution: − cos x
Notice that the fourth derivative of sin x is again sin x, and 2959 = 4 × 739 + 3. So, the 2959-th
derivative of sin x is the same as the third derivative of sin x, which is just − cos x.
2. If f (x) = [x] is the greatest integer function, what is f  (3.7)?
Solution: 0
Notice that f is constant on the interval (3, 4), so f  = 0 on (3, 4). Thus, f  = 0 on (3, 4), so
f  (3.7) = 0.
3. Suppose that f is a monotonically increasing continuous function defined on the real numbers. We
2
know that f (0) = 0 and f (2) = 3. Let S be the set of all possible values of 0 f (x) dx. What is the
least upper bound of S?
Solution: 6
Since f is monotonically increasing, we know that for all x ≤ 2, f (x) ≤ 3. Also, for all x ≥ 0, f (x) ≥ 0.
2
So, 6 is an upper bound of 0 f (x) dx. Now, we will show that it is the least upper bound. Consider
functions fn defined by 
3nx, x ≤ n1
fn (x) =
3, x > n1
2 1 2 3
Then, 0 f (x) dx = 0n 3nx dx + 1 3 dx = 2n + 6 − n3 = 6 − 2n3
, so we have found monotonically
n
2
increasing functions whose integrals are arbitrarily close to 6. Thus, the least upper bound of 0 f (x) dx
is 6.
 5 x2
4. Evaluate −4 |x| dx.
41
Solution: 2
= 20.5 = 20 21
For x ≤ 0, |x| = −x. For x ≥ 0, |x| = x. So,
 5  0  5 2  0  5
x2 x2 x
dx = dx + dx = −x dx + x dx
−4 |x| −4 −x 0 x −4 0
0 5
1 1 1 1
= − x2 + x2 = −0 + · 16 + · 25 − 0
2 −4 2 0 2 2
25
= 8+
2
and this is 41
2 .
∞ 1
5. 1 x2 +3x =?
1 ln 4
Solution: 3
ln 4 = 3
1
u
1 1

Solution: The integral given is equal to 3 limu→∞ 1 x − x+3
1
= 3 limu→∞ (ln u − (ln(u + 3) − ln 4))
1
= 3 ln 4.
2 2
6. Given a point (p, q) on the ellipse xa2 + yb2 = 1 where p = 0, find the x-intercept of the tangent line at
(p, q) in terms of p, a, and b. (Note that a, b = 0.)
a2
Solution: p
 2
If q = 0, then we are either at the point (a, 0) or (−a, 0), so in either case, we see that ap gives the
correct answer.
 Otherwise, suppose q > 0. Then, we are in the top half of the ellipse, which has
x2 b2
equation y = b2 − a2 .
dy
Remember that the slope of the tangent line at (p, q) is just dx at (p, q).

2 2
−1/2 2 2 2
By the chain rule, dy
dx = − 12 b2 − xa2b 2xb
a2 = − ya2 . So, at (p, q), dx = − qa2 . This means that
xb dy pb

2
the tangent line at (p, q) looks like y = − qa
pb
2 x + c for some c. We know that (p, q) is a point on the
2
pb2
line, so q = − qa
pb
2 p + c, and this tells us that c = q + qa2 p. So, the equation of the tangent line is
pb2 pb2
y= − qa2x +q+ qa2 p.
2
pb2
So, the x-intercept of the tangent line is the x satisfying 0 = − qa
pb
2x + q + qa2 p. Solving this linear
q2 a2 +p2 b2
equation, we find that x = pb2 . However, this is not yet in terms of p, a, and b. Remember that
2 2
(p, q) is a point on the elliptic curve, so ap2 + qb2 = 1. This means that p2 b2 + q 2 a2 = a2 b2 ; substituting
2 2 2
this into the numerator of x, we get x = apbb2 = ap .
Now, notice that the tangent line to (p, −q) is just the tangent line of (p, q) reflected over the x-axis,
so it has the same x-intercept.
7. Find the number of real solutions to sin(6πx) = x, where x is in radians.
Solution: 11
1
There is one solution at x = 0. Now, consider the graph of sin(6πx)−x. At 12 , sin(6πx) = sin(π/2) = 1,
3 1
so sin(6πx) − x > 0. At 12 , sin(6πx) − x < 0, so there is a solution to sin(6πx) = x between 12 and
3 3 5 5 7 7 9 9 11
12 . Similarly, there are solutions between 12 and ,
12 12 and ,
12 12 and 12 , and 12 and 12 , giving
5 solutions with x positive. There is not a solution between 11 11
12 and 1, since sin(6π · 12 ) = −1 and
sin(6π) = 0, and there are no solutions for x > 1, since then x > 1 > sin(6πx). Since the graph of
sin(6πx) − x is symmetric around the origin, there must be 5 negative solutions as well, for a total of
11 solutions.
8. If a, b, and c are positive real numbers such that a + b + c = 16 and a2 + b2 + c2 = 160, what is the
maximum possible value of abc?
√
Solution: 12827
(7 7 − 10)
First, notice that a+b = 16−c and a2 +b2 = 160−c2 , so 2ab = (a+b)2 −(a2 +b2 ) = (16−c)2 −(160−c2 ) =
2c2 − 32c + 96, and ab = c2 − 16c + 48. Then, abc = c3 − 16c2 + 48c. To find the maximum, we want
the derivative of c3 − √16c2 + 48c to be√
0, so we want 3c2 − 32c + 48 = 0. Using the quadratic formula,
16±4 7 16+4 7
we find that c = 3 . If c = 3 , then abc < 0, which is clearly not the maximum. So, we take
16−4 7
√ √
c= 3 . Substituting into abc = c − 16c2 + 48c, we find that abc = 128
3
27 (7 7 − 10).

9. For a given k > 0, n ≥ 2k > 0, consider the square R in the plane consisting of all points (x, y) with
0 ≤ x, y ≤ n. Color each point in R gray if xy
k ≤ x + y, and blue otherwise. Find the area of the gray
region in terms of n and k.


Solution: 2k n + k ln n−k k
2 2
k ≤ x+y iff xy ≤ kx+ky iff xy−kx−ky ≤ 0 iff (x−k)(y−k) = xy−kx−ky+k ≤ k .
First, observe that xy
2
So, we really want to find the area of the part of (x − k)(y − k) ≤ k that is contained in R. Notice that
the graph (x − k)(y − k) = k 2 is just a hyperbola, and the bottom half of the hyperbola lies entirely
outside the first quadrant, so we only need to look at the top half of the hyperbola.

(k, 0)
 k2 k2
The graph of the top half is y = x−k + k, which is strictly decreasing. Notice that x−k + k = n for
k2 k2
x= n−k +k = nk
n−k , so x−k + k ≥ n for all x ≤ nk
n−k . Thus, we can express the gray area as
 nk   
n−k n
k2 n2 k
n dx + +k dx = + (k 2 ln(x − k) + kx)|nnk
0 nk
n−k
x−k n−k n−k

n2 k k2 nk
= + [k 2 ln(n − k) + kn] − (k 2 ln +k )
n−k n−k n−k
k2
= 2kn + k 2 ln(n − k) − k 2 ln
n−k
= 2kn + k 2 ln(n − k) − 2k 2 ln k + k 2 ln(n − k)
 
n−k
= 2k n + k ln
k
 n
10. Let f (x) = x2 − 1 , where n is a positive integer. Determine, in terms of n, (a, b, c), where a,b,
and c are the number of distinct roots of f (n) (x) in the intervals (−∞, −1), (−1, 1), and (1, +∞),
respectively.
Solution: (0, n, 0)
A polynomial of degree k has at most k roots and differentiating a polynomial decreases the multiplicity
of each existing root by 1. Also, the derivative of any differentiable function with two different real 0s
has a real 0 strictly between them by Rolle’s theorem (the mean value theorem). This means that for
0 ≤ k ≤ n, the kth derivative of f has roots of multiplicity n − k at -1 and 1 and k roots of multiplicity
1 in (−1, 1). So, for k = n, we see that there are roots of multiplicity 0 (i.e. no root) at -1 and 1, and
n roots in (−1, 1). Since this polynomial is of degree n, these must be the only roots.
 2001 Stanford Math Tournament
Geometry
1. Find the coordinates of the points of intersection of the graphs of the equations y = |2x| − 2 and
y = −|2x| + 2.
Solution: (1, 0), (−1, 0)
(x, y) is a point of intersection iff y = |2x| − 2 = −(|2x| − 2), so y must be 0. Then, |2x| = 2, so x = ±1.
2. Jacques is building an igloo for his dog. The igloo’s inside and outside are both perfectly hemispherical.
254
The interior height at the center is 2 feet. The igloo has no door yet and contains 2187 π cubic yards
of hand-packed snow. What is the circumference of the igloo at its base in feet?
14
Solution: 3
π
Let R be the radius of the igloo at its base, in feet. Then, the amount of snow in the igloo is, in cubic
feet, 12 43 πR3 − 43 π · 23 = 23 π(R3 − 8). This is equal to 2187
254
π cubic yards, or 254 3 254
81 = 3 · 2187 cubic
2 3 254 3 343 7
feet. Therefore, 3 π(R − 8) = 81 π, so R = 27 , and R = 3 . So, the circumference of the igloo is
2πR = 14 3 π feet.

3. Find the area of the convex quadrilateral whose vertices are (0, 0), (4, 5), (9, 21), (−3, 7).
165
Solution: 2
We can split this into two triangles, one with vertices (0, 0), (4, 5), (9, 21) and the other with vertices
(0, 0), (9, 21), (−3, 7). To find the area, we will first find a formula for the area of a triangle with vertices
(0, 0), (a, b), (c, d) and then apply this formula to our two triangles.
(c, d)

(a, b) C

A B

As the figure demonstrates, we can find the area of the triangle by first finding the area of the large
triangle (the one with vertices (0, 0), (a, b), (a, 0)) and then subtracting the area of the regions A, B,
and C. The large triangle clearly has area cd 2 . Region A has area 2 , region B has area (c − a)b, and
ab

region C has area (c−a)(d−b)

2 . So, the area of the triangle is cd 2 − 2 − (c − a)b −
ab (c−a)(d−b)
2 = ad−bc
2 .
Of course, the area should only be positive, so we need to take the absolute value of this. (If you’re
familiar with determinants, think about what this means in terms of determinants.)
Using the formula, the area of the triangle with vertices (0, 0), (4, 5), (9, 21) is |4·21−5·9|
2 = 39
2 and the
|9·7+21·3|
area of the triangle with vertices (0, 0), (9, 21), (−3, 7) is 2 = 63. So, the total area of the
quadrilateral is 165
2 .
4. E is a point in the interior of rectangle ABCD. AB = 6, triangle ABE has area 6, and triangle CDE
has area 12. Find (EA)2 − (EB)2 + (EC)2 − (ED)2 .
Solution: 0
Let l be the line through E perpendicular to AB and CD. Let F and G be the intersections of l with
AB and CD, respectively. Then, AF = DG, call this a, and BF = CG, call this b. Since triangle
ABE has area 6, EF = 2. Since triangle CDE has area 12, EG = 4.
A a F b B

D a b C
G
 By the Pythagorean Theorem, we have the following equalities:

a2 + 4 = (EA)2 , b2 + 4 = (EB)2 , b2 + 16 = (EC)2 , a2 + 16 = (ED)2

Thus, (EA)2 − (EB)2 + (EC)2 − (ED)2 = 0.

5. Two identical cones, each 2 inches in height, are held one directly above another with the pointed end
facing down. The upper cone is completely filled with water. A small hole is punctured in the bottom
of the upper cone so that the water trickles down into the bottom cone. When the water reaches a
depth of 1 inch in the bottom cone, what is its depth in the upper cone?
√
Solution: 3 7
Let r be the radius of the base of one cone. Then, the amount of water is V = 13 2πr2 = 23 πr2 . When
the height of the water in the bottom cone is 1, using similar triangles, the volume of water in the
 2
bottom cone is Vb = 13 1π 2r = 12 1
πr2 . So, the volume of water is the upper cone is Vu = V − Vb =
7 2 1 2
12 πr = 3 hnew πrnew where hnew is the depth of the water in the upper cone and rnew is the radius of
the cone formed by the remaining water. Again, by similar triangles, rnewr = hnew
2 . Substituting,

 2
7 2 1 hnew
Vu = πr = πhnew r .
12 3 2
√
Solving this equation, we find that hnew = 3 7.
6. Find the radius of a circle inscribed in the triangle determined by the lines 4x + 3y = 24, 56x − 33y =
−264, and 3x − 4y = 18.
Solution: 4
First, observe that the lines 4x + 3y = 24 and 3x − 4y = 18 are perpendicular, since their slopes
are negative reciprocals. We can also find their intersection point; 4x + 3y = 24 can be rewritten as
y = 24−4x
3 , while 3x − 4y = 18 can be rewritten as y = 3x−18 4 . Setting these two equal, we have
24−4x 3x−18
3 = 4 . Cross-multiplying, 4(24 − 4x) = 3(3x − 18), so 96 − 16x = 9x − 54, and 150 = 25x,
and x = 6. Then, y = 24−4x3 = 0. Thus, one vertex of the triangle is (6, 0). Similarly, we find that the
other two vertices of the triangle are (0, 8) and (− 66
5 , − 72
5 ).
Now, we must
find the lengths of the legs of the triangle. The leg between the vertices (6, 0) and (0, 8)
has length (6 − 0)√2 + (0 − 8)2 = 10, while the other leg has length 24. Thus, the hypotenuse of the
triangle has length 102 + 242 = 26.
Let the radius of the incircle be x. Using the fact that the tangents from a point to a circle have the
same length, we have the following diagram:

10−x

10−x
x 24−x
x

24−x

Thus, (10 − x) + (24 − x) = 26, so x = 4.

7. In the figure, AB is tangent at A to the circle with center O; point D is interior to the circle; and DB
intersects the circle at C. If BC = DC = 3, OD = 2 and AB = 6, then find the radius of the circle.
 D

O
C

A B
√
Solution: 22
Let θ = ∠ODB and r be the radius of circle O.

D
2 θ 3
O
C
r 3

A 6 B

Observe that ABO is a right triangle, so by the Pythagorean theorem. BO2 = 36 + r2 . By the law of
cosines with triangle ODC, r2 = 4 + 9 − 12 cos θ, so −12 cos θ = r2 − 13. By the law of cosines with
triangle ODB, BO2 √ = 4 + 36 − 24 cos θ = 40 + 2(r2 − 13) = 14 + 2r2 . Therefore, 36 + r2 = 14 + 2r2 ,
2
so r = 22, and r = 22.

8. Let S be the solid tetrahedron with boundary points (0, 0, 0), (2, 4, 0), (5, 1, 0), (3, 2, 10). Let z1 =
max{q | ( 12 23 9 5
5 , 10 , q) ∈ S} and let z2 = max{r | ( 2 , 4 , r) ∈ S}. Find z1 − z2 .
15
Solution: 4
A tetrahedron is made up of four planes. Notice that z1 and z2 must lie on the boundary of S, so each
lies on one of the four planes. However, we can ignore the plane defined by (0, 0, 0), (2, 4, 0), (5, 1, 0)
because z1 and z2 will not be on this plane (as it is lower than all the rest, and we want the maximum).
Also notice that a vertical line intersects each of the other 3 planes exactly once, but only the lowest
intersection point is actually on the tetrahedron.
To find the planes, recall that any plane that doesn’t go through the origin can be represented as
ax + by + cz = 1. Let P1 be the plane defined by (2, 4, 0), (5, 1, 0), (3, 2, 10). Then, 2a + 4b = 1,
5a + b = 1, and 3a + 2b + 10c = 1. Solving, we find that P1 is the plane 16 x + 16 y + 60 1
z = 1.
A plane that does go through the origin can be written as x + by + cz = 0. Let P2 be the plane
defined by (0, 0, 0), (2, 4, 0), (3, 2, 10). Then, using the same method as above, we find that the plane
is x − 12 y − 15 z = 0. Let P3 be the plane defined by (0, 0, 0), (5, 1, 0), (3, 2, 10). Then, it has equation
7
x − 5y + 10 z = 0.
Using these equations, we find that ( 12 23 12 23 25 12 23
5 , 10 , 13) is on P1 , ( 5 , 10 , 4 ) is on P2 , and ( 5 , 10 , 13) is on P3 .
Only the lowest is on the tetrahedron, so z1 = 4 . Similarly, we find that z2 = 2 , so z1 − z2 = 15
25 5
4 .

9. Circles A and B are tangent and have radii 1 and 2, respectively. A tangent to circle A from the
point B intersects circle A at C. D is chosen on circle B so that AC is parallel to BD and the two
segments BC and AD do not intersect. Segment AD intersects circle A at E. The line through B and
E intersects circle A through another point F . Find EF .
 √
Solution: 2 3/3
Draw a line from A perpendicular to BD. Let G the point of intersection of this line with BD. Now,
notice that ACBG is a rectangle.

A B

E
F G

Since C is on circle A, AC = 1. Also, AB =√3 since circle A has radius 1 and circle B has radius√2.
By the pythagorean theorem, BC = AG = 2 2. Then, AGD is a right triangle with legs 1 and 2 2,
so AD = 3. Then, since A, E, and D are collinear, AD = AE + ED, so ED = 2. This means that
ED = BD = 2, so DEB is an isosceles triangle with vertex D. Notice that AEF is also isosceles, since
E and F are both on circle A. Also, ∠AEF = ∠BED, so the triangles AEF and DEB are similar.
In particular, the ratio of a side of DEB to a side of AEF is 2 to 1 because BD = 2 and AF = 1, so
F E = 12 EB, and F E = 13 BF .
So, now we just need to find BF . Let θ = ∠AF E. Then, ∠AEF = θ, so ∠F AE = 180◦ −2θ. Also, since
AEF and BED are similar, ∠BDE = 180◦ − 2θ. Since AGD is a right triangle, ∠GAD = 2θ − 90◦ ,
so ∠GAF = ∠GAD + ∠F AE = (2θ − 90◦ ) + (180◦ − 2θ) = 90◦ . Since CAG is also a right angle, √ this
means that C, A, and F are collinear. So, BCF is a right triangle, and it has legs 2 and 2 2. So,
√ √
BF = 2 3, and F E = 13 BF = 2 3 3 .
10. E is a point inside square ABCD such that ∠ECD = ∠EDC = 15◦ . Find ∠AEB.
Solution: 60◦
We will use symmetry to solve this problem. Place points F, G, and H inside the square so that
triangles EDC, F CB, GBA, and HAD are congruent.

A B

H F

D C

Then triangles EDH, F CE, GBF , and GAH are isosceles with vertex angle 60◦ , so they are equilateral.
So, GHEF is a rhombus. Then, since ∠DEC is 150◦ and ∠DEH and ∠CEF are both 60◦ , ∠HEF
is 90◦ . By symmetry, the other angles of GHEF are also 90◦ , so GHEF is a square.
Since AHG is equilateral and GHEF is a square, AEH is an isosceles triangle with vertex angle 150◦ .
Thus, ∠AEH is 15◦ . Similarly, ∠BEF is 15◦ , so ∠AEB is 60◦ .
 2001 Stanford Math Tournament
General Test
1. If a & b is defined as 2a − ba , what value is associated with 3 & 2?
Solution: −2
3 & 2 = 2 · 3 − 23 = 6 − 8 = −2.
2. Find the coordinates of the points of intersection of the graphs of the equations y = |2x| − 2 and
y = −|2x| + 2.
Solution: (1, 0), (−1, 0)
(x, y) is a point of intersection iff y = |2x| − 2 = −(|2x| − 2), so y must be 0. Then, |2x| = 2, so x = ±1.
3. ABCD is a 4-digit number. What is the largest number that can be formed with AB a prime 2-digit
number and C and D each a prime 1-digit number?
Solution: 9777
The largest 2-digit prime number is 97, and the largest 1-digit prime is 7, so the largest number with
the specified properties is 9777.
4. Let A be the set of all non-composite positive integers. Let B be the set of all squares of integers. Let
C be the set of all multiples of 3. What is the size of A ∩ (B ∪ C)?
Solution: 2
B ∪ C = all squares and multiples of 3, so A ∩ (B ∪ C) = {1, 3}, and the size of this is 2.
5. What is i2959 ?
Solution: −i
Notice that i2 = −1, i3 = −i, and i4 = 1. Since 2959 = 4 × 739 + 3, i2959 = (i4 )739 i3 = i3 = −i.
6. The points Q = (9, 14) and R = (a, b) are symmetric with respect to the point P = (5, 3). What are
the coordinates of point R?
Solution: (1, −8)
P is the midpoint of the segment QR, so a+9
2 = 5 and b+14
2 = 3. Thus, a = 1 and b = −8, so
R = (1, −8).
7. If F (x) = 3x3 − 2x2 + x − 3, find F (1 + i).
Solution: −8 + 3i
Notice that (1 + i)2 = 2i, so (1 + i)3 = 2i − 2. So, 3(1 + i)3 − 2(1 + i)2 + (1 + i) − 3 = 3(2i − 2) −
2(2i) + (1 + i) − 3 = −8 + 3i.
8. Express the absolute value of the difference between 0.36 and 0.36 as a common fraction.
Solution: 1/275
36 4
Let x = 0.36. Then, 100x = 36.36, so 99x = 36, and x = 99 = 11 . So, the difference between 0.36 and
4 9 1
0.36 is 11 − 25 = 275 .
9. Jonathan chooses 10 cards without replacement from a standard 52-card deck of cards (without jokers).
What is the probability that he does not draw the 3 of clubs and he does not draw the King of spades?
287
Solution: 442
 
The number of ways to draw 10 cards that are not either one of those cards is 50 10 , and the total
52 (50) 50!
number of ways to draw 10 cards is 10 . So, the probability is 10 = 10!40! = 42·41 287
52·51 = 442 .
(52
10)
52!
10!42!
10. Julie has a 12 foot by 20 foot garden. She wants to put fencing around it to keep out the neighbor’s
dog. Normal fenceposts cost $2 each while strong ones cost $3 each. If Julie needs one fencepost for
every 2 feet and has $70 to spend on fenceposts, what is the largest number of strong fenceposts she
can buy?
Solution: 6
Let x be the number of normal fenceposts and y be the number of strong fenceposts. The perimeter
of Julie’s garden is 64 feet, and she needs one fencepost every 2 feet, so x + y = 32. The total cost of
the fenceposts is 2x + 3y, and we want 2x + 3y ≤ 70. Since 2x + 2y = 64, we find that y ≤ 6.
11. Anne has a cube that she wants to paint. If she decides to paint each face a different color and she has
6 colors, how many distinct ways can she paint the cube? (If one painted cube can be rotated to get
another, they are the same.)
Solution: 30
Let’s label the colors 1 through 6. Anne wants to paint each face a different color, so she will use each
color exactly once. Since rotated paintings are the same, let’s always rotate our painting so that color
1 is on the bottom. Then, there are 5 colors, and 5 faces left. She can color these in 5! = 120 ways.
However, notice that she can spin the cube around (always keeping color 1 on the bottom), and these
painting are all the same. There are 4 such rotations, so there are actually only 120
4 = 30 ways to paint
the cube.
12. Find, in degrees, the sum of angles 1, 2, 3, 4, and 5 in the star-shaped figure shown.

2
3

5 4

Solution: 180◦
Let’s label the points where the star meets the circle A, B, C, D, and E.
B

2 C
3

A 1

5 4

D
E

The intercepted arc of angle 1 is CD, so the measure of angle 1 is half the arc-length of CD. We can
do the same for the other angles, so the sum of the angles is half the sum of the arc lengths AB, BC,
CD, DE, and EA. These make up the whole circle, so their sum is 360◦ . Therefore, the sum of angles
1, 2, 3, 4, and 5 is 180◦ .
13. Each valve A, B, and C, when open, releases water into a tank at its own constant rate. With all three
valves open, the tank fills in 1 hour, with only valves A and C open it takes 1.5 hours, and with only
valves B and C open it takes 2 hours. How many hours will it take to fill the tank with only valves A
and B open?
6
Solution: 1.2 = 5
Let rA , rB , rC be the rates of valves A, B, and C in tanks per hour, respectively. Then, 1(rA +rB +rC ) =
1, 1.5(rA + rC ) = 1, and 2(rB + rC ) = 1. So, rA + rB + rC = 1, rA + rC = 23 , and rB + rC = 12 . So,
rA + rB = 2(rA + rB + rC ) − (rA + rC ) − (rB + rC ) = 2 − 23 − 12 = 56 . Thus, it takes 65 = 1.2 hours to
fill the tank with only valves A and B open.
14. Lattice paths are paths consisting of one unit steps in the positive horizontal or vertical directions.
How many distinct lattice paths are there from the origin to the point (5, 4)?
Solution: 126
To get to (5, 4) from the origin, you must go to the right 5 times and up  4 times. The four vertical
steps can be taken in any order in the 9 total steps, and thus there are 94 = 126 distinct lattice paths.
15. Three circles, each of area 4π, are all externally tangent. Their centers form a triangle. What is the
area of the triangle?
√
Solution: 4 3
Each√ circle has radius 2, so the triangle is equilateral with side length 4. Thus, the area of the triangle
is 4 3.
16. Two equilateral triangles sharing an edge have a combined area of π. What is the square of the length
of their shared edge?
√
2π 3
Solution: 3
√ √
s2 3 2
3
Let s be the length the shared edge. An equilateral triangle of side length s has area 4 , so 2 s 4 = π,
√
and s2 = 2π3 3 .
17. Frogger wants to cross a stream. He starts on one bank and jumps onto the first passing log, which is
traveling at 9 feet per second. The following logs are at speeds such that the second is twice is fast as
the first, the third is twice as fast as the second, the fourth is one-third as fast as the third, and the
last (fifth) is one-third as fast as the fourth (all of the logs travel in the same direction). If there is a
2 second interval between jumps, then how far does he travel down river from his original point, once
he reaches the opposite bank?
Solution: 158
The first log travels at 9 feet per second, so the second travels at 18 feet per second, the third travels
at 36 feet per second, the fourth travels at 12 feet per second, and the last travels at 4 feet per second.
Frogger spends 2 seconds on each log, so he travels 2(9 + 18 + 36 + 12 + 4) = 158 feet.
18. Rice tuition was $7200 one semester. Each student takes a minimum of 12 credits and a maximum of
20 credits. For each credit, a class meets for 3 hours a week. A semester is 15 weeks with no breaks.
What is the difference in cost in dollars per hour of class that a student taking the maximum load and
a student taking the minimum load pays, to the nearest penny?
Solution: $5.33
A student taking the minimum load has 12 · 3 · 15 = 540 class hours over the semester so pays
7200 40
540 = 3 = $13.33 per class hour. A student taking the maximum load has 20 · 3 · 15 = 900 class
hours so pays 7200
900 = $8 per class hour. So, the difference is $13.33 − $8 = $5.33.

19. Suppose a1 , a2 , a3 , . . . is a sequence of numbers such that a1 = 1 and an+1 = an + (2n + 1) for all
positive integers n. Find a20 .
Solution: 400
In general, an = n2 . This is certainly true for a1 , and if an = n2 , then an+1 = an + (2n + 1) =
n2 + 2n + 1 = (n + 1)2 , so it is true for all n (this sort of proof is called induction).
20. My spouse and I have 9 kids. Each child gets married and has exactly 9 children of their own. By the
time I have 100,000 descendants (including every generation) what is the longest title that applies to
me? (Assuming no married blood relatives.)

a. grandparent
b. great grandparent
c. great great grandparent
 d. great great great grandparent
e. great great great great grandparent
f. great great great great great grandparent
g. great great great great great great grandparent

Solution: great great great great grandparent

I have 9 children, 92 grandchildren, 93 great grandchildren, etc. So, the number of descendants I have
n+1
in the n-th generation is 9 + 92 + . . . + 9n = 9 8 −1 − 1. The smallest value of n for which this is at
least 100,000 is 6, which makes me a great great great great grandparent.
21. Find the greatest integer x for which 320 > 32x .
Solution: 6
First, notice that 32x = 25x , so 320 > 32x if and only if 34 > 2x . 34 = 81, and it is easy to check that
26 is the largest power of 2 less than 81.
22. If x + y = xy, with x and y real, what value can x not have?
Solution: 1
x + y = xy if and only if xy − x − y + 1 = 1, or (x − 1)(y − 1) = 1. This is impossible if x = 1.
23. Instead of using two standard cubical dice in a board game, three standard cubical dice are used so
that the game goes more quickly. In the regular game, doubles are needed to get out of the “pit”. In
the revised game, doubles or triples will get you out. How many times as likely is it for a player to get
out of the “pit” on one toss under the new rules as compared to the old rules?
Solution: 8/3
The probability of rolling two different numbers with two different dice is 6·5 5
6·6 = 6 , so the probability
5 1
of rolling doubles is 1 − 6 = 6 . The probability of rolling three different numbers with three different
dice is 6·5·4 5 5 4
6·6·6 = 9 , so the probability of rolling doubles or triples with three dice is 1 − 9 = 9 . So, the
4 1 8
ratio of the two is 9 / 6 = 3 .
24. A circle is drawn with center at the origin and radius 2.5. Find the coordinates of all intersections
of the circle with an origin-centered square of side length 4 whose sides are parallel to the coordinate
axes.
Solution: (±1.5, ±2), (±2, ±1.5)
The square consists of all points (x, y). with −2 ≤ x, y ≤ 2 and at least one of x and y equal to 2 or
-2. The points on the circle are all points (x, y) with x2 + y 2 = 6.25. So, the intersection of these two
is (±2, ±1.5) and (±1.5, ±2).
(7017 )!
25. Write 2010 (170 8 )!
as a decimal in base 6. The subscript indicates the base in which the number is
written (i.e., 156 is 15 base 6, so 156 = 1110 )
Solution: 0.16 or 0.1
The easiest way to solve this problem is to convert everything into base 10.

7017 = 7 × 171 + 0 × 170 = 119

1708 = 1 × 82 + 7 × 81 + 0 × 80 = 120
(7017 )! 119! 1
2010 = 20 =
(1708 )! 120! 6

1
Converting to base 6, 6 = 0.16 .
26. Jacques is building an igloo for his dog. The igloo’s inside and outside are both perfectly hemispherical.
254
The interior height at the center is 2 feet. The igloo has no door yet and contains 2187 π cubic yards
of hand-packed snow. What is the circumference of the igloo at its base in feet?
14
Solution: 3
π
Let R be the radius of the igloo at its base, in feet. Then, the amount of snow in the igloo is, in cubic
feet, 12 43 πR3 − 43 π · 23 = 23 π(R3 − 8). This is equal to 2187
254
π cubic yards, or 254 3 254
81 = 3 · 2187 cubic
2 3 254 3 343 7
feet. Therefore, 3 π(R − 8) = 81 π, so R = 27 , and R = 3 . So, the circumference of the igloo is
2πR = 14 3 π feet.

27. How many solutions are there to x7 + y 7 = z 7 with x, y, z real?

Solution: ∞
For any real x, y, x7 + y 7 is real, so x7 + y 7 has a 7-th root. If we let z be this root, then (x, y, z) is a
solution to x7 + y 7 = z 7 . So, for any x, y, there is a solution. There are infinitely many x, y, so there
must be infinitely many solutions.
28. Find all prime factors of 318 − 218 .
Solution: 5, 7, 19, 577, 1009

318 − 218 = (39 − 29 )(39 + 29 ) = (33 − 23 )(36 + 33 · 23 + 26 )(33 + 23 )(36 − 33 · 23 + 26 )

= 19 · 1009 · 35 · 577 = 5 · 7 · 19 · 577 · 1009
√
29. If a = 1 and a
10000a = 10a , find a. The subscript indicates the base in which the number is written.
Solution: 4
√
10000a = a4 and 10a = a, so a4 = a. Thus, a4/a = a, and 4/a = 1, so a = 4.
a

30. Suppose 100 students have at least one of math, applied math, or statistics as one of their majors.
There are 24 statistics majors, 46 applied math majors, and 55 pure math majors. There are 14 who
are at least pure & applied math, 10 at least applied math & statistics and 7 at least statistics & pure
math. How many triple majors (people with all three majors) are there, if any?
Solution: 6
Using the principle of inclusion-exclusion, the number of triple majors is given by 100 − (46 + 24 +
55) + (14 + 10 + 7) = 6.
31. The sum of 3 real numbers is 0. If the sum of their cubes is π e , what is their product?
πe
Solution: 3
Let x, y, and z be the 3 real numbers. Then x + y + z = 0, so z = −(x + y). Also,

πe = x3 + y 3 + z 3 = x3 + y 3 − (x + y)3
= −3x2 y − 3xy 2 = −3xy(x + y) = 3xyz

So, xyz = 13 π e .

x+3
32. Given a real-valued function f (x) = |2x+6| , what is the least integer value which lies in the domain
of the function?
Solution: −2
We have x ≥ −3, as the value under the radical must be nonnegative, and the absolute value is non
negative, so x + 3 ≥ 0. But, the denominator must be non-zero, so we must exclude x = −3, as then
2x + 6 = 0. Hence, the least integer value is -2.
33. E is a point in the interior of rectangle ABCD. AB = 6, triangle ABE has area 6, and triangle CDE
has area 12. Find (EA)2 − (EB)2 + (EC)2 − (ED)2 .
Solution: 0
Let l be the line through E perpendicular to AB and CD. Let F and G be the intersections of l with
AB and CD, respectively. Then, AF = DG, call this a, and BF = CG, call this b. Since triangle
ABE has area 6, EF = 2. Since triangle CDE has area 12, EG = 4.
A a F b B

D a b C
G

By the Pythagorean Theorem, we have the following equalities:

a2 + 4 = (EA)2 , b2 + 4 = (EB)2 , b2 + 16 = (EC)2 , a2 + 16 = (ED)2

Thus, (EA)2 − (EB)2 + (EC)2 − (ED)2 = 0.

34. What is the radius of the incircle of a right triangle with legs of lengths 7 and 24?
Solution: 3
Let ABC be a right triangle with AB = 24 and BC = 7. Then, AC = 25. Let x be the radius of the
incircle of ABC. Let D, E, and F be the points on AB, BC, and AC, respectively, that are tangent
to the incircle. Then, AD = AF = 24 − x and CE = CF = 7 − x. So, AC = 31 − 2x = 25, and x = 3.
35. If logA B + logB A = 3 and A < B, find logB A.
√
3− 5
Solution: 2
1
Let x = logB A. Then B x = A, so B = A x . Thus, logA B = x1 . So, x + 1
x = 3. Multiplying through
by x, x2 + 1 = 3x, so x2 − 3x + 1 = 0. By the quadratic formula,
√
3± 5
x=
2
Since A < B, logB A < 1, so √
3− 5
x=
2
.
36. ABC is an equilateral triangle with edge of length 256. Let n be the maximum number of non-
overlapping equilateral triangles of edge length 12 that can be fit into ABC. What is the prime
factorization of n?
Solution: 218
For a triangle of side length n, we can fit in 4 triangles with sides of length n2 (maybe draw a figure).
256 1 9 18
29 = 2 , so we can do this 9 times. Each time, the number of triangle increases 4-fold. 4 = 2 .

37. If x2 + x12 = 7, find x3 + x13 .
Solution: 18

Observe that (x + x1 )2 = x2 + 2 + x12 = 9. so x + x1 = 3. Cubing, x3 + 3x + x3 + x13 = 27. Now,
x3 + x13 has the same
3 sign as 3x + x3 because
if x is positive,
both
are positive,
and if x is
negative,
both
are negative. So, x + 3x + x + x13 = x3 + x13 + 3 x + x1 = 27. Since x + x1 = 3, x3 + x13 = 18.
3
38. Eleven pirates find a treasure chest. When they split up the coins in it evenly among all the pirates,
they find that there are 5 coins left. They throw one pirate overboard and split the coins again, only
to find that there are 3 coins left over. So, they throw another pirate over and try again. This time,
the coins split perfectly. What is the least number of coins there could have been?
Solution: 423
Let n be the number of coins in the chest. Then, n ≡ 5 (mod 11) (this means that the remainder of
n when divided by 11 is 5), n ≡ 3 (mod 10), and n ≡ 0 (mod 9). The Chinese Remainder Theorem
gives an elegant solution to this problem, or you can solve it as follows. Since n ≡ 3 (mod 10),
we know that the last digit of n must be 3. Also, since n ≡ 0 (mod 9), n is divisible by 9, so
n = 9q for some q. Since n ends in a 3, q must end in a 7. That means that q looks like 10p + 7, so
n = 9(10p + 7) = 90p + 63. Now, you can just try p = 1, 2, . . ., and you’ll find that p = 4 works.
 2001 Stanford Math Tournament
Team Test
1. ABCD is a square with sides of unit length. Points E and F are taken on sides AB and AD respectively
so that AE = AF and the quadrilateral CDF E has maximum area. What is this maximum area?
Solution: 5/8
2 2
Let x = AE. Then, BE = 1 − x, so the area of CDF E is 1 − x2 − 1−x 2 = 1 − x −x+1
2 , which is
2 2 1
maximized when x − x + 1 is minimized. x − x + 1 is minimized at x = 2 . Plugging this into our
formula for the area of CDF E, we find that the maximum area of CDF E is 58 .
2. How many positive integers between 1 and 400 (inclusive) have exactly 15 positive integer factors?
Solution: 3
In general, if a number has prime factorization pn1 1 pn2 2 · · · pnk k , then its positive integer factors are
exactly those of the form pj11 pj22 . . . pjkk where 0 ≤ ji ≤ ni for each i between 1 and k, inclusive. In
particular, there are ni + 1 choices for ji , so there are (n1 + 1)(n2 + 1) · · · (nk + 1) positive integer
factors. In our case, since 15 = 1 · 15 = 3 · 5 are the only factorizations, either n1 = 0 and n2 = 14 or
n1 = 2 and n2 = 4. So, a positive integer has exactly 15 positive integers if and only if it is of the form
p14 for some prime p or of the form p21 p42 for two distinct primes p1 and p2 . 214 > 400, so there are
no numbers of the first form between 1 and 400. The only ones of the second form that are between
1 and 400 are 32 · 24 = 144, 52 · 24 = 400, and 22 · 34 = 324. Therefore, there are 3 positive integers
between 1 and 400, inclusive, that have exactly 15 positive integer factors.
3. Find the 2000th positive integer that is not the difference between any two integer squares.
Solution: 7998
First, we will try to find all numbers that can be expressed as a difference between two integer squares.
For any integer n, (n+ 1)2 − n2 = 2n+ 1, so any odd number can be expressed as the difference between
two integer squares. Similarly, (n + 2)2 − n2 = 4n + 4 = 4(n + 1), so any number divisible by 4 can be
expressed as the difference between two integer squares.
What about the other numbers, the ones of the form 4n + 2? It is easy to see that none of these can
be the difference between two integer squares. Assume that it could; then x2 − y 2 = 4n + 2 for some
x and y. Then, either x and y are both even, or they are both odd, since their difference is even.
However, then x2 − y 2 would be divisible by 4, while 4n + 2 is not. So, the integers that cannot be
expressed as the difference between two integer squares are exactly those of the form 4n + 2. The first
is 4(0) + 2 = 2, so the 2000th one is 4(1999) + 2 = 7998.
4. For what values of a does the system of equations

x2 = y 2 , (x − a)2 + y 2 = 1

have exactly 2 solutions?

√
Solution: ± 2
We can plot this system to get a rough idea of what it looks like. x2 = y 2 is the same as x = ±y, or
a pair of perpendicular lines. (x − a)2 + y 2 = 1 is a circle with center (a, 0) and radius 1. We can see
that the system will have exactly 2 solutions if and only if the circle is tangent to both lines:

O A
 Let’s label the origin O, the center of the circle A = (a, 0), and the two points of tangency B and
C. Then, OBAC is a rectangle, √ and AB =√AC = 1. Thus, OBAC must be a square, and AO, its
diagonal, must have length 2. Thus, a = 2. Now, √ of course, the picture is exactly the same if the
circle is on the other side of the√origin, so a = − 2 is also a possible solution. Thus, the values of a
giving exactly 2 solutions are ± 2.
5. What quadratic polynomial whose coefficient of x2 is 1 has roots which are the complex conjugates of
the solutions of x2 − 6x + 11 = 2xi − 10i? (Note that the complex conjugate of a + bi is a − bi, where
a and b are real numbers.)
Solution: x2 + (−6 + 2i)x + (11 − 10i)
Let a, b be the solutions of x2 −6x+11 = 2xi−10i. Then, a and b are the roots of x2 −(6+2i)x+(11+10i),
so x2 − (6 + 2i)x + (11 + 10i) = (x − a)(x − b) = x2 − (a + b)x + ab. Therefore, a + b = 6 + 2i and
ab = 11 + 10i.
Let a, b the complex conjugates of a, b, respectively. Then, the quadratic polynomial with one as the
coefficient of x2 and a, b as roots is (x − a)(x − b) = x2 − (a + b)x + ab = x2 − (a + b)x + ab =
x2 − (6 − 2i)x + (11 − 10i) = x2 + (−6 + 2i)x + (11 − 10i) = 0.
6. Find the least n such that any subset of {1, 2, . . . , 100} with n elements has 2 elements with a difference
of 9.
Solution: 55
Let A = {18a + 1, 18a + 2, . . . , 18a + 9 : a = 0, 1, 2, 3, 4, 5}. Then, notice that A has 54 elements, and
no 2 elements have a difference of 9. So, we know n ≥ 55. Now, we just need to convince ourselves
that any subset with 55 elements must have 2 elements with a difference of 9. To do this, take the
sets {x, x + 9} for x = 1, 2, . . . , 91. Notice that there are 91 such sets. The numbers 1, 2, . . . , 9 and
92, . . . , 100 each fall into exactly one of the set, while all other numbers fall into two of them. So, there
are 18 numbers that fall into exactly one set, and then there are 73 other sets. That means that, in a
set of 55 elements, at least 37 fall into two sets. As 37 · 2 + 18 = 92 > 91 (the number of sets), by the
pigeonhole principle, two of them fall in the same set, and thus have a difference of 9.
7. The median to a 10 cm side of a triangle has length 9 cm and is perpendicular to a second median of
the triangle. Find the exact value in centimeters of the length of the third median.
√
Solution: 3 13
Call the triangle ABC, where AB = 10, BC = 2x, and CA = 2y. Let D, E, and F be the midpoints of
AB, BC, and CA, respectively. Let G be the centroid (the point where the three medians intersect).

C
x
y
E
F
G x
y
B
5
5 D
A
CG AG
We are given that CD = 9. One property of the centroid is that GD = GE = BG
GF = 2. So, CG = 6
and GD = 3. We are given that CD is perpendicular to another median, say AE. Let a = GE; then
2 2 2 2 2
AG = 2a. By the Pythagorean Theorem, √ a + 36 = √x , (2a) + 36 = (2y) , and 9 + (2a) = 25. The
last tells us that a = 2, so then x = 2 10 and y = 13.
Let θ = ∠BCA. By the law of cosines with triangle ABC, 100 = (2y)2 + (2x)2 − 2(2x)(2y) cos θ,
so xy cos θ = 100−52−160 = 14. By the law of cosines with the triangle BCF , BF 2 = y 2 + (2x2 ) −
−8 √
2y(2x) cos θ = 13 + 160 − 4 · 14 = 117. So, BF = 3 13.
 8. Janet and Donald agree to meet for lunch between 11:30 and 12:30. They each arrive at a random
time in that interval. If Janet has to wait more than 15 minutes for Donald, she gets bored and leaves.
Donald is busier so will only wait 5 minutes for Janet. What is the probability that the two will eat
together? Express your answer as a fraction.
43
Solution: 144
We will solve this problem using a graph. Let the x-axis be time for Janet, and let the y-axis be time
for Donald. Then, the region we are interested in is the square with both x and y between 11:30 and
12:30. Since Janet will wait no more than 15 minutes for Donald, we have x ≤ y + 15. Since Donald
will wait no more than 5 minutes for Janet, y ≤ x + 5.

12:30

Donald
11:30 12:30
Janet

452
The area of the whole square is 602 , the area of the upper triangle is 2 , and the area of the lower
2 2
2
55 602 − 452 − 552 43
triangle is 2 , so the probability that the two will eat together is 602 = 144 .
9. What is the minimum number of straight cuts needed to cut a cake in 100 pieces? The pieces do not
need to be the same size or shape but cannot be rearranged between cuts. You may assume that the
cake is a large cube and may be cut from any direction.
Solution: 9
First, we will find the number of regions that n lines split a plane into. Notice that 0 lines split the
plane into 1 region, 1 line splits the plane into 2 regions, 2 lines splits it into 4 regions, etc. In general,
the k-th line can intersect each of the first k − 1 lines exactly once, creating k more regions. So, the
number of regions that n lines split a plane into is


n
n(n + 1)
1+ k= +1
2
k=1

Now, cutting the cake is the same idea, only in 3 dimensions instead of a plane. Again, 0 planes split
the cake into 1 piece, 1 plane gives 2 pieces, 2 planes give 4 pieces, etc. Notice that the k-th plane can
intersect each of the first k − 1 planes exactly once, but now the intersection is a line. Then, by the
planar case, these lines divide the new plane into (k−1)k
2 + 1 regions; each of these planar regions splits
one piece of the cake into two, so we’ve added (k−1)k
2 + 1 pieces. So, the number of regions that n lines
split a cube into is
n 


1
2 1


n n n
(k − 1)k
1+ +1 = 1+ k − k+ 1
2 2 2
k=1 k=1 k=1 k=1
1 n(n + 1)(2n + 1) 1 n(n + 1)
= 1+ − +n
2 6 2 2
2n3 + 3n2 + n n2 + n
= n+1+ −
12 4
2n3 + 10n + 12 n3 + 5n + 6
= =
12 6

n3 +5n+6
Then, the least n so that 6 ≥ 100 is 9.
10. You know that the binary function  takes in two non-negative integers and has the following properties:
 • 0a=1
• aa=0
• If a < b, then a  b = (b − a)[(a − 1)  (b − 1)].
Find a general formula for x  y, assuming that y ≥ x > 0.
Solution: (y − x)x
We just need to check that this satisfies the three given properties. First, 0  a = (a − 0)0 = 1. Second,
a a = (a− a)a = 0. Finally, if a < b, then a b = (b − a)a = (b − a)[(b − a)a−1 ] = (b − a)[(a− 1) (b − 1)].
11. Christopher and Robin are playing a game in which they take turns tossing a circular token of diameter
1 inch onto an infinite checkerboard whose squares have sides of 2 inches. If the token lands entirely
in a square, the player who tossed the token gets 1 point; otherwise, the other player gets 1 point. A
player wins as soon as he gets two more points than the other player. If Christopher tosses first, what
is the probability that he will win? Express your answer as a fraction.
1
Solution: 2
First, notice that the player tossing wins a point if the center of the coin lands in the center 1 × 1 by
square of each 2 × 2 square of the checkerboard. So, the player tossing wins a point with probability
1
4 . Now, a player wins when the other player has n points and he has n + 2, making a total of 2n + 2
tosses. We can think about the tosses in pairs; the first two are a pair, the second two are a pair,
etc. The first player to win one of these pairs also wins the game, so the probability that Christopher
wins the game in 2n + 2 turns is the probability that, in the first n pairs, Christopher gets 1 point and
Robin gets 1 point, and in the last pair, Christopher wins both points.
Christopher tosses first in each pair, so his probability of winning the first toss is 14 , and his probability
of winning the second toss by Robin not winning is 34 . So, the probability that he wins the first toss
and loses the second toss is 14 · 14 = 16
1
. The probability that he loses the first toss and wins the second
is 4 · 4 = 16 . So, the probability that Christopher and Robin each get 1 point in the pair is 10
3 3 9
16 . The
 n
probability that this happens for the first n pairs is 10 . The probability that Christopher wins the
1 3
16  10 n 3
last pair is 4 · 4 . So, the probability that Christopher wins in 2n + 2 turns is 16 16 . This means
that, overall, the probability that Christopher wins is
∞  n   2  3 

10 3 3 10 10 10
= 1+ + + + ...
n=0
16 16 16 16 16 16

This is a geometric series with sum

3 1 1
· 10 =
16 1 − 16 2

(Notice that the fact that the person tossing wins with probability 14 is not important – no matter
what that probability is, the overall probability of winning in such a game is still 12 !!)
12. A binary string is a string consisting of only 0’s and 1’s (for instance, 001010, 101, etc.). What is
the probability that a randomly chosen binary string of length 10 has 2 consecutive 0’s? Express your
answer as a fraction.
55
Solution: 64
We will first find the number of binary strings of length 10 that don’t have 2 consecutive zeros. Let Cn be
the set of binary strings of length n that don’t have 2 consecutive zeros. Let An be the set of all strings
in Cn that end in 1, and let Bn be the set of all strings in Cn that end in 0. Notice that Cn = An ∪ Bn .
Now, if a string is in An+1 , then, the first n letters must be a string in Cn ; conversely, appending a 1 to
a string in Cn gives a string in An+1 . So, |An+1 | = |Cn | = |An | + |Bn |. Similarly, a string is in Bn+1 iff
the first n letters are in An , so |Bn+1 | = |An |. Thus, |An+2 | = |An+1 | + |Bn+1 | = |An+1 | + |An |, which
gives us a Fibonacci sequence which starts with |A1 | = 1, |A2 | = 2. It’s easy to find that |A10 | = 89 and
|B10 | = 55, so |C10 | = 89 + 55 = 144. Therefore, there are 210 − 144 = 1024 − 144 = 880 strings without
2 consecutive 0’s. So, the probability that a randomly chosen string of length 10 has 2 consecutive 0’s
880
is 1024 = 55
64 .
13. You have 2 six-sided dice. One is a normal fair die, while the other has 2 ones, 2 threes, and 2 fives.
You pick a die and roll it. Because of some secret magnetic attraction of the unfair die, you have a
75% chance of picking the unfair die and a 25% chance of picking the fair die. If you roll a three, what
is the probability that you chose the fair die?
1
Solution: 7
The probability of picking the fair die is 14 , and the probability of rolling a 3 with it is 16 . So, the
1
probability of picking the fair die and rolling a 3 with it is 24 . The probability of picking the unfair
3 1 1
die and rolling a 3 with it is 4 · 3 = 4 . So, the probability that we chose the fair die is
1
24 1
1 1 =
24 + 4
7

14. Find the prime factorization of 1≤i<j≤100 ij.
2
Solution: 3 · 5 · 11 · 101 · 151
 n(n+1)
We will find 1≤i<j≤n ij for any n. Since (1 + 2 + . . . + n) = 2 , we square and get that

2  2
(1 + 2 + . . . + n)2 = n(n+1)
2 . Since 2 1≤i<j≤n ij = (1 + 2 + . . . + n) − (12 + 22 + . . . + n2 ), we
have

 
1 n2 (n + 1)2 n(n + 1)(2n + 1)
ij = −
2 4 6
1≤i<j≤n

3n4 + 6n3 + 3n2 − 2(2n3 + 3n2 + n)

=
24
3n4 + 2n3 − 3n2 − 2n n(3n3 + 2n2 − 3n − 2)
= =
24 24
n(3n + 2)(n2 − 1) (n − 1)n(n + 1)(3n + 2)
= =
24 24

99·100·101·302
So, for n = 100, we find that the sum is 24 = 33 · 25 · 101 · 151 = 3 · 52 · 11 · 101 · 151.
15. Let ABC be an isosceles triangle with ∠ABC = ∠ACB = 80◦ . Let D be a point on AB such that
∠DCB = 60◦ and E be a point on AC such that ∠ABE = 30◦ . Find ∠CDE in degrees.
Solution: 30◦
Let θ = ∠CDE.
A

20

E θ 40

50
110 110
70

20
30
60 50

C B

Since we’re only worried about angles in this problem, we can just pretend that AD = 1. Then, since
ADC is an isosceles triangle with side 1 and base angle 20◦ , AC = 2 cos 20◦ . Since ABC is an isosceles
triangle with vertex angle 20◦ and side 2 cos 20◦ , BC = 4 cos 20◦ sin 10◦ . Notice that ECB is isosceles,
so EC = BC = 4 cos 20◦ sin 10◦ . Also, ADC is isosceles, so DC = AD = 1.
Now, let’s look at the triangle EDC. Notice that ∠DEC = 110 − θ. By the law of sines,
sin θ sin(160◦ − θ)
=
EC DC
so
sin θ
= sin(160◦ − θ)
4 cos 20◦ sin 10◦
= sin(20◦ + θ)
= sin 20◦ cos θ + cos 20◦ sin θ

Then,    
◦ 1 1 − 4 cos2 20◦ sin 10◦
cos θ sin 20 = sin θ − cos 20◦ = sin θ
4 cos 20◦ sin 10◦ 4 cos 20◦ sin 10◦
So,

1 1 − 4 cos2 20◦ sin 10◦

=
tan θ 4 cos 20◦ sin 10◦ sin 20◦
1 − 4 cos2 20◦ sin 10◦
= since sin(2φ) = 2 sin φ cos φ
2 sin 40◦ sin 10◦
cos 10◦ − 4 cos2 20◦ sin 10◦ cos 10◦
=
2 sin 40◦ sin 10◦ cos 10◦
cos 10◦ − 2 cos2 20◦ sin 20◦
=
sin 40◦ sin 20◦
sin 80◦ − cos 20◦ sin 40◦ 2 sin 40◦ cos 40◦ − cos 20◦ sin 40◦
= ◦ ◦
=
sin 40 sin 20 sin 40◦ sin 20◦
◦ ◦
2 cos 40 − cos 20 2 cos(30 + 10 ) − cos(30◦ − 10◦ )
◦ ◦
= ◦
=
sin 20 sin(30◦ − 10◦ )
2(cos 30◦ cos 10◦ − sin 30◦ sin 10◦ ) − (cos 30◦ cos 10◦ + sin 30◦ sin 10◦ )
=
sin 30◦ cos 10◦ − cos 30◦ sin 10◦
√
3 ◦ 3
2 cos 10 −√2 sin 10
◦ √
= = 3
1 ◦ 3 ◦
2 cos 10 − 2 sin 10

√
3
So, tan θ = 3 , which means that θ = 30◦ .

 



 
  

   

  
   






  
 

 


 

 



    

              

    
!    
   " 
    #  
    
Æ  !       $       %



&%

&!     

 ' 

  
  








 "   
  
 #  (
) 
 

 ¾
       ) 
!     #  #   

 ) * +  !  
  )    )
 ,"! 
   "!    #  +
! "  ) 
! 
 
) 
¾ 
 -  ..       .       
   % !  ! 
&        %  "  
&         
" 
       /

  
  0                     + "

 .
  * "  !  )
  *   

 
"     .!     
1!  )
  *     ) 
  * )
 !   )
  
(   
 ) %  &
 +            
    

  
  1 " #   #"  


   ) * " ¾    %
    
    
 #  &!     )     
 +

    !  2        "!     $    
   3 


  *  
     
   
  
 "  !
!  ! "  
 ! 


 )




   )
(

 )


              
  4  

%

&%


&%

& )

% 
 &
  %

  &




, ! !
!      
 

(
  ) *  

 
)
   !   
  !     
  

 ) *!     

  +
!     
    "             
  
 , 
 
! 
*533  "! "  
   
 " 

**
 #   #      
/

 

   0            1 %
*  & )
**   *   1! 
    " 

**   #   !  
**  " * !  #" 
 
"    1!     !     * !  
*       

6      *  !          %*  &!   
* 
6
 )
   
7 +
 

)

  ½ 
Ü·




 ¿
¾
          #   



 )



%


& 

  ½







Ü·





)

%


& 



)
 





 $


) 
  !   


) * 
!
) *    
!   "
   

) ¾¿ 

6 0         
   
 )
    0   "   %&

 
  0
)  ) * 1        )  "  #  # !  $   %*& )

¾

  %*& )  8  %*& )
   
  


  
)
  
 9 %
&

: )
  1  %
& )



;   .     # !  $      
 
   
 )
   
   "  %*& ) !   "  9 %
&

: )
 !   %
& )
      
  
 
   
 )
   !    %& )    )  
3 <"  
 
 
 !  )
 
 
 !
)

 

 

 !  )


 ! 
 )
 
 
  8 ) 
) 7  )
*!    "  /

 
  =
2  %
 
 
 & !        )   

 !  
  )  
   
¿

*             
!  
  

      
  + !         )    !        1 #
 
   "      )%  &  %
 &  + !  
   )  )
  

   !      )
  %>    )  )

  $   ! 

"      &
 
¿ 
1    "   " 
 ) 7     
    

 #   
%   
 &  % 
 
 &  ) 7   
  "  )
)
 $  # ! "   $      



  
 

  
  

   
        


x
5

1
1
1 1

  


   
 
   
  !   "   
  
             

x
5 c
a b

1
1
1 1

 # $
  $ $     
  $    $   % 

 
    &                  


 '(     $              )    %  $

## $
           %  $ %          $ 

   


  
 
 %                      %






* + ,
      
$   +   - 
      $

$        .      -


 


   
 $      
$ $    $ 
 $ $
 $  
    $ 



/ 0%   $$    $ %  

  #          
 

 






       $    $$  $ $    
    ,
           $   * 
  # 
   $  * $$
 $ $  $  12   $$    

 .$ 




        *   
 
$         %





3 + $ #   #   $  #     %
 
   .  $
#    $$  $       
 3

 

 & 

4 
 3 5 
 -     #  #     -

3 6 3  ! / 3       3   #
 +      
#%   %   
 $  # % 


 + -   ,
                        7$ 
         
$  - $  $   $$      

 

 '       $ #  $    8     - 
$   $$       
 $#  $   $$  $ $  

   $  
  $     $ 

#    $$   %
 4         
 .$  -   
#       $  

          


! )   

1        &         ,
   #
  
  92#  $ :

 



 5  $ $
   )      %$   $   #  
   7$ %$  
         #     
  
     ;             $   #     $ #
 ## $
        9     #  :          

        
 
 $  

 

  
 
         

$  
 $  
     #   



 
$   

 
       ,
   # 

 
 
    


" 5
 
 
 #         %      * /  3  
%
    #   
   %
     &        
 # 

 


 
 <
#       #   $  
#   %
      
%
   #$
 #  # 
 -    
  
      
   $
;           
 #  

  
9* 4 / 4 3 :  

  
  


= +               

 >            

    

     >    
$   

  
&   
   # $    #    

 




 +##                %  


x x



4 

 
 <   $  

   $    #    #
 %             
     # $  ,




$$    

  $        $   

6       $ $$   $   

  ,
  $ ,
  $
  

  $$  
##  # $
      
 ,
 
   
           






 )   
 $      #$
   ,
$ ?%
   
,
   $%    -   %
 @          
   $      #             
   ,
             .$  
Æ$  $ $
 
     
          
         



 + $   ,
   
     




*

 
  

     
  





  



  


 









 
  

 

 
 
 



      
 

         !" # #$ !   % ##&  ' " !  
( ! #!
  )  " !!  !   ((  
  ##      !
( ¾  "
 " *   *!  *    +    " ##   ! #"   #   
 " " !# #     * , # ! "     "  # # 

 

 
 -    "  *  " , # !  * ." !#   !  "!  !



 
/ ." !  ! *
(   ! #!    ! " ! !#  ! " , # *  ."!
((

(    / $"#" $ # ! ! " 0# 0  
 1 !" #  !    *
# 2! # 3  ! * " !   $"  ! " !
* "  # !4




¾ 
 1!  *# * " ! "  
 * "  !    
    !

    5 $" " #"   ! ! "
 

  1 $   
 
!







  
 
 



 ¾
 6!   !!#  !  7 $"  # ! 
( (
 
8 
 (        "
! 3* $   
  " !#  $      "    "  $" ! "
   * 
   "   9( :
" $"  74




 ."    * 
 #  "#   !   ! " ! * "  ! * " $
 ! *   $  "  ! # *     " $ ! ! * "  7   
"  * "    !# "  ! # *    ! 
  "   * "  
!   "   * "   !   "    !
( ! "    * 
 !

 

/ ! 

 
 

 $"      "      
/  ( * 
 )  !!   

!
 

      

 ! $ ##  

  
  

   

  
  

  ! " ! 

" 0 $       ! $ 
   
/
 
   
/

  (
."! #   7  * 
* " # Æ#  * #"   ! 7  !  /  " 
!!  ! * "    ." #    & !
 
   

"  !!
!$ !


 





8  

     


 ;$ "  ##  ."       ! "    $ 
   
 


   
$"#" ! #!  *  <!" * "  ##  $  ! $" !
  " " 
."    ! "  #    

= >" ! "   *
 !!
 (4

   
 
Ü
 ! $ $
 
  
    ;?   "!  !!  " #"  $

    
  


  

    

   

  
 
 
 
 ! 

  

 1 -@'!@!  $ ! "

 ¾ ¿ ¾ ¿
  
 
  
  

 




 


 !





A  "!   "  " !" *  #   $ %  $" ! ." # "! " "
(
*   ! *  * ' ! "   * "    " ! !  !#" "
*  ." ! !    !"  !$
" !  # ! !   #!  * / 
 $"  "7 #!!<! # *  ##    # #!!<! # $"#" *$! " " *
     !   %     " " * 8 * $" $  1!! " ! !&! 0#& 
$"  ! " $   " !$  ! $"  " " " !  *4





 
 ;  "  * !  " "  
! " "  
 " ! * "
##  "  * " !  
 ."  
  
 
 


 !# 
  
 !  !
."  * ! "    
      

 * *#!  $ " 
  



!# $   


 (/
B$      "  * $   "        " " " * " $  C!
 *          


&    $ $  ! * 
 >"  
 
$ "    
 * " ! $  !
 ! ! ."! 
   
 
  
  

  ¿  
 ." " !  " !  "  %! !    
 $"#"  ! !

/ 


 




 /  AA


 


   


( >" !
 

 
 

    4

 

 -   
 

 
 

    ." 

   
 

  
  

    





   
 
B$  # * " ! * $  !  ! " *











 
 


  
."! * $&! *

! $   !  $    
       ' #
     !       


 
     
      
  

   
    
     


 
   
   
 
  
 
 
 
    
      

  
  

  


  

  

   

 !




  

      


  
   "



 


 
    
 


     

#
$% 
   
&
#
$%



'  &
#
$%

  &

#
$% &
#
(  ÊÖ &




# )
*
 
   
     
 
    
 

 '

  


  

 
   
  
+ 
, 
 
  % % 
 #  # -


    
  
     
  
   %    - 
%    - 

 


 

  


 
.  
 
  
 

 
%       



 /

 


 
 
 
 
 
'

 

   

 
 



(  

 



      
 0    

   
0 ' '
     

 
 /
    

 

 1
 


 
 
   

  

 

, 
     
    
 
  '      
  


  




 '
*        
 

  ' 


,   

    



 '
 
    
*    

  

        
    '
  


        

,   
      




¿ &     



- . #%%
2 


  
#0       3
  




 
  
#0  
 
4
 
4 
       
 #0 
   #%%
2
   


  

     #0Ò 
 #%%
2 "
 #%%
 


56 
 
#0  0 
  #0
   
  #0  
 
#0  #%%
  

 #0 

  #%%
2
,  56 7 0 &  

(
6 
  
 
8
   
   
 
    
 '
 * 


 

* 


0 

 
 "   


, *  
 # 
  
  9 
8
 

 

 
*   

   
( 
  

  
6 
 
 :
* 

  # 

    9 "  
* 


  


    
     
 
 


  
 
 

 
 !
 


  * 
   

0      

  

 6 

 
( (#  
 
 
 
   
    


  (# 7
0

(
 

   
 
 

    
  
  
 
0 


 
6  
 

(# 7
0

( 
0 
6

( 7 - 

6

( 
- 
6

'
   
   


   &
%  
 
 
    *
 
4 

 '
&  
$ 

 

 

%% 

 
  
   
   
   
   
  
      
  

  

 







 
 ; 
  <5 

  0  


 

%% 
 # 0 # 0 -    
<5 <<
%% 

  
 
    

  =>  

   


 
 
   . 
   
 
 
 # << 
%% 
 


 
# 
<< 
  
 

%%
 

  
     
    - 
 

 
 0  $   
0
$5  $ &
<5  

  


  
   


 
 
  
  


  
 

 '    
 
 
 

 


 :

   
 
 # - ( 6 5


 <6 <5
%%  
 

 
  
4  
%%   
 
 0  
  

  
  $6

    



6 . 
0 - 
 
 
 
  =
 
 
>     ? 
 4 
  

 
 
 
 @   =
 
 
>  

    
  


4 
  
  
*


11
00 00
11
00
11 00
11
00
11
11
00

 
 
*   
  
  =
 
 
> 


A 
 

7 # 7 0 7 - &
%

     

 
  

   -   
 # 
   
0   

 
7 # 7 0 & $ 
    

 
    



$%   =
 
 
> B 
  
=
 
 
>   

   


4 
  

   @




- 
       ( 

      #% 


4 
   
5  
 


4 
 
0
   


$ =
 
 
>   
 

4 
 
0  

   

4 
  
*     
.  A1, C 




$%  #% 
5 7 $ & 
 
=
 
 
>
  
5 1  


¡ 7 ¡ 7 ¡ 7


7 Ò
Ò 7




 


 B
  
 
4     
   7   

 7   

 7


  
  
 

 
  
  


    
7  7   & 



#
 < ;  ;      
 /
 , #% 3 
 
 

  .   ,    

 

,

 
  

         

    9  

  

 
 D


  
,

  
 
 
 


   




 





,

 
  

      
    $ 

 
*  
  

,    
,   
 


 !
  


,

  

 

      
  "
  

   % 
   

   
  


  B
  
 
 
   7


   
 
 
 
 
 




*    
   & ½ ½



 
 
 
 
   

  9    
  
 
  7
 &    '

Ò  
  & Ò  
  

 &
 7

½

  7

Ò 
&

 ½

 7

 

$ Ò 


½
 ½

& 
  7  
$ Ò  Ò 


&  7

$


 
½ Ò    
   
 
  



4 
    &   7



   &   @

 

,

 
  

   ,    $ &  

 

 ;

 

 




% !
  
 &  # 

 & 

!
 Ò 

Ò 
Ò  '

Ò  & Ò  0Ò 7 0 Ò   Ò

& Ò  0Ò

 

Ò & Ò  7 0Ò

& Ò Ò 
7 0

B  
 
7 0 &
 
 
 &  &  &


& 
   
 &

 

 
   
 

4     )
:
E      



0

 
 

   
  

  
                     
                           
       !   "                #




$      !             " 
  
  
%  &  '    !     (" %

)

% *
% +   *
)
*!
 
) 

( '                ,       
    -      -!           (  
         -       !        *   - -
  ,       
.        (.         
 #

 
$        /          (    
             (      0!    
 
 1 

) 
   !
) 
2 % ( )  
 3           ! ,  !      3      
      /( 
0(  !  /( 
0     4  1         




$  1         ( 
    (! !  " '    

   
 


! !     -!        

2 5   6 !       / 70  /" (0          5  
/
.0       !      #

 
$        / 70  /" (0     !

      
   / 
  
0 ) / 
 "0          2
% 
) "!    


      6
.    /

.0 !  )   
" &       !             
                   #

 
$ 8                 9  !    
       !            
(      
                    !      
 )   ! 
) (!   )  
 

 
* 9       /  0         .    2! .    2 
  6    / 0               
   #

 
$                !  .     2 ! 

   .      

 (  2 

     !        :   /0   
         /. .0  / 0!   /0 ) / 
0 % (/0!  /0   ;
 !   -  

 < ; (..(  (..(
(..
(..


 
$ 8   ) (..(!    ,     / %
0/ 
0 ) /
 / %
0/ 
00 )
/
 /

00 ) !    

7 3 --      -           / !  ! 
 0       /   0         
          #

 
$ 8                     4  
    - !     - 
(
*.  ! 
)
(
*.






!  )
(
*.
  )
(
*. )
( "!   )  

+ = >   ?
 ;                   
        3    ?
  1   !      -
    - #  1    


  

$ 3      %
          
%  % 

% )     
     ,         )
+       
     !  -  (.     - !       2   )  

.   2      =  "    !       =     =   

.   #

 
$ 8   @      @ =       ) 2 %     " ) (
/  "0!
   ) +  5
.   !     


 5   A!     "     3          "
        !                # =
!                /B       
,      0

 


$    -        !         
              , 2        
      ,
  C           / 
(* 0   "      C     !  ("      !


  "
(" )
("   !          -  
      
 "        
              !    




( >   
(   (2           9 3      
?
.!      (             ,  ?(!       
          ? 9 =        ?
(!     
             ,  ?!            
     ?2    C      >        
 #

 
$           <(     9 3!   
7 
   
7    !       
. %
7  ( %
7   ) ?
.. & 9
= !   
(     
(    !       
(%
(  %
(  2 ) ?+*
 C        ? 

(

 B  7
   ,   ,    !         
B            (             
                        
      #


$ 4  1   !    1  (<     , (< 6        !
 -                 5 !       
         !    !          (<!   
    !             3   
          +!             !   
    -        
                -  8  /0       
       ,         
  1   !  
    ! !  !   !        D /        
  0              
               !        
 / 
0 !      
        ! /0  / 
0
 
 !           / 
0   !   ! /0 ) / 
0 
/.0 )
/            D           E0!
    /
0 ) ! /(0 ) +!  /0 ) (<!        7
 /3     
  ! /0 )  0 !         !         

2 3            
.             
    !           D !       
 #

 
$ 8                   !   
$         !    ,     
.   !
    ( . )
.(2     !         .   !  
 !       

" 3                       
               ("#

 
$   
% 
) *(" 4   -         !  -

"! (.! ("  <! (2! ("      ("      3      
            


*   
 (
) / % 0   4            
      

  

$    @  @   
   (
  %  ) .  
   /

           @   0!    ) 



 4
    !  ,         !     1     
  +
%
( 
(  .     
      

< F 1        /  0              
          G  , !   /. .0! /
 (0!  /( 20    
       !       .    2  .    2! 
         #

 


 5            
.         
 
   $
 
 
 
 
 
     

         $   "  

!
                

7 3 @      
  3  -      !   

                -#

    

$           
           
    

      

+ 5      1   /0     
!     #

 
$ + 
 + ) <(+ 
. ) 
(.
 ! H  H              

...!   
           +#

 
$      
.      +     ! H     +
   
..    !      +     
..     +  
  
..  !     
       +
(
 8              
.. 8               < 8 
                     D    /   0#


$          ! *  +    <  /   0 I   
        !          
..          ! *!
 +  ! 
! <! *
! *<!  +<         <  <!
<! <! 2<! *<!  +< 
 * % *   ) 
(( 4       "
%  % *

       
  

$                (  

  
!             @     

     

 Æ !      1     /
   (0/
 0!     
 D         
(      2<  2< ,        #

 

$ 5  ) 2<  
. ) 2< 
.. ) 2<<!  +. ) 2!      )   ! 


   )  
 
  
(2                      )
  5   "  
    @  !      #

 

2
 $                 !      )    ) 
 !
   "       @  !      ) "   !     ) "
   !
    "
 * %
) .!     )
 )     )  !   )
!  )  
(" =   H  
.            "(6   -    /
- 0                      <  #





$     <  !        +     3   
  !
    <     !        2<   !



  
          -   
.       
/0

   
)
 
/0
(*     !  
.6++  !            @  

..  @   @        #

 

$ 8               /
. %  0
)
..
% (. % 
 ! 

      
..   @  
!   
..   (. !  @  "
        !    "!    "  .  
.        " 

7     .  "!  (        /".  ""0!   
.%
7  ( ) 

  
(< A    -    ,       !        ,
        !                   
           # /5          - -
 !        0



$ 3     !  !   !   A         

4 !      /      0 5         1 
 !            6        !   
  


: ,!            1     
) *   -  
 !         :             !  
       !     2        ( % *  ( )

     
(7 ;  !    !    !       -        
    !  - 1  "          !  - 1 
 
(
       -  1  -       #

 
 
$ 8                  -   !      ) 

  /  % 0 )
 ! /  %  % 0 )  %  )  


 !   -    1  -


(+ 8       6       D     
           /6
!.0    /!"0      
   /       D    0#

 

$      "      2  D   /      0
   
     +      "    !   /    
  0
    5      !     2  !   D 



            D      7    
        
(* % "
7 ) 

 

"
.    
         @      1     )
  )  %   
     4  

 
$    J6   J  1    @   :   )   %  
)

%   ( %  
)   %    %   ( %  
 5   -    !    
 )  %
% ( %  %  
       
% ( %  %  
 
 !       
  ) 
 %
          ) 

 %
)  

 4          ( (< 

 
$ (< )   3! (      (
  (
) "
(!    ) (2  
    "
(
(  
..         *      3 3 2"
  = = 
 .    >    > 4   ! (

 /      , 0  3   =  
"    =  
 > +                3  
>#

 
$ &     6 ,!   @     !   
        3  >$

/* % 2" % .0  /(
%
" % 0 % + )
..

 1    ) 

 5   %  
) 2    ! 1   

  
$ 8   )       )      
 ) 
   
) 
  !
 
) 
      % 
) 2 H   ,!    @  
 2 % ( ) .

   (
(   !        )   
   

 
 
 

*


 

  



  



     

              





 
Ô
Ô






 

 
   

 

  

    

    
  
 
! "

         #         

   

$
 

Ô 


% "%  $
          " "


$
   

 
&
 
  


 Ô
 Ô





 '     
(

) 

   * 
 + )
,
    -

 



 
"              
 &     
   $"

  $"    
               
"
$
      
   
     .      
  

) / % $


   

     $


    

 

-

 
  

 & $
 "
0
  


 

.






)

  " %    
$


    

 


 ) 1
  )
 $
  
 
 %    
   $


    

 )
  
* #    %  
2

    $
 -

  

 #   ""
   
2

 $
    
 "
"    ! 


   %
 (" $    "
"   "    (  

 &  

  

 
"$  %   
  %   "
"    %    !  


      &  
  
 %  "" "   

3 #  
   


-

1
4

3
2

 

 '  
 
    
  $ 
 
  $
"    

 5

1 1
5
4
2
4
3

3
2

4
 
  
 
       
   
   

$    
  

%
 
 
       
  
      
          ! 
 
 
   "  
    

$     "  $$ 
 ("%  
  
   "
 $$ 
 ("%   
   ! 5
   


 

  
  3
  


" 
   !
 
     $  
 
 )  &   



   )
  3



  


+ '        "  

  
 
 
 
   67

 
8*+*.
  

 
   %   
(   %     ("% )
   ,
   " 

 28
 -

 
#       
 9   6   ::: 

 ::: 

 :::  :::7  
" :::
   2

   
  
 

    
  9  !     2

   ::: "    *    &
 
 
  % %     $$        

   #   
  
  
 6 "

7  !
 " "" 
 :::    
    
  *2  
     ("% )     
     % :::  
 *   ;

    
<     "$  
 $
    6"

7 
" 


% 
             %  / "       

 !    ::: 
 *    #    

 ).)   % :::

 ).) 

   " 


  ¼     ¼ 

. = !         
    4$ 

 

     &    4" ¼
           ¼
   
 
> #   * 
     
  

  

 *
" 
-

 

 
"  5
"   
 %   
     
   
"    $$%

       
   
¾ ¾


   % %   

 

    "


  
   
  ($
   5
  
  
"  %   " % $

 
¾
   
   

¾  

   
 *
*   
" *   * *  8+   


   


 

 
  
  * )    )
8
  "
" $%
    "   
 
    3  ) 
*   *  )

       

 '         
    &  
   
  $
"  
 
   $

  
 %   5


       

>  *      3
.  8  )   
*
+* 
+  *    )

&  
 ?        
)  
  
+    
    
-

 ,


 
 )


 


&  $  $%    
9       $
 "
" 5
 @
%  
$%
  



       
  % 
  
"     $


   
     ; "    
<   %   $ 
 " %
"
  
   
 5 

    





 &  $              

  !  

       
 

    






        





 & A  B  # 5
% % 
     

  >8
 
C
 
 " !   %    $
  $


       
 % *> 



 


    ""
 #   
 %  
 
 5
%  

*>

 
 
 >8


*>   >8 
 




  
>8 
 

*
 >8 
 


**  
 *


>8   

**  
  *


@ %  A  B  # 5

%






 

   


 

 *

>>   




**  
  *


     


)

     
 

    
 D "      (
 " "  
"


  
    D  
) $

%  9"9
 E <   
  "
     
3 ""  9"9
 E D<   "  
"
  
 ""
 
 
 
    
3 ""  
 %% 
 
 1 

 $
 
   D   
  9"9   
     "
  
D
   D 
  "     
  $

%  D   

 -

 



  <
  "  D 
 # % 
  "  ""  D  
  
          "
     

  B 
   ""  D
  
          "
 
   
  / "   $

%  D  
    
 D
   D       ""     
  
 
 

  
$

%  D   
         " 

  %  $

% 
      " &



  3 

).

1
"   
<      9    

) 



    ½·½



Ô

  





  
 5   "
  "
  
       

 
 
&
 %
  5


   
 
      
3 &   


"   
  
   
"   "
  "
 "  
  




    ½·½


 

  
 ½·½

  

  
  / "  
      

* ,
    $


   " 

  *

  $
 -

 

 C
  
 
"  " 
%  
0    
  *
"   



 % ) 3 .


)
. 
8  ($ 
) 

    ) " 


  *
 #




  % )   
 
    )   
 
 )    

 *) '
9 
 



)   
      )   
 
    )    

 ** , 9 


 * 
    *
  "$
  F " 5
"9% " "9    

    
.
 )8   $
 


3 '       5 
(  
   "
    $$       
   &   ¼
  
 
$%  
(   % 
E  

 " %   
  $ "
  ,
   
      ¼     % 
E  

 ""
 ¼
  


  ¼    ¼  
   

    " "9
   $"   
  9  
"  
"  
    
  
 
 
% 
( 
¼      ¼
  
  ¼    ¼   

*
 Algebra Test
2004 Stanford Mathematics Tournament
February 28, 2004
1. How many ordered pairs of integers (a,b) satisfy all of the following inequalities?

a2 + b2 < 16
a2 + b2 < 8a
a2 + b2 < 8b

2. Find the largest number n such that (2004!)! is divisible by ((n!)!)!.

3. Compute:
20053 20033
b − c.
2003 · 2004 2004 · 2005
Where bxc denotes the greatest integer less than x.
4. Given that:

x > 0, y > 0, z > 0,

xy − 3x − 7y + 15 = 0,
xz − 2x − 7z + 8 = 0,
yz − 2y − 3z + 2 = 0.

Solve for x, y, and z.

5. There exists a positive real number x such that cos(tan−1 (x)) = x. Find the value of x2 .
6. Dan, Shravan, and Jake attempt to clean their room for the first time in six months. It would take
Dan 3 hours to clean it by himself. It would take Shravan 10 hours and Jake 15 hours to clean the
room. How long does it take the three of them together to clean the room?
7. Find all real x such that
x(x − 2)(x − 4)
< 0.
(x − 1)(x − 3)(x − 5)

8. Let x be a real number such that x3 + 4x = 8. Determine the value of x7 + 64x2 .

9. A sequence of positive integers is defined by a0 = 1 and an+1 = a2n + 1 for each n ≥ 0. Find
gcd(a999 , a2004 ).
10. There exists a polynomial P of degree 5 with the following property: if z is a complex number such
that z 5 + 2004z = 1, then P (z 2 ) = 0. Calculate the quotient PP(−1)
(1)
.

1
 Algebra Test Solutions
2004 Stanford Mathematics Tournament
February 28, 2004
1. Answer: 6
This is easiest to see by simply graphing the inequalities. They correspond to the (strict) interiors of
circles of radius 4 and centers at (0, 0), (4, 0), (0, 4), respectively. So we can see that there are 6 lattice
points in their intersection (circled in the figure).

2. Answer: 6
For positive integers a, b, we have
a! | b! ⇔ a! ≤ b! ⇔ a ≤ b.
Thus,
(n!)!)! | (2004!)! ⇔ (n!)! ≤ 2004! ⇔ n! ≤ 2004 ⇔ n ≤ 6.

3. Answer: 8
Let x = 2004. Then the expression inside the floor brackets is

(x + 1)3 (x − 1)3 (x + 1)4 − (x − 1)4 8x3 + 8x 16x

− = = 3 =8+ 3 .
(x − 1)x x(x + 1) (x − 1)x(x + 1) x −x x −x

Since x is certainly large enough that 0 < 16x/(x3 − x) < 1, the answer is 8.
4. Answer: x = 10, y = 5, z = 2
Factoring,

(x − 7)(y − 3) = 6
(x − 7)(z − 2) = 6
(y − 3)(z − 2) = 4

This implies that

x−7 = 3
y−3 = 2
z−2 = 2

Thus x = 10, y = 5, z = 4.

1
 √
−1+ 5
5. Answer: 2
Draw a right triangle with legs 1, x; then the angle θ opposite x is tan−1 x, and √ we can compute
cos(θ) = √x12 +1 . Thus, we only need to solve x = √x12 +1 . This is equivalent to x x2 + 1 = 1. Square
both sides√ to get x4 + x2 = 1 ⇒ x4 + x2 − 1 = 0. Use the quadratic formula to get the solution
x2 = −1+2 5 (unique since x2 must be positive).
6. Answer: 2hours.
1 1 1 1
Adding the individual rates, we get 3 + 10 + 15 = 2 of the room is cleaned per hour so the whole room
takes two hours.
7. Answer: 0 < x < 1 or 2 < x < 3 or 4 < x < 5
The sign of one of the terms switches every time x moves from the range (I, I + 1) to (I + 1, I + 2).
When x is less than zero, all terms are negative so the LHS is negative. Also note that x is undefined
at 1, 3, 5.
8. Answer: 128
For any integer n ≥ 0, the given implies xn+3 = −4xn+1 + 8xn , so we can rewrite any such power of x
in terms of lower powers. Carrying out this process iteratively gives

x7 = −4x5 + 8x4
= 8x4 + 16x3 − 32x2
= 16x3 − 64x2 + 64x
= −64x2 + 128.

Thus, our answer is 128.

9. Answer: 677
If d is the relevant greatest common divisor, then a1000 = a2999 + 1 ≡ 1 = a0 (mod d), which implies
(by induction) that the sequence is periodic modulo d, with period 1000. In particular, a4 ≡ a2004 ≡ 0.
So d must divide a4 . Conversely, we can see that a5 = a24 + 1 ≡ 1 = a0 modulo a4 , so (again by
induction) the sequence is periodic modulo a4 with period 5, and hence a999 , a2004 are indeed both
divisible by a4 . So the answer is a4 , which we can compute directly; it is 677.
10. Answer: − 2010012
2010013
Let z1 , . . . , z5 be the roots of Q(z) = z 5 + 2004z − 1. We can check these are distinct (by using the
fact that there’s one in a small neighborhood of each root of z 5 + 2004z, or by noting that Q(z) is
relatively prime to its derivative). And certainly none of the roots of Q is the negative of another, since
z 5 + 2004z = 1 implies (−z)5 + 2004(−z) = −1, so their squares are distinct as well. Then, z12 , . . . , z52
are the roots of P , so if we write C for the leading coefficient of P , we have

P (1) C(1−z12 )...(1−z52 )

P (−1) = C(−1−z12 )...(−1−z52 )
[(1−z1 )...(1−z5 )]·[(1+z1 )...(1+z5 )]
= [(i−z1 )...(i−z5 )]·[(i+z1 )...(i+z5 )]
[(1−z1 )...(1−z5 )]·[(−1−z1 )...(−1−z5 )]
= [(i−z1 )...(i−z5 )]·[(−i−z1 )...(−i−z5 )]
(15 +2004·1−1)(−15 +2004·(−1)−1)
= (i5 +2004·i−1)(−i5 +2004·(−i)−1)
(2004)(−2006)
= (−1+2005i)(−1−2005i)
2
= − 2005 −1
20052 +1
= − 4020024 2010012
4020026 = − 2010013 .

2
 Stanford Mathematics Tournament
February 28, 2004

Individual Round: Geometry Subject Test — Problems

1. In the diagram below, the outer circle has radius 3, and the inner circle has radius 2. What
is the area of shaded region?

2. A parallelogram has 3 of its vertices at (1, 2), (3,8), and (4, 1). Compute the sum of the
possible x-coordinates for the 4th vertex.
3. AC is 2004. CD bisects angle C. If the perimeter of ABC is 6012, find (AC × BC)/(AD ×
BD).
C

A
D B

4. P is inside rectangle ABCD. P A = 2, P B = 3, and P C = 10. Find P D.

5. Find the area of the region of the xy-plane defined by the inequality |x| + |y| + |x + y| ≤ 1.
6. We inscribe a square in a circle of radius 1 and shade the region between them. Then we
incribe another circle in the square and another square in the new circle and shade the
region between the new circle and square. After we have repeated this process infinitely
many times, what is the area of the shaded region?

1
 7. Yet another trapezoid ABCD has AD parallel to BC. AC and BD intersect at P . If
[ADP ]/[BCP ] = 1/2, find [ADP ]/[ABCD]. (Here the notation [P1 · · · Pn ] denotes the area
of the polygon P1 · · · Pn .)

8. A triangle has side lengths 18, 24, and 30. Find the area of the triangle whose vertices are
the incenter, circumcenter, and centroid of the original triangle.

9. Given is a regular tetrahedron of volume 1. We obtain a second regular tetrahedron by

reflecting the given one through its center. What is the volume of their intersection?

10. Right triangle XY Z has right angle at Y and XY = 228, Y Z = 2004. Angle Y is trisected,
and the angle trisectors intersect XZ at P and Q so that X, P, Q, Z lie on XZ in that order.
Find the value of (P Y + Y Z)(QY + XY ).

2
 Geometry Solutions
2004 Stanford Mathematics Tournament
February 28, 2004
√
1. Answer: 4 2
Consider the following diagram, where O is the center of the larger circle: Since the smaller circle has
Q

O P x

radius 2, we must have x = 2 and OP = 1. And OQ is a radius of the larger circle, so OQ = 3.

Applying the Pythagorean theorem to triangle OP Q, we find that

12 + y 2 = 32 .
√
Solving this for y gives us y = 2 2.
The shaded region is a triangle with base 2y and height x. Therefore its area is
1 √
(x)(2y) = xy = 4 2.
2

2. Answer: 8
There are 3 possible locations for the 4th vertex. Let (a, b) be its coordinates. If it is opposite to
vertex (1, 2), then since the midpoints of the diagonals of a parallelogram coincide, we get ( a+1 b+2
2 , 2 )=
3+4 8+1
( 2 , 2 ). Thus (a, b) = (6, 7). By similar reasoning for the other possible choices of opposite vertex,
the other possible positions for the fourth vertex are (0, 9) and (2, −5), and all of these choices do give
parallelograms. So the answer is 6 + 0 + 2 = 8.
3. Answer: 4
Let AC = d, BC = c, AD = a, BD = b so ad = cb . Now a+b c = ac + bc = ac + ad = a( 1c + d1 ),
a+b a b b b 1 1 c+d 1 1 1 cd a+b c+d 2 cd
d = d + d = c + d = b( c + d ), a+b = 1 + 1 × ( a + b ) = c+d × ab . This implies that ( a+b ) = ab .
c d
cd
Now a + b = 2004, c + d = 6012 − 2004 = 4008 so ab = ( 4008 2
2004 ) = 4.
√
4. Answer: 95

A y E z B

w w
P
H F

x x

D y G z C

Draw perpendiculars from P to E on AB, F on BC, G on CD, and H on DA, and let AH = BF = w,
HD = F C = x, AE = DG = y, and EB = GC = z. Then P A2 = w2 + y 2 , P B 2 = w2 + z 2 , P C 2 =
x2 + z 2 , and 2 2 2 2 2 2 2
√ P D = x + y . Adding and subtracting, we see that P D = P A − P B + P C = 95,
so P D = 95.

1
 Sector Inequality Simplified inequality
x ≥ 0, y ≥ 0, x + y ≥0 x + y + (x + y) ≤ 1 x + y ≤ 12
x ≥ 0, y ≤ 0, x + y ≥0 x − y + (x + y) ≤ 1 x ≤ 12
x ≥ 0, y ≤ 0, x + y ≤0 x − y − (x + y) ≤ 1 y ≥ − 12
x ≤ 0, y ≥ 0, x + y ≥0 −x + y + (x + y) ≤ 1 y ≤ 12
x ≤ 0, y ≥ 0, x + y ≤0 −x + y − (x + y) ≤ 1 x ≥ − 12
x ≤ 0, y ≤ 0, x + y ≤0 −x − y − (x + y) ≤ 1 x + y ≥ − 21

3
5. Answer: 4
To graph this region we divide the xy-plane into six sectors depending on which of x, y, x + y are ≥ 0,
or ≤ 0. The inequality simplifies in each case:
We then draw the region; we get a hexagon as shown. The hexagon intersects each region in an isosceles
right triangle of area 18 , so the total area is 6 · 18 = 34 .

6. Answer: 2π − 4
√
The first square has side length of 2. The second circle has half of the area of the first one. Thus
the shaded region has half of the area of the first one. The second shaded area is half of the size of the
first. But the sum of 21n from zero to infinity is 2 so the shaded area is 2(π − 2) = 2π − 4.
√
7. Answer: 3 − 2 2
A homothety (scaling) about P takes triangle ADP √ into BCP , since AD, BC are parallel and A, P, C;
B, P, D are collinear. The ratio
√ of homothety is thus 2. It follows that, if we rescale to put [ADP√] = 1,
then [ABP ] = [CDP ] = 2, just by the ratios of lengths of bases. So [ABCD] = 3 + 2 2, so
[ADP ] 1√
√
[ABCD] = 3+2 2 . Simplifying this, we get 3 − 2 2.

8. Answer: 3
There are many solutions to this problem, here is one. The given triangle is a right 3-4-5 triangle, so
the circumcenter is the midpoint of the hypotenuse. Coordinatizing for convenience, put the vertex at
(0, 0) and the other vertices at (0, 18) and (24, 0). Then the circumcenter is (12, 9). The centroid is
at one-third the sum of the three vertices, which is (8, 6). Finally, since the area equals the inradius
18·24
2
times half the perimeter, we can see that the inradius is 18+24+30 = 6. So the incenter of the triangle
2
is (6, 6). So the small triangle has a base of length 2 and a height of 3, hence its area is 3.
1
9. Answer: 2
Imagine placing the tetrahedron ABCD flat on a table with vertex A at the top. By vectors or
otherwise, we see that the center is 34 of the way from A to the bottom face, so the reflection of this
face lies in a horizontal plane halfway between A and BCD. In particular, it cuts off the smaller
tetrahedron obtained by scaling the original tetrahedron by a factor of 12 about A. Similarly, the
reflections of the other three faces cut off tetrahedra obtained by scaling ABCD by 12 about B, C, and
D. On the other hand, the octahedral piece remaining remaining after we remove these four smaller
tetrahedra is in the intersection of ABCD with its reflection, since the reflection sends this piece to
itself. So the answer we seek is just the volume of this piece, which is

2
 (volume of ABCD) − 4 · (volume of ABCD scaled by a factor of 12 )
= 1 − 4( 21 )3 = 12 .

10. Answer: 1370736

X
P
Q

Y Z

The triangle’s area is 228·2004

2 = 228456. All the angles at Y are 30 degrees, so by the sine area formula,
the areas of the three small triangles in the diagram are QY 4·Y Z , P Y 4·QY , and XY 4·P Y , which sum to
the area of the triangle. So expanding (P Y + Y Z)(QY + XY ), we see that it equals

4 · 228456 + XY · Y Z = 6 · 228456 = 1370736.

3
 Calculus Test
2004 Stanford Mathematics Tournament
February 28, 2004
√
1. Evaluate limx→∞ ( 4x2 + 7x − 2x).
2. Suppose the function f (x) − f (2x) has derivative 5 at x = 1 and derivative 7 at x = 2. Find the
derivative of f (x) − f (4x) at x = 1.

3. An object moves along the x-axis with its position at any given time t ≥ 0 given by x(t) = 5t4 − t5 .
During what time interval is the object slowing down?
4. For x > 0, let f (x) = xx . Find all values of x for which f (x) = f 0 (x).
5. The highway department of North Eulerina plans to construct a new road between towns Alpha and
Beta. Town Alpha lies on a long abandoned road running east west. Town Beta lies 3 miles north and
5 miles east of Alpha. Instead of building a road directly between Alpha and Beta, the department
proposes renovating part of the abandoned road (from Alpha to some point P ) and then bulding a new
road from P to Beta. If the cost of restoring each mile of old road is $200,000 and the cost per mile of
a new road is $400,000, how much of the old road should be restored in order to minimize costs?

6. Consider the two graphs y = 2x and x2 − xy + 2y 2 = 28. What is the absolute value of the tangent of
the angle between the two curves at the points where they meet?
7. A mouse is sitting in a toy car hooked to a spring launching device on a negligibly small turntable.
The car has no way to turn, but the mouse can control when the car is launched and when the car
stops (the car has brakes). When the mouse chooses to launch, the car will immediately leave the
turntable on a straight trajectory at 1 m/s. Suddenly someone turns on the turntable; it spins at 30
rpm. Consider the set of points the mouse can reach in his car within 1 second after the turntable is
set in motion. What is the area of this set?
8. A spherical cow is being pulled out of a deep well. The bottom of the well is 100 feet down and the
cow and all the water on him weighs 200 pounds. He is being hauled up at a constant rate with a chain
which weighs 2 pounds per foot. The water on the cow drips off the cow at the rate of 21 pounds per
foot as he is being hauled up. How much work is required to rescue the cow in foot-pounds? Remember
work is force times distance.
9. The base of a solid is the region between the parabolas x = y 2 and 2y 2 = 3 − x. Find the volume of
the solid if the cross-sections perpendicular to the x-axis are equilateral triangles.

10. Find the positive constant c0 such that the series

∞
X n!
(cn) n
n=0

converges for c > c0 and diverges for 0 < c < c0 .

1
 Calculus Solutions
2004 Stanford Mathematics Tournament
February 28, 2004
1. Answer: 74
√ √ √ 1
( 4x2 +7x+2x) 7x
limx→∞ ( 4x2 + 7x − 2x) = limx→∞ ( 4x2 + 7x − 2x) · √
( 4x2 +7x+2x)
= limx→∞ √
( 4x2 +7x+2x)
· x
1 =
x
limx→∞ √ 7 7 = 47 .
( 4+ x +2)

2. Answer: 19
The derivative of f (x) − f (2x) is f 0 (x) − 2f 0 (2x). So f 0 (1) − 2f 0 (2) = 5, f 0 (2) − 2f 0 (4) = 7. Thus

f 0 (1) − 4f 0 (4) = (f 0 (1) − 2f 0 (2)) + 2(f 0 (2) − 2f 0 (4)) = 5 + 2 · 7 = 19,

the answer.
3. Answer: [3,4] or (3,4) or from t=3 to t=4
The velocity of the object is given by v(t) = x0 (t) = 20t3 − 5t4 , and the acceleration function is
a(t) = v 0 (t) = 60t2 −20t3 . The object is slowing down when the velocity is positive and the acceleration
is negative, or vice versa. v(t) is positive from t = 0 to t = 4 and is negative after that. a(t) is positive
from t = 0 to t = 3 and negative afterward. These only differ in sign from t = 3 to t = 4.
4. Answer: 1
f 0 (x)
Let g(x) = log f (x) = x log x. Then f (x) = g 0 (x) = 1 + log x. Therefore f (x) = f 0 (x) when
1 + log x = 1, that is, when x = 1.
√
5. Answer: 5 − 3 miles
p
Let x be the amount of old road restored. Then the length of the new road is 9 + (5√− x)2 using the
2
Pythagorean Theorem. Thus the total cost of the plan is C(x) = 200000x √ + 400000 x − √ 10x + 34.
The minimum cost occurs at one of √ the critical points which are x = 5 ± 3. Clearly 5 + 3 is not a
valid answer and one can check 5 − 3 is indeed a minimum.
6. Answer: 2
The two graphs intersect at x2 − 2x2 + 8x2 = 28 or rather x = ±2 with y = ±4. At x = +2, m1 = 2
and m2 = y 0 (2). Using implicit differentiation on the second graph, we find y 0 (x) = y−2x
4y−x and plugging
m2 −m1
in (2, 4) gives a slope of 0. If α is the angle between the graphs then | tan(α)| = | 1+m 1 m2
|. Plugging
in the values yields the answer 2. x = −2 yields the same value.
7. Answer: π/6
The mouse can wait some amount of time while the table rotates and then spend the remainder of
the time moving along that ray at 1 m/s. He can reach any point between the starting point and the
furthest reachable point along the ray, (1 − θ/π) meters out. So the area is
Z π Z π
(1/2)(1 − θ/π)2 dθ = (1/2)(1/π)2 θ2 dθ = π/6.
0 0

8. Answer: 27500 foot-pounds

Let x indicate the distance the cow has yet to travel. Then the work for a distance dx is (2x + 200 −
1
R 100 5
2 (100 − x))dx. Thus the total work is 0 ( 2 x + 150)dx = 27500 foot-pounds.
√
3 3
9. Answer: 2
The base region is bounded on the left by x = y 2 and on the right by 2y 2 = 3 − x. The intersection
points are (1, 1) and (1, −1). Each cross-section, say x = a, is an equilateral triangle. The length

1
 √ q
3−x
of a side is 2y where y = x for a ≤ 1 but it is y = for a ≥ 1. The area of an equilateral
2
√
3s2
R1 √ √ 2 R3 √ q
triangle is 4 where s is the side length. Thus the volume is 0 3(24 x) dx + 1 43 (2 3−x 2
2 ) dx =
√ R1 √ R3 √
3 0 xdx + 23 1
(3 − x)dx = 3 2 3 .
1
10. Answer: e
The ratio test tells us that the series converges if
(n+1)!  n
(c(n+1))n+1 1 n
lim n!
= · lim
n→∞
(cn)n
c n→∞ n+1

is less than one and diverges if it is greater than one. But

 n  −n
n 1 1
lim = lim 1+ = .
n→∞ n+1 n→∞ n e
1 1
Then the limit above is just ce , so the series converges for c > e and diverges for 0 < c < 1e .

2
 Advanced Topics Test
2004 Stanford Mathematics Tournament
February 28, 2004
     
1 1 1 3 1 5 1 99
1. Find ··· .
0 1 0 1 0 1 0 1
2. Write −.110 + .19 − .18 + .17 as a repeating decimal in base 6.
3. Express 2 cos(10◦ ) + sin(100◦ ) + sin(1000◦) + sin(10000◦) as sin(x◦ ) where −90 ≤ x ≤ 90.
4. Find x in the diagram below.

5 9
x
6
8 30

5. How many rearrangements of the letters ”STANFORD” contain the letters D, A, F, T in that order
(although not necessarily consecutively)?
6. Write F2006 + F2002 + F2000 + F1998 + · · · + F2 + F0 as aFb where a and b are positive integers and
where F0 = 1, F1 = 1 and Fn = Fn−1 + Fn−2 for n ≥ 2.
7. Adam, Bill, and Chris are playing a game of King of the Hill. Given who is king one day, the probability
of who is king the next day is given by the chart below. If Adam is king on day 1, which day is Chris
expected to first become king? (Leave your answer as a simplified fraction.)
XX n + 1:
X
Day n:XX
X Adam Bill Chris
1 1 1
Adam 3 3 3
1 1 1
Bill 2 4 4
Chris 0 0 1
√
8. Compute (i + 1)2004 − (i − 1)2004 , where i = −1
9. How many solutions does 2004 cos x = x2 have?
10. Allen and Zorro decide to play a game using a deck of cards. They alternate turns picking a card from
the deck with Allen going first. The first person to draw an ace wins. What is the probability that
Allen wins?

1
 Advanced Topics Solutions
2004 Stanford Mathematics Tournament
February 28, 2004
 
12500
1. Answer:
0 1
    
1 a 1 b 1 (a + b)
First note that = . This implies the given product is
0 1 0 1 0 1
1 502
     
1 (1 + 3 + 5 + ... + 99) 1 2500
= = .
0 1 0 1 0 1

2. Answer: .010136 .
1 1
9 − 10 + 17 − 18 = 9·8·7·5
73
. Note this is all base 10. To convert to a base 6 decimal, we get each digit by
multiplying the fraction by 6 and taking the integer portion. This is the reverse of converting integers
to base 6. Note that the decimal begins to repeat after the first five places. The answer is .010136 .
3. Answer: sin(80◦ )
The expression is equivalent to cos(10◦ )+cos(10◦ )+sin(100◦ )+sin(280◦ +2·360◦ )+sin(280◦ +27·360◦ ) =
cos(10◦ ) + cos(10◦ ) + sin(100◦ ) + sin(280◦ ) + sin(280◦ ) = cos(10◦ ) + cos(10◦ ) + cos(10◦ ) − cos(10◦ ) −
cos(10◦ ) = cos(10◦ ) = sin(80◦ ).
√
4. Answer: 3 2
Using the law of cosines in the right triangle, we find 92 = 62 + 52 − 2(5)(6) cos θ, where
√
θ is the angle
1 2 2 8
the triangles have in common. Thus cos θ = − 3 . Since sin θ + cos θ = 1, sin θ = 3 . Using the law
√
of sines in the left triangle yields sinx30◦ = sin8 θ . Thus, x = 3 2.
5. Answer: 1680
There are 8! orders for RICEOWLS and each of the 4! orders of WISE are equally likely within them,
so 8!
4! = 8 ∗ 7 ∗ 6 ∗ 5 = 1680 have the correct order.

6. Answer: 2F2005 .
F0 = 1 = F1 . Then F1 + F2 = F3 . Then F3 + F4 = F5 . And so on until we get F2006 + F2003 . This
equals F2005 + F2004 + F2003 = 2F2005 .
17
7. Answer: 4
Let X be the number of days until Chris is king if Adam is king and Y be the number of days until
Chris is king if Bill is king. Then X = 1 + X Y 0 X Y 0
3 + 3 + 3 and Y = 1 + 2 + 4 + 4 . Solving for X we get
13 17
4 but we want expected day, not ”days until” so we add 1 day to get 4 .

8. Answer: 0
(i+1)4 is in the direction of -1. Also, (i−1)4 is in the direction of -1 and they have the same magnitude.
Call (i + 1)4 = n. Then this is n501 − n501 = 0.
9. Answer: 30
x2
√
It’s best to rewrite it as cos x = 2004 and first to consider only positive values. Clearly, x < 2004 in
agreement with the range of cos x. We’ll definitely have 2 solutions for every interval [2π∗(n−1), 2π∗(n)]
for
√
n = 1, 2, ·, m for some m. It’s not hard to see that m is the largest integer that does not exceed
2004 2 2 22 22.5
2π . Since 44 < 2004 < 45 , 44 < x < 45. The expression is hence between π and π . Note
22 22 22.5
that 7 > 3.142 > π, so π > 7 and m = 7 (since π < 8). But at the end of the seventh inteval,
x2 2
π2 3.152
x = 14π, and 2004 = 49π
501 < 10 < 10 < 1 = cos 14π. Hence there must be at least one more solution.
x2
There cannot be more than one in [14π, 15π] since cos x decreases and 2004 increases. Note also that
(15π)2 2 2 2

2004 = 225π 225π π

2004 > 2025 = 9 > 1, so x ≥ 15π yields no solutions. There are a total of 2(7) + 1 = 15
solutions for positive x and hence 30 overall.

1
 433
10. Answer: 833
Instead of thinking about people picking cards, we will place
the aces in the deck. So we assign a
number between 1 and 52 to each of the aces. There are 52 4 ways to do this. Now we examine what
happens if the first ace is in an odd numbered slot. If the first ace is number 1, we have 51


3 possibilities
for the other 3 aces. Similarly, for slot 3 we have 49


3 and so forth. So the probability can be written as
(51 49 47 3
3 )+( 3 )+( 3 )+...+(3)
P∞ 1
. We can put the numerator into summation form as n=1 6 (2n + 1)(2n)(2n − 1).
(52
4)
2 2
This is 2∗25 ∗266 −25∗26 = 844350 = 140725. Then 140725 = 433
6 (52
4)
833

2
 Stanford Mathematics Tournament
February 28, 2004

Individual Round: General Test — Problems

1. Add this to that, divide by three,
The square of this of course you’ll see,
If that to this is eight to one,
Find that and you’ll be done.
2. In the Championship World Series of the future, Rice and Stanford play three games or
until one team wins two games. In each game, both teams have a 1/3 chance of winning,
and there is a 1/3 chance that they will tie. What is the probability that Rice wins the
championship?
3. Find all natural numbers n that satisfy the equation:
(n − 1) + (n − 1) = n!

4. Find the ordered 4-tuple of positive integers (a, b, c, d) so that 37/13 = a + 1/[b + 1/(c + 1/d)]
5. Each vertex of a given square is the center of one of four circles. The circles are all congruent
and each one is tangent to two others. What is the probability that a randomly chosen point
in the figure is inside both a circle and the square?
6. There are 1000 rooms in a row along a long corridor. Initially the first room contains 1000
people and the remaining rooms are empty. Each minute, the following happens: for each
room containing more than one person, someone in that room decides it is too crowded and
moves to the next room. All these movements are simultaneous (so nobody moves more
than once within a minute). After one hour, how many different rooms will have people in
them?
7. What is the largest whole number that is equal to the product of its digits?
8. Suppose f is a function that assigns to each real number x a value f (x), and suppose the
equation
f (x1 + x2 + x3 + x4 + x5 ) = f (x1 ) + f (x2 ) + f (x3 ) + f (x4 ) + f (x5 ) − 8
holds for all real numbers x1 , x2 , x3 , x4 , x5 . What is f (0)?
9. How many ways can you mark 8 squares of an 8 × 8 chessboard so that no two marked
squares are in the same row or column, and none of the four corner squares is marked?
(Rotations and reflections are considered different.)
√
10. A rectangle has perimeter 10 and diagonal 15. What is its area?
11. Find the ordered quadruple of digits (A, B, C, D), with A > B > C > D, such that
ABCD
− DCBA
= BDAC.

1
12. Let ACE be a triangle with a point B on segment AC and a point D on segment CE such
that BD is parallel to AE. A point Y is chosen on segment AE, and segment CY is drawn.
Let X be the intersection of CY and BD. If CX = 5, XY = 3, what is the ratio of the area
of trapezoid ABDE to the area of triangle BCD?

13. You have a 10 × 10 grid of squares. You write a number in each square as follows: you write
1, 2, 3, . . . , 10 from left to right across the top row, then 11, 12, . . . , 20 across the second row,
and so on, ending with a 100 in the bottom right square. You then write a second number in
each square, writing 1, 2, . . . , 10 in the first column (from top to bottom), then 11, 12, . . . , 20
in the second column, and so forth.
When this process is finished, how many squares will have the property that their two
numbers sum to 101?

14. Urn A contains 4 white balls and 2 red balls. Urn B contains 3 red balls and 3 black balls.
An urn is randomly selected, and then a ball inside of that urn is removed. We then repeat
the process of selecting an urn and drawing out a ball, without returning the first ball. What
is the probability that the first ball drawn was red, given that the second ball drawn was
black?

15. A floor is tiled with equilateral triangles of side length 1, as shown. If you drop a needle
of length 2 somewhere on the floor, what is the largest number of triangles it could end up
intersecting? (Only count the triangles whose interiors are met by the needle — touching
along edges or at corners doesn’t qualify.)

16. Find the largest number n such that (2004!)! is divisible by ((n!)!)!.

17. Andrea flips a fair coin repeatedly, continuing until she either flips two heads in a row (the
sequence HH) or flips tails followed by heads (the sequence T H). What is the probability
that she will stop after flipping HH?

18. How many ordered pairs of integers (a, b) satisfy all of the following inequalities?

a2 + b2 < 16
a2 + b2 < 8a
a2 + b2 < 8b

19. A horse stands at the corner of a chessboard, a white square. With each jump, the horse can
move either two squares horizontally and one vertically or two vertically and one horizontally

2
 (like a knight moves). The horse earns two carrots every time it lands on a black square,
but it must pay a carrot in rent to rabbit who owns the chessboard for every move it makes.
When the horse reaches the square on which it began, it can leave. What is the maximum
number of carrots the horse can earn without touching any square more than twice?

20. Eight strangers are preparing to play bridge. How many ways can they be grouped into two
bridge games — that is, into unordered pairs of unordered pairs of people?

21. a and b are positive integers. When written in binary, a has 2004 1’s, and b has 2005 1’s
(not necessarily consecutive). What is the smallest number of 1’s a + b could possibly have?

22. Farmer John is grazing his cows at the origin. There is a river that runs east to west 50 feet
north of the origin. The barn is 100 feet to the south and 80 feet to the east of the origin.
Farmer John leads his cows to the river to take a swim, then the cows leave the river from
the same place they entered and Farmer John leads them to the barn. He does this using
the shortest path possible, and the total distance he travels is d feet. Find the value of d.

3
23. A freight train leaves the town of Jenkinsville at 1:00 PM traveling due east at constant
speed. Jim, a hobo, sneaks onto the train and falls asleep. At the same time, Julie leaves
Jenkinsville on her bicycle, traveling along a straight road in a northeasterly direction (but
not due northeast) at 10 miles per hour. At 1:12 PM, Jim rolls over in his sleep and falls
from the train onto the side of the tracks. He wakes up and immediately begins walking at
3.5 miles per hour directly towards the road on which Julie is riding. Jim reaches the road
at 2:12 PM, just as Julie is riding by. What is the speed of the train in miles per hour?

24. Given is a regular tetrahedron of volume 1. We obtain a second regular tetrahedron by

reflecting the given one through its center. What is the volume of their intersection?

25. A lattice point is a point whose coordinates are both integers. Suppose Johann walks in a
line from the point (0, 2004) to a random lattice point in the interior (not on the boundary)
of the square with vertices (0, 0), (0, 99), (99, 99), (99, 0). What is the probability that his
path, including the endpoints, contains an even number of lattice points?

4
 1

General Solutions
2004 Stanford Math Tournament
February 28, 2004
1. Answer: 24
Pretty straightforward. If a and b represent this and that respectively, a + b = 3a2 , and ab = 8. It follows that
3a = a2 , and a = 0 or 3. The former implies b = 0, contradicting ab = 8 since 00 is undefined. So a = 3, and
b = 24, which is the answer.
8
2. Answer: 23
The chances that either team wins are equal, so it’s best to examine the probability that neither team wins.
This occurs when there are three consecutive ties or one tie, one Rice victory and one Stanford victory. This
yields 7 cases. There would be 3 × 3 × 3 = 27 total possible cases, but a third game is not played if a team wins
the first two. So there are only 23 total cases. A team victory occurs in 23 − 7 = 16 of these cases, and Rice
8
wins in half of them. The answer is 23 .
3. Answer: 2
2n − 2 = n! Try the first few values.
(a) 2n − 2 = 0, 1! = 1
(b) 2n − 2 = 2, 2! = 2
(c) 2n − 2 = 4, 3! = 6
Now as we increase n, n! increases by much more than 2n − 2. So the only solution is 2.
4. Answer: (2, 1, 5, 2)
Since they’re all positive and since 3 > 37 1
1 is between
13 > 2, it’s a good idea to let a = 2. It follows that b + c+ d
1 and 2, so we should let b = 1. Likewise, one can confirm that c = 5 and d = 2. It is easy to see that this is
the only solution.
π
5. Answer: 4+3π
The area of the figure is the sum of the areas of the circles and that of the square less the overlap area. The
overlap area is simply four congruent quarters of a circle. It is easy to see that the answer is independent of the
π
radius of the circle. We get 4+3π .
6. Answer: 31
We can prove by induction on n that the following pattern holds for 0 ≤ n ≤ 499: after 2n minutes, the first
room contains 1000 − 2n people and the next n rooms each contain 2 people, and after 2n + 1 minutes, the first
room contains 1000 − (2n + 1) people, the next n rooms each contain 2 people, and the next room after that
contains 1 person. So, after 60 minutes, we have one room with 940 people and 30 rooms with 2 people each.
7. Answer: 9
Suppose the number n has k + 1 digits, the first of which is d. Then the number is at least d · 10k . On the
other hand, each of the digits after the first is at most 9, so the product of the digits is at most d · 9k . Thus, if
n equals the product of its digits, then
d · 10k ≤ n ≤ d · 9k
which forces k = 0, i.e., the number has only one digit. So n = 9 is clearly the largest possible value.
8. Answer: 2
Plug in x1 = x2 = x3 = x4 = x5 = 0. Then the equation reads f (0) = 5f (0) − 8, so 4f (0) = 8, so f (0) = 2.
9. Answer: 21600
In the top row, you can mark any of the 6 squares that is not a corner. In the bottom row, you can then mark
any of the 5 squares that is not a corner and not in the same column as the square just marked. Then, in the
second row, you have 6 choices for a square not in the same column as either of the two squares already marked;
then there are 5 choices remaining for the third row, and so on down to 1 for the seventh row, in which you
make the last mark. Thus, altogether, there are 6 · 5 · (6 · 5 · . . . · 1) = 30 · 6! = 30 · 720 = 21600 possible sets of
squares.
 2

10. Answer: 5
p √
If the sides are x and y, we have 2x + 2y = 10, so x + y = 5, and x2 + y 2 = 15, so x2 + y 2 = 15. Squaring
the first equation gives x2 + 2xy + y 2 = 25, and subtracting the second equation gives 2xy = 10, so the area is
xy = 5.
11. Answer: (7, 6, 4, 1)
Since D < A, when A is subtracted from D in the ones’ column, there will be a borrow from C in the tens’
column. Thus, D + 10 − A = C. Next, consider the subtraction in the tens’ column, (C − 1) − B. Since C < B,
there will be a borrow from the hundreds’ column, so (C −1+10)−B = A. In the hundreds’ column, B −1 ≥ C,
so we do not need to borrow from A in the thousands’ column. Thus, (B − 1) − C = D and A − D = B. We thus
have a system of four equations in four variables A, B, C, D, and solving by standard methods (e.g. substitution)
produces (A, B, C, D) = (7, 6, 4, 1).
12. Answer: 39/25

5
D B
K X
3

E A
L Y

Draw the altitude from C to AE, intersecting line BD at K and line AE at L. Then CK is the altitude of
CY CL
triangle BCD, so triangles CKX and CLY are similar. Since CX = 58 , CK = 85 . Also triangles CKB and CLA
CA 8 AE
are similar, so that CB = 5 , and triangles BCD and ACE are similar, so that BD = 58 . The area of ACE is
1 1 64
2 (AE)(CL), and the area of BCD is 2 (BD)(CK), so the ratio of the area of ACE to the area of BCD is 25 .
39
Therefore, the ratio of the area of ABDE to the area of BCD is 25 .
13. Answer: 10
The number in the ith row, jth column will receive the numbers 10(i − 1) + j and 10(j − 1) + i, so the question
is how many pairs (i, j) (1 ≤ i, j ≤ 10) will have
101 = [10(i − 1) + j] + [10(j − 1) + i] ⇔ 121 = 11i + 11j = 11(i + j).
Now it is clear that this is achieved by the ten pairs (1, 10), (2, 9), (3, 8), . . . , (10, 1) and no others.
14. Answer: 7/15
This is a case of conditional probability; the answer is the probability that the first ball is red and the second
ball is black, divided by the probability that the second ball is black.
First, we compute the numerator. If the first ball is drawn from Urn A, we have a probability of 26 of getting
a red ball, then a probability of 21 of drawing the second ball from Urn B, and a further probability of 63 of
drawing a black ball. If the first ball is drawn from Urn B, we have probability 63 of getting a red ball, then 21
of drawing the second ball from Urn B, and 35 of getting a black ball. So our numerator is
 
1 2 1 3 3 1 3 7
· · + · · = .
2 6 2 6 6 2 5 60

We similarly compute the denominator: if the first ball is drawn from Urn A, we have a probability of 21 of
drawing the second ball from Urn B, and 36 of drawing a black ball. If the first ball is drawn from Urn B, then
we have probability 36 that it is red, in which case the second ball will be black with probability 21 · 35 , and
probability 63 that the first ball is black, in which case the second is black with probability 21 · 25 . So overall, our
denominator is
  
1 1 3 3 1 3 1 2 1
· + · + · = .
2 2 6 6 2 5 2 5 4
 3
7
7
Thus, the desired conditional probability is 60
1 = 15 .
4

15. Answer: 8
Let L be the union of all the lines of the tiling. Imagine walking from one end of the needle to the other. We
enter a new triangle precisely when we cross one of the lines of the tiling. Therefore, the problem is equivalent
to maximizing the number of times the needle crosses L. Now, the lines of the tiling each run in one of three
directions. It is√clear that the needle cannot cross more than three lines in any given direction,
√
since the lines
3 3 3
are a distance 2 apart and the needle would therefore have to be of length greater than 2 > 2. Moreover,
it cannot cross three lines in each of two different directions. To see this, notice that its endpoints would have
to lie in either the two light-shaded regions or the two dark-shaded regions shown, but the closest two points
of such opposite regions are at a distance of 2 (twice the length of a side of a triangle), so the needle cannot
penetrate both regions.

Therefore, the needle can cross at most three lines in one direction and two lines in each of the other two
directions, making for a maximum of 3 + 2 + 2 = 7 crossings and 7 + 1 =√8 triangles intersected. The example
shows that 8 is achievable, as long as the needle has length greater than 3 < 2.

16. Answer: 6
For positive integers a, b, we have

a! | b! ⇔ a! ≤ b! ⇔ a ≤ b.

Thus,

((n!)!)! | (2004!)! ⇔ (n!)! ≤ 2004! ⇔ n! ≤ 2004 ⇔ n ≤ 6.

1
17. Answer: 4
The only way that Andrea can ever flip HH is if she never flips T , in which case she must flip two heads
immediately at the beginning. This happens with probability 14 .
18. Answer: 6
This is easiest to see by simply graphing the inequalities. They correspond to the (strict) interiors of circles of
radius 4 and centers at (0, 0), (4, 0), (0, 4), respectively. So we can see that there are 6 lattice points in their
intersection (circled in the figure).
 4

(0,4)

(0,0) (4,0)

19. Answer: 0
The horse must alternate white and black squares, and it ends on the same square where it started. Thus it
lands on the same number of black squares (b) as white squares (w). Thus, its net earnings will be 2b − (b + w) =
b − w = 0 carrots, regardless of its path.

20. Answer: 315

Putting 8 people into 4 pairs and putting those 4 pairs into 2 pairs of pairs are independent. If the people are
numbered from 1 to 8, there are 7 ways to choose the person to pair with person 1. Then there are 5 ways to
choose the person to pair with the person who has the lowest remaining number, 3 ways to choose the next, and
1 way to choose the last (because there are only 2 people remaining). Thus, there are 7 · 5 · 3 · 1 ways to assign 8
people to pairs and similarly there are 3 · 1 ways to assign 4 pairs to 2 pairs of pairs, so there are 7 · 5 · 3 · 3 = 315
ways.
21. Answer: 1
Consider the following addition:

111 . . . 100 . . . 01
+ 11 . . . 11
= 1000 . . . . . . . . . 00

By making the blocks of 1’s and 0’s appropriately long, we can ensure that the addends respectively contain
2004 and 2005 1’s. (To be precise, we get a = 24008 − 22005 + 1 and b = 22005 − 1.) Then the sum has only one
1. And clearly it is not possible to get any less than one 1.
 5
√
22. Answer: 40 29
Suppose we move the barn to its reflection across the river’s edge. Then paths from the origin to the river and
then to the old barn location correspond to paths from the origin to the river and then to the new barn location,
by reflecting the second part of the path across the river, and corresponding paths have the same length. Now
the shortest path from the origin to the river and then to the new barn location
√ is a straight line.
√ The new barn
location is 200 feet north and 80 feet east of the origin, so the value of d is 2002 + 802 = 40 29.

200 reflection of barn

shortest path

river

80

23. Answer: 62.5or 125

2
Julie’s distance is (10 mph) · ( 65 hrs) = 12 miles. Jim’s walking distance, after falling off the train, is (3.5 mph) ·
√ √
(1 hr) = 3.5 miles at a right angle to the road. Therefore, Jim rode the train for 122 + 3.52 = 21 242 + 72 = 25 2
miles, and its speed is ( 25
2 mi)/( 1
5 hr) = 62.5 mph.
1
24. Answer: 2
Imagine placing the tetrahedron ABCD flat on a table with vertex A at the top. By vectors or otherwise, we see
that the center is 34 of the way from A to the bottom face, so the reflection of this face lies in a horizontal plane
halfway between A and BCD. In particular, it cuts off the smaller tetrahedron obtained by scaling the original
tetrahedron by a factor of 12 about A. Similarly, the reflections of the other three faces cut off tetrahedra
obtained by scaling ABCD by 21 about B, C, and D. On the other hand, the octahedral piece remaining
remaining after we remove these four smaller tetrahedra is in the intersection of ABCD with its reflection, since
the reflection sends this piece to itself. So the answer we seek is just the volume of this piece, which is

(volume of ABCD) − 4 · (volume of ABCD scaled by a factor of 1/2) = 1 − 4(1/2)3 = 1/2.

3
25. Answer: 4
If Johann picks the point (a, b), the path will contain gcd(a, 2004 − b) + 1 points. There will be an odd number
of points in the path if gcd(a, 2004 − b) is even, which is true if and only if a and b are both even. Since there
are 492 points with a, b both even and 982 total points, the probability that the path contains an even number
of points is

982 − 492 492 (22 − 12 ) 3

2
= = .
98 492 (22 ) 4
 Team Test
2004 Stanford Mathematics Tournament
February 28, 2004
2002 5π 9π
1. Find sin x − cos x if sin 2x = 2003 and 4 <x< 4 .

2. Dave and Daly decide to play a game using two dice. Dave will roll first, then they will continue
alternating turns. If Dave gets a total of exactly 6 before Daly gets a total of exactly 7, then Dave
wins. Otherwise Daly wins. What is the probability that Dave wins?
3. Suppose ABCDEF is a regular hexagon with area 1, and consider the diagonals AC, BD, CE, DF ,
EA and F B. A star is formed by alternating between the vertices of the hexagon and the points of
intersection of the diagonals. What is the area of the star?

4. A pair of positive integers is golden if they end in the same two digits. For example (139, 2739) and
(350, 850) are golden pairs. What is the sum of all two-digit integers n for which (n2 , n3 ) is golden?
5. A polynomial is monic if the leading coefficient is 1. Suppose p(x) is a cubic monic polynomial. Let a,
b, and c be the roots of p(x). If a + b + c = 1 and a2 + b2 + c2 = 5 and a3 + b3 + c3 = 16, determine
p(x).

6. Thirteen students, four of whom are named Bob, are going out to a math conference. They have three
distinct cars, two of which will hold 4 students and the other will hold 5. If the students randomly
scramble into the cars, what is the probability that every car has at least one Bob in it?
P∞
7. Find all x such that k=1 kxk = 20.
8. Let 4ABC be an equilateral triangle with side length 2004. For any point P inside 4ABC, let d(P )
be the sum of the distances of P to each of the sides of 4ABC. Let m be the minimum value of d(P ).
Let M be the maximum value of d(P ). Compute M − m.
9. There are 30 owls in a row starting from left to right and Sammy (their coach) is facing them. They
have a warmup routine involving trading places in rounds. Each round involves three phases. In the
first phase of each round, the owls in the odd positions rotate to the right while the owls in even
positions do not move. For example, the first owl moves to the third owl’s spot who is moving to
the fifth owl’s spot, etc. Note the 29th owl will wrap around and move to the first owl’s spot. The
second phase occurs after the first is completed and in this phase the owls in positions which leave a
remainder of 1 when divided by 3 rotate to the right. All other owls stay fixed. For example, the owl
in position 1 would move to position 4. In the last phase of each round only owls in positions that
leave a remainder of 1 when divided by 5 move. These owls however move left! The next round then
begins back at phase 1. The warmup continues until the owls are back in their original order. How
many rounds does this take?
10. Let bxc denote the floor function (the largest integer less than or equal to x). Find the average value
of the quantity b2x3 − 2bx3 cc on the interval (− 32 , 32 ).
11. Suppose a king has 25! grains of rice after collecting taxes. He is feeling generous so he instructs his
accountant to divide the rice into two piles, the first of which he will keep and the second of which he
will distribute to the poor. The accountant places the nth grain of rice in the first pile if n and 25! are
not relatively prime and the second pile otherwise. What fraction of the grains get distributed to the
poor?
12. Given that x2 − 3x + 1 = 0, find x9 + x7 + x−9 + x−7 .
13. How many ordered triples of integers (b, c, d) are there such that x4 − 5x3 + bx2 + cx + d has four (not
necessarily distinct) non-negative rational roots?

1
 2004
14. Find the length of the leading non-repeating block in the decimal expansion of 7∗52003 . (For example,
5
the length of the leading non-repeating block of 12 = .416666666 is 2).
15. Suppose we play a variant of Yahtzee where we roll 5 dice, trying to get all sixes. After each roll we
keep the dice that are sixes and re-roll all the others. What is the probability of getting all sixes in
exactly k rolls as a function of k?

2
 Team Test Solutions
2004 Stanford Mathematics Tournament
February 28, 2004
1. Answer: √−1
2003

Since (sin x − cos x)2 = sin2 x + cos2 x − 2 sin x cos x = 1 − sin 2x = 2003
1
, we are looking for when
(sin x − cos x) = 2003 . For the interval given, sin x < cos x thus our answer is √−1
2 1
2003
.
30
2. Answer: 61
5
There is a chance that Dave’s roll will be a 6 and a 16 chance that Daly’s roll will be a 7. So we can
36
5 31 5 5 31 5 31 5 5
write the probability that Dave wins as: 36 + 36 6 36 + 36 6 36 6 36 + .... So, this is a geometric series
5
30
whose sum is 36
1− 56 31 = 61 .
36

2
3. Answer: 3
Notice that the figure is a hexagon in union with six congruent triangles. It’s easy to see that triangles
ACE and BDF are equilateral, so each of these six congruent triangles has a 60◦ angle opposite the
side it shares with the smaller hexagon. The other two sides are obviously equivalent, so the triangle is
equilateral. ”Folding” these six triangles across their common side with the hexagon shows that their
area of their union is the same as that of the hexagon. If V is the intersection of AC and BF , then
AV B is 120◦ and BV = AV = JV where J is the intersection of AE and BF . This proves that F J,
JV and V B are all equivalent, so JV is 31 the length of BF . If s is the side length of ABCDEF , then
√
it’s easy to see that BF is s 3, so the smaller hexagon has length √s3 . The ratio of the area of the
smaller hexagon to that of the larger is thus 13 , and the answer to the question is 32 .
4. Answer: 772
Suppose n is a two-digit integer and that s is its square. Writing n as (10a + b) where a and b are
digits, we need the rightmost two digits of (20ab + b2 ) and (30ab2 + b2 ) to be the same. We have four
cases: b = 0, 1, 5, 6. The first case gives all two-digit multiples of ten. The second case yields nothing.
The third case yields a ∈ {2, 4, 6, 8}. The final case gives a ∈ {2, 7}. The answer is 450 + 25 + 45 +
65 + 85 + 26 + 76 = 772.
5. Answer: p(x) = x3 − x2 − 2x − 3
The coefficients of p(x) = x3 − lx2 + mx − n are l = a + b + c m = ab + ac + bc n = abc l = 1 is given.
l2 = (a + b + c)2 = a2 + b2 + c2 + 2ab + 2ac + 2bc = a2 + b2 + c2 + 2m. Since l2 = 1 and a2 + b2 + c2 = 5,
m = −2. l3 = (a + b + c)3 = a3 + b3 + c3 + 3a2 b + 3ab2 + 3a2 c + 3ac2 + 3b2 c + 3bc2 + 6abc. Note that
lm = a2 b + ab2 + a2 c + ac2 + b2 c + bc2 + 3abc, so l3 = a3 + bc + c3 + 3lm − 3n. We know the values of
everything except n, so we can solve to show n = 3.
80
6. Answer: 143
The students can be grouped into the cars in a total of 13

9
5
2
4 4 5 = 13 · 11 · 7 · 5 · 3 · 2 ways. We
must arrange the Bobs so that two are seated in one car and the other two are in the other cars. If
the car holding the two Bobs is the 5-seater, then there are 93 42 63 21 33 11 = 7 · 5 · 32 · 26 ways.

This is the number of ways of choosing the non-Bob people in the 5-seater times the number of ways of
choosing the two Bobs times the number of ways of choosing non-Bob people in one of the four seaters
times the number of ways of choosing Bob times the number of ways of choosing non-Bob people in
the other four seater times
the


number


of
ways of choosing the last Bob. If the two Bobs double up
in a 4-seater, there are 92 42 73 21 44 11 = 7 · 5 · 33 · 24 possible groupings. There are two different
7·52 ·32 ·25 80
4-seaters. Adding all these up yields 7 · 52 · 32 · 25 . Thus the probability is 13·11·7·5·3 2 ·2 = 143 .

4
7. Answer: 5

1
 1
Use the identity for geometric series: 1−a = 1+a+a2 +... Notice that (a+a2 +a3 +...)(1+a+a2 +...) =
1 1
P∞ 1
a + 2a2 + 3a3 + ... Thus ( 1−a − 1)( 1−a ) = k
k=1 kx = 20. Set y = 1−a and we want to solve
(y − 1)(y) = 20 The two solutions are 5 and −4. Then a = 45 or 54 . The second solution gives a series
which diverges, hence the only answer is 54 .
8. Answer: 0
Fix P a set distance from BC. Draw the line through P which is parallel to BC. It intersects AB
and BC at B 0 and C 0 respectively.
√
The perpendiculars
√
from P to AB 0 and AC 0 are P X and P Y
respectively. Now P B = 23 (P X) and P C 0 = 23 (P Y ), so P X + P Y = √23 (P B 0 + P C 0 ) which is
constant. So, d(P ) does not change when we slide the point in a direction parallel to a side. We can
use this tranlstion twice to get to any other point, so we know that d(P ) is a constant.
9. Answer: 60
Writing out what happens in the round one will show us the pattern. Owls in positions 2, 8, 11, 12,
14, 18, 20, 24 and 30 stay in the same spot each round. The owls in sports 16 and 19 just trade spots.
Thus every even numbered round they are back in the right places. The four owls in spots 21, 26, 28
and 23 just rotate, in that order. Thus every fourth round they are back in the original order. Lastly,
the remaining 15 owls form one huge rotation. The order is 1, 6, 9, 7, 4, 29, 27, 25, 22, 17, 15, 13, 10,
5, 3, and 1 again. Thus every 15 rounds these owls return to their original order. One can see each
of these three groups that rotate will be back in their original order at the gcd of 2, 4 and 15. The
answer is 60.
1
10. Answer: 2
Recall that (1) b−xc = −dxe (2) If x is not an integer dxe = bxc + 1. (3) bx + nc = bxc + n for integer
n. Let F (x) = b2x3 − 2bx3 cc. Then as long as x3 is not an integer or half-integer we have:
F (−x) = b−2x3 − 2b−x3 cc = b−2x3 + 2dx3 ec = b−2x3 + 2bx3 c + 2c = 2 + b−2x3 + 2bx3 cc =
2 − d2x3 − 2bx3 ce = 1 − b2x3 − 2bx3 cc = 1 − F (x).
Thus F (x) + F (−x) = 1 except at isolated points. Hence the average is 21 .
110592
11. Answer: 676039
The people lose 12
of the grains on the first step. Then 13 of the remaining numbers are multiples of 3,
212 ∗33
and so on. So we just need to compute (1 − 21 )(1 − 13 )(1 − 15 )...(1 − 23
1
) = 7∗13∗17∗19∗23 = 110592
676039 .

12. Answer: 6621

The equation implies that x + x1 = 3. The expression to be solved can be factored as (x8 + x−8 )(x + x1 ).
The second term is 3 while the first one can be found by squaring x + x1 , solving for x2 + x−2 and
repeating the procedure twice. The first term turns out to be 2207. The answer is 6621.
13. Answer: 6
Because -5 is the second coefficient, the sum of the roots is 5. 5 can be written as the sum of 4
nonnegative integers 6 distinct ways: (0,0,0,5), (0,0,1,4), (0,0,2,3), (0,1,1,3), (0,1,2,2), (1,1,1,2) Since
5 is the sum of the roots, and the roots must be integers by the rational roots theorem, these are
the only roots the polynomial can have. But since the polynomial is monic, fixing the roots fixes the
polynomial.
14. Answer: 2003 Notice that this can be written with partial fractions as A7 + 52003
B
. A7 is a repeating
B
decimal while 52003 is a string of 2003 digits which do not repeat. So the length of the leading non-
repeating block is 2003.
15. Answer: (1 − ( 65 )k )5 − (1 − ( 56 )k−1 )5
Put all 5 dice in independent cups. Roll all of them k times. Then the probability that one of them
is 6 somewhere in k rolls is 1 − ( 65 )k . Then the probability that all 5 of them have a 6 somewhere
in k rolls is (1 − ( 56 )k )5 . To get the probability of getting all sixes in exactly k rolls, we subtract the
probability that all 5 dice have a 6 in k − 1 rolls. This gives (1 − ( 56 )k )5 − (1 − ( 56 )k−1 )5 .

2
 Algebra Test
2006 Stanford Math Tournament
February 25, 2006
1. A finite sequence of positive integers mi for i = 1, 2, . . . , 2006 are defined so that m1 = 1 and mi =
10mi−1 + 1 for i > 1. How many of these integers are divisible by 37?
2. Find the minimum value of 2x2 + 2y 2 + 5z 2 − 2xy − 4yz − 4x − 2z + 15 for real numbers x, y, z.

3. A Gaussian prime is a Gaussian integer z = a + bi (where a and b are integers) with no Gaussian
integer factors of smaller absolute value. Factor −4 + 7i into Gaussian primes with positive real parts.
i is a symbol with the property that i2 = −1.
a3 b3 c3
4. Simplify: (a−b)(a−c) + (b−a)(b−c) + (c−a)(c−b)

5. Jerry is bored one day, so he makes an array of Cocoa pebbles. He makes 8 equal rows with the pebbles
remaining in a box. When Kramer drops by and eats one, Jerry yells at him until Kramer realizes he
can make 9 equal rows with the remaining pebbles. After Kramer eats another, he finds he can make
10 equal rows with the remaining pebbles. Find the smallest number of pebbles that were in the box
in the beginning.
6. Let a, b, c be real numbers satisfying:

ab − a = b + 119
bc − b = c + 59
ca − c = a + 71

Determine all possible values of a + b + c.

7. Find all solutions to aabb = n4 − 6n3 , where a and b are non-zero digits, and n is an integer (a and b
are not necessarily distinct).
8. Evaluate:
10
X 2
x=2
x(x2 − 1)
.
9. Principal Skinner is thinking of two integers m and n and bets Superintendent Chalmers that he will
not be able to determine these integers with a single piece of information. Chalmers asks Skinner the
numerical value of mn + 13m + 13n − m2 − n2 . From the value of this expression alone, he miraculously
determines both m and n. What is the value of the above expression?
∞
k
P
10. Evaluate: ak−1
for |a| < 1.
k=1

1
 Algebra Solutions
2006 Stanford Math Tournament
February 25, 2006
1. Answer: 668
Note that 111 = 3 · 37. It follows that mi is divisible by 37 for all i = 3, 6, 9, . . . , 2004. The others will
clearly leave remainders of 1 or 11.
2. Answer: 10
The expression can be written as (x − 2)2 + (x − y)2 + (y − 2z)2 + (z − 1)2 + 10. This clearly must be
at least 10. Indeed, if x = 2, y = 2, z = 1, this value is achieved.
3. Answer: (1 + 2i)(2 + 3i)
We write −4 + 7i = (a + bi)(c + di). The solution can be intuitive after the first line of expansion, in
the same way as factoring of polynomials. However, we can assume a = 1 and then move factors from
(c + di) back to (a + bi) if we don’t end up with integers (fortunately, in this case we’re lucky).

−4 + 7i = (a + bi)(c + di)
= ac − bd + (ad + bc)i
= c − bd + (d + bc)i

We then know c should be positive (and not too large), so we can try c = 1, giving 1 − bd = −4 and
b + d=7, which clearly has no rational solution. We then try c = 2, giving 6 = bd and 2b + d = 7, which
is easily solved giving the final solution.
4. Answer: a + b + c

a3 b3 c3 a3 (c − b) + b3 (a − c) + c3 (b − a)
+ + =
(a − b)(a − c) (b − a)(b − c) (c − a)(c − b) (a − b)(b − c)(c − a)
a (c − b) + a(b3 − c3 ) + bc3 − cb3
3
=
(a − b)(b − c)(c − a)
(c − b)(a3 − a(b2 + bc + c2 ) + b2 c + c2 b
=
(a − b)(b − c)(c − a)
b (c − a) + b(c2 − ac) + a3 − ac2
2
=−
(a − b)(c − a)
(c − a)(b + bc − ac − a2 )
2
=−
(a − b)(c − a)
(a + b)(a − b) + c(a − b)
=
a−b
=a+b+c

5. Answer: 352
Let N represent the number of remaining pebbles after Kramer eats the second. Then N is divisible
by 10, and N + 1, which must end in 1, is divisible by 9. Put N + 1 = 100a + 10b + 1, where a and b
are digits summing to 8 or 17 (so the sum of the digits will be divisible by 9 - hence the number will
be divisible by 9). Now we need N + 2 to be divisible by 8. Try 82, 172, 262, and 352 to get 352 as
the answer.

1
 6. Answer: 31, −25
From the first equation:
ab − a = b + 119
a(b − 1) = (b − 1) + 120
(a − 1)(b − 1) = 120

Similarly, (b − 1)(c − 1) = 60 and (a − 1)(c − 1) = 72. Therefore a−1 2

c−1 = 2, and so 2(c − 1) = 72. This
gives c = 7, and then it is easy to find a = 13 and b = 11. The other solution is c = −5, so a = −11,
and b = −9. The sums are 31 and -25.
7. Answer: a = 6, b = 5
Since 11|aabb, aabb = 11 · a0b. Factor n4 − 6n3 = (n − 6)n3 , so clearly n > 6, as aabb > 0. Also,
a0b < 1000, so unless n = 11, n < 10. Trying n = 7, 8, 9 yields no solutions, so n = 11 must be the
only solution, if it exists. Indeed we get 6655 = (11 − 6) · 113 .
27
8. Answer: 55 
2 1 1 1 1 1
x(x2 −1) = x x−1 − x+1 = x(x−1) − x(x+1)
1
Let f (x) = x(x−1) . Then:
P10 2
P10 P10 P11 1 1
x=2 x(x2 −1) = x=2 (f (x) − f (x + 1)) = x=2 f (x) − x=3 f (x) = f (2) − f (11) = 2·1 − 11·10 =
1 1 27
2 − 110 = 55

9. Answer: 169
Let A be the value of the expression. We have: m2 + n2 − 13m − 13n − mn + A = 0. Multiplying by
2 yields:
m2 − 2mn + n2 + m2 − 26m + n2 − 26n + 2A = 0
(m − n)2 + (m − 13)2 + (n − 13)2 = 2 · 13 · 2 − 2A
In order for there to be a single solution, the sum of the squares must equal zero, yielding A = 169.
If instead the sum is a positive integer with a solution (m, n), then (n, m) will provide an additional
solution unless m = n. In that case, (26−m, 26−n) is an additional solution. Hence, it is both sufficient
and necessary that the sum of the squares equal zero in order that the solution be unambiguous.
 2
a
10. Answer: a−1

∞
X k 1 2 3 4
= + + 2 + 3 + ···
nk−1 1 k k k
k=1
     
1 1 1 1 1 1 1 1 1
= + + + 3 + ··· + + + 3 + ··· + + 3 + ··· + ···
1 k k2 k k k2 k k2 k
     
1 1 1 1 1 1 1 1 1 1 1
= + + 2 + 3 + ··· + + + 2 + ··· + 2 + + ··· + ···
1 k k k k 1 k k k 1 k
  
1 1 1 1 1 1 1 1
= + + + 3 + ··· + + + 3 + ···
1 k k2 k 1 k k2 k
 2
1
=
1 − 1/k
 2
k
=
k−1

2
 Geometry Test
2006 Stanford Math Tournament
February 25, 2006
1. Given a cube, determine the ratio of the volume of the octahedron formed by connecting the centers
of each face of the cube to the volume of the cube.
2. Given square ABCD of side length 1, with E on CD and F in the interior of the square so that
EF ⊥ DC and AF ∼= BF ∼= EF , find the area of the quadrilateral ADEF .
3. Circle γ is centered at (0, 3) with radius 1. Circle δ is externally tangent to circle γ and tangent to the
x-axis. Find an equation, solved for y if possible, for the locus of possible centers (x, y) of circle δ.
4. The distance AB is l. Find the area of the locus of points X such that 15o ≤ 6 AXB ≤ 30o and X is
on the same side of line AB as a given point C.
5. Let S denote a set of points (x, y, z). We define the shadow of S to be the set of points (x, y) for which
there exists a real number z such that (x, y, z) is in S. For example, the shadow of a sphere with radius
r centered on the z axis is a circle in the xy plane centered at√the origin with radius r. Suppose a cube
has a shadow consisting of a regular hexagon witih area 147 3. What is the side length of the cube?
6. A circle of radius R is placed tangent to two perpendicular lines. Another circle is placed tangent to
the same two lines and the first circle. In terms of R, what is the radius of a third circle that is tangent
to one line and tangent to both other circles?
7. A certain 2’ by 1’ pool table has pockets, denoted [A, . . . , F ] as shown. A pool player strikes a ball
at point x, 14 of the way up side AC, aiming for a point 1.6’ up the opposite side of the table. He
makes his mark, and the ball ricochets around the edges of the table until it finally lands in one of the
pockets. How many times does it ricochet before it falls into a pocket, and which pocket? Write your
answer in the form {C, 2006}.

2
8. In triangle 4P QR, the altitudes from P, Q and R measure 5, 4 and 4, respectively. Find QR .
9. Poles A, B, and P1 , P2 , P3 ,. . . are vertical line segments with bases on the x-axis. The tops of poles A
and B are (0,1) and (200,5), respectively. A string S connects (0,1) and (200,0) and intersects another
string connecting (0,0) and (200,5) at point T. Pole P1 is constructed with T as its top point. For each
integer i > 1, pole Pi is constructed so that its top point is the intersection of S and the line segment
connecting the base of Pi−1 (on the x-axis) and the top of pole B. Find the height of pole P100 .
BQ
10. In triangle 4ABC, points P, Q and R lie on sides AB, BC and AC, respectively, so that PAP
B = QC =
CR 1
RA = 3 . If the area of 4ABC is 1, determine the area of the triangle formed by the points of
intersection of lines AQ, BR and CP .

1
 Geometry Solutions
2006 Stanford Math Tournament
February 25, 2006
1. Answer: 16
Let s represent the side length of the cube. The octahedron has a volume
√ equivalent to the volume
of two pyramid with height 2s and a square base with side length 2s 2. The volume is therefore
√
2
2 · 31 · 2s 2 · 2s = 61 · s3 , or 1/6 of the cube volume.
2. Answer: 13
32
Let EF = x.

From pythagorean theorem:

( 12 )2 + (1 − x)2 = x2
1 + 4x2 − 8x + 4 = 4x2
8x = 5
x = 85
1 ( 12 )( 83 ) 1 3 13
area of ADEF = area of ADEG− area of AF G = 2 − 2 = 2 − 32 = 32
2
3. Answer: y = x8 + 1
Since circle δ is tangent to the x-axis, its radius is y. Thus from the Pythagorean Theorem:
(3 − y)2 + x2 = (y + 1)2
9 − 6y + x2 = 2y + 1
8 + x2 = 8y
x2
1+ =y
8
 √ 
4. Answer: 12 + π 1 + 1112 3
Since inscribed angles intercept arcs of measure twice that of the inscribed angle, this is the area above
line AB between circles centered at P and Q, with 6 AQB = 60◦ and 6 AP B = 30◦ , A, B on both
circles, and P, Q on the√ perpendicular bisector of AB. Let M be the midpoint of AB. 4AQB is then
equilateral, so QM = 23 , so the radius of circle Q is 1. We see that since 6 AP B = 30◦ , P is on circle
√  √ 2
2 √
Q, so P M = 1 + 23 , and by the Pythagorean theorem, (P A)2 = 1 + 23 + 12 = 2 + 3. We find
the area in circles P and Q above line AB by taking the major sector AB of each the circles above the
line and adding in the areas of 4AP B and 4AQB respectively:
√ ! √ !!
 √  330 1 3 2 300 1 3
A=π 2+ 3 + (1) 1 + − π·1 + (1)
360 2 2 360 2 2

1
 This simplifies to the given answer.
√
5. Answer: 7 3
Let the side length of the cube be s. It is apparent that in order for the shadow to be a regular hexagon,
the cube must have two vertices with the same x and y coordinates; call these vertices A and √ B. Let T
be another vertex of the cube. Clearly, √ 4ABT is a right triangle with hypotenuse AB = s 3 the space
diagonal of the cube, and legs s and s 2. Notice that a segment from T to AB has for its shadow a
segment between the center of the hexagon and one of its vertices; thus the distance from T to AB√ is
the same as the center to vertex distance. Using similar triangles, this length can be found to be s 3 6 .
√ √ √
Thus the area of the hexagon is s2 3 = 147 3 and therefore s = 7 3.
√
6. Answer: R 2
(3 − 2 2)

Let
√ the radius of √
the second circle be r.
R 2 −√R = r + r 2
√
r = R(√2+1
2−1)
= R(3 − 2 2)
Let
p the radius of the third
p circle be ρ. p
(r + ρ)2 − (r −√ρ)2 + (R + ρ)2 − (R − ρ)2 = (R + r)2 − (R − r)2
√ √
4rρ +√ 4Rρ = √ 4Rr
√
rρ + Rρ = Rr
√
√ Rr
ρ = √R+ √
r √ √ √ √
Rr√ R2 (3−2 2) R2 (3−2 2) R2 (3−2 2) R
ρ= = √ √ √ = √ √ = 2R = 2 (3 − 2 2)
R+r+2 Rr R(4−2 2)+2R 3−2 2 R(4−2 2)+2R( 2−1)

7. Answer: {D, 6}

Note that each time the ball bounces up the wall, it is equivalent to forming a mirror image of the table
and extending the path taken. Set up sides AF and AC as the x and y coordinate axes, respectively.

2
 Since the ball is hit at (0,0.5), it can travel in an imaginary straight line through imaginary images of
the table until it hits an integer coordinate (i.e. a pocket). Therefore,
11
0.5 + (1.6 − 1.5)x = y1 + · x = 2y
5
It is clear that the first instance of integer (x, y) occurs when x = 5 and y = 6. Simply counting, 5 units
in the x direction ends up on side DF , and 6 units in the y direction would be on side CD. Therefore,
the ball must have fallen in at this intersection, into pocket D. Drawing iterations of the pool table to
fill the rectangle from (0, 0) to (5, 6), we see that the ball has crossed four vertical boundaries and two
horizontal boundaries, making 6 ricochets.
8. Answer: 400 21
A
If M is the midpoint of QR, then P M · QR = 2A, where A is the area of the triangle. So QM = 5 and,
by the same logic, P Q = A2 . Use the Pythagorean Theorem on triangle 4P QM to get A = √50 ⇒
21
QR = √20 .
21

1
9. Answer: 21
Arbitrarily label the heights of poles A and B as a and b, respectively. Suppose poles A and B are p
and q units, respectively, from pole P1 (as measured along the x-axis). Then the height of P1 , call it x,
q p
satisfies: xa = p+q and xb = p+q ⇒ x = a+bab
. The same procedure yields the height of P2 : just replace
ab ab ab ab
a by a+b in the above equation to get 2a+b . Generalize by replacing a by na+b to get (n+1)a+b as the
1
height of Pn+1 . Now put a = 1, b = 5 and n = 100 to get 21 .
4
10. Answer: 13
Let O be the intersection of AQ and BR. Our goal is to find the area of 4ABO = 1 · BQ AO 1 AO
BC · AQ = 4 · AQ .
Using mass points, place a mass of 1 at B and therefore a mass of 31 at C since 1BQ = 13 QC. Likewise,
vertex A bears a mass of 19 . Replace the masses at B and C with a mass of 1 + 13 = 34 at Q. Thus,
AO 4/3 AO 12 1 12 3
OQ = 1/9 = 12. Hence, AQ = 13 . Thus, the area of 4ABO = 4 · 13 = 13 , which is independent of
the side lengths of 4ABC. There are two additional nonoverlapping triangles like 4ABO that must
3 3 4
also have an area of 13 . The area of the central triangle is 1 − 3 · 13 = 13 .

3
 Calculus Test
2006 Stanford Math Tournament
February 25, 2006
1. Evaluate:
d sin x
dx x
lim
x→0 x
2. Given the equation 4y 00 + 3y 0 − y = 0 and its solution y = eλt , what are the values of λ?
3. Find the volume of an hourglass constructed by revolving the graph of y = sin2 (x) + 10
1
from − π2 to π
2
about the x-axis.
4. Evaluate
ln(x + 1)
lim 1
x→0 x · ((1 + x) 2 − e)

(x tan−1 x)dx
R
5. Evaluate:
6. Evaluate
π/2
sin3 x
Z
dx
0 sin3 x + cos3 x
.
7. Find Hn+1 (x) in terms of Hn (x), Hn0 (x), Hn00 (x), . . . for

2 dn  −x2 
Hn (x) = (−1)n ex e
dxn

8. A unicorn is tied to a cylindrical wizard’s magic tower with an elven rope stretching from the unicorn
to the top of the tower. The tower has radius 2 and height 8; the rope is of length 10. The unicorn
begins as far away from the center of the tower as possible. The unicorn is startled and begins to run
as close to counterclockwise as possible; as it does so the rope winds around the tower. Find the area
swept out by the shadow of the rope, assuming the sun is directly overhead. Also, you may assume
that the unicorn is a point on the ground, and that the elven rope is so light it makes a straight line
from the unicorn to the tower.
ex −e−x
9. Define the function tanh x = ex +e−x Let tanh−1 denote the inverse function of tanh. Evaluate and
simplify:
d
tanh−1 tan x
dx
10. Four ants Alan, Bill, Carl, and Diane begin at the points (0, 0), (1, 0), (1, 1), and (0, 1), respectively.
Beginning at the same time they begin to walk at constant speed so that Alan is always moving
directly toward Bill, Bill toward Carl, Carl toward Diane, and Diane toward Alan. An approximate
solution finds that after some time, Alan is at the point (0.6, 0.4). Assuming for the moment that
this approximation is correct (it is, to better than 1%) and so the pont lies on Alan’s path, what is
the radius of curvature at that point. In standard Cartesian coordinates, the radius of curvature of a
function y(x) is given by:
  2 3/2
dy
1 + dx
R=  
 d2 y 
 dx2 

1
 Calculus Solutions
2006 Stanford Math Tournament
February 25, 2006
1. Answer: − 13
d sin x
dx x x cos x − sin x
lim = lim
x→0 x x→0 x3
cos x − x sin x − cos x
= lim
x→0 3x2
sin x cos x 1
= lim − = lim − =−
x→0 3x x→0 3 3
2. Answer: 14 , −1
You substitute the solution of y = eλt into the differential equation. You then get 4λ2 eλt +3λeλt −eλt =
0 where you can divide through by eλt and end up with 4λ2 + 3λ − λ = 0. You get (4λ − 1)(λ + 1) = 0
where λ must be equal to 41 , −1.
97π 2
3. Answer: 200
The volume is
Z π   Z π  
2
2 1 2
41 2 1
π sin x + dx = π sin x + sin x + dx
−π
2
10 −π
2
5 100
Z π2 
π2

1
= +π sin2 x(1 − cos2 x) + sin2 x dx
100 −π2
5
π 
π2
Z 
2 6 1
= +π sin2 x − sin2 2x dx
100 −π 5 4
 2
π2 97π 2

6π 1π
= +π − =
100 52 42 200
1
4. Answer: e
1
ln(x+1) x 1 ln(y)
Rearrange to get limx→0 1 . Let y = ln(x + 1) x . Using L’Hopital’s rule, limx→0 y−e =
(1+x) 2 −e
y0
limx→0 y
y0 = limx→0 y1 . To evaluate the limit in the denominator,
 1 1
1 ln (1+x) x ln(1+x) 1+x
lim (1 + x) x = lim e = elimx→0 x = elimx→0 1 =e
x→0 x→0
1
so the answer is e
1+x2
5. Answer: − x2 + 2
tan−1 x

Z Z
x 1
(x tan−1 x)dx = − 1 + 2x tan−1 x dx


+
2 2
1 + x2
Z  
x 1 −1
=− + + 2x tan x dx
2 2 1 + x2
Z  
x 1 d d
=− + (1 + x2 ) (tan−1 x) + (1 + x2 ) tan−1 x dx
2 2 dx dx
2
x 1+x
=− + tan−1 x
2 2

1
 Integration by parts can also lead to the solution.
6. Answer: π4
Let u = π2 − x. Substituting then changing u to x gives
π/2 π/2
sin3 x cos3 x
Z Z
3 dx = dx
0 sin x + cos3 x 0 cos3 x + sin3 x

Adding the two integrals

π/2 π/2
sin3 x
Z Z
2· 3 dx = dx
0 sin x + cos3 x 0

so
π/2
sin3 x
Z
π 1 π
3 dx = · =
0 sin x + cos3 x 2 2 4

7. Answer: 2xHn (x) − Hn0 (x)

2 d  −x2 
Hn+1 (x) = −ex e Hn (x)
dx
 
2 2 2
= −ex −2xe−x Hn (x) + e−x Hn0 (x)
= 2xHn (x) − Hn0 (x)

8. Answer: 18 + 9π
The shadow of the rope has length 6 by the Pythagorean theorem; we can work the rest of the problem
pretending we have a horizontal rope of length 6 from the unicorn to the tower. The unicorn travels a
quarter-circle before the rope begins to wrap around the tower; from then on, if the rope has wrapped
around an angle θ of the tower, the rope remaining is 6 − 2θ long. Using the Pythagorean theorem,
we find that if r is the unicorn’s distance from the center of the tower, r2 = 22 + (6 − 2θ)2 . The area
swept out is the initial quarter-circle travelled by the unicorn plus the area swept out by the line from
the unicorn to the center of the tower minus the area of the tower covered in the angle traversed:
6/2
π62
Z
1 6/2
+ (22 + (6 − 2θ)2 )dθ − π22
4 2 0 2π

9. Answer: sec 2x
First we solve for the inverse function; let x = tanh−1 y.

ex − e−x
y = tanh x =
ex + e−x
e2x − 1
y = 2x
e +1
e + 1 y = e2x − 1
2x

e2x (y − 1) = y + 1
1 y+1
x = ln
2 y−1

2
 Now we return to the given expression:
d d 1 tan x + 1
tanh−1 tan x = ln
dx dx 2 tan x − 1
1 d
= (ln (tan x + 1) − ln (tan x − 1))
2 dx
sec2 x sec2 x
 
1
= −
2 tan x + 1 tan x − 1
1 (tan x − 1) − (tan x + 1)
= sec2 x
2 tan2 x − 1
1
sec2 x cos2 x
= =
1 − tan2 x sin2 x
1 − cos 2x

1
= = sec 2x
cos2 x − sin2 x

1
10. Answer: 5
By symmetry, Bill’s path must be the same as Alan’s, rotated 90◦ counter-clockwise and shifted to
start at (1, 0). If Alan’s position is given by (x, y), Bill’s is then given by (1 − y, x). The slope of Alan’s
x−y
path at a point (x, y) is then y 0 = 1−y−x . This is undefined at the given point, so noting that the
sign of the derivatives in the radius of curvature does not matter, we switch to the function x(y), with
(−1−x0 )(x−y)−(x0 −1)(1−y−x)
x0 = 1−y−x 1−0.6−0.4
x−y = 0.6−0.4 = 0 at the given point. Differentiating, x =
00
(x−y)2 =
(−1)(.6−.4)−(−1)(1−.6−.4)
(.6−.4)2 = −5. The radius of curvature is then
  2 3/2
dx
1+ dy (1 + 0)3/2 1
R=   = =
 d2 x 
 dy2  |−5| 5

The answer can also found using the parametric form of the radius curvature with (x0 (t), y 0 (t))

3
 Advanced Topics Test
2006 Stanford Math Tournament
February 25, 2006
1. A college student is about to break up with her boyfriend, a mathematics major who is apparently
more interested in math than her. Frustrated, she cries, ”You mathematicians have no soul! It’s all
numbers and equations! What is the root of your incompetence?!” Her boyfriend assumes she means
the square root of himself, or the square root of i. What two answers should he give?
   
0 1 8 6 4 2 0
2. Define A = . Find a vertical vector v such that (A + A + A + A + I)v = (where
3 0 11
I is the 2×2 identity matrix).
2006
k
P

3. Simplify: 10 (Your answer should contain no summations but may still contain binomial coeffi-
k=10
cients/combinations).

4. Rice University and Stanford University write questions and corresponding solutions for a high school
math tournament. The Rice group writes 10 questions every hour but make a mistake in calculating
their solutions 10% of the time. The Stanford group writes 20 problems every hour and makes solution
mistakes 20% of the time. Each school works for 10 hours and then sends all problems to Smartie
to be checked. However, Smartie isn’t really so smart, and only 75% of the problems she thinks are
wrong are actually incorrect. Smartie thinks 20% of questions from Rice have incorrect solutions, and
that 10% of questions from Stanford have incorrect solutions. This problem was definitely written and
solved correctly. What is the probability that Smartie thinks its solution is wrong?
∞
1
P
5. Evaluate: √ √
k k+2+(k+2) k
k=1

6. Ten teams of five runners each compete in a cross-country race. A runner finishing in nth place
contributes n points to his team, and there are no ties. The team with the lowest score wins. Assuming
the first place team does not have the same score as any other team, how many winning scores are
possible?
7. A lattice point in the plane is a point whose coordinates are both integers. Given a set of 100 distinct
lattice points in the plane, find the smallest number of line segments AB for which A and B are distinct
lattice points in this set and the midpoint of AB is also a lattice point (not necessarily in the set).

n+1 (ni)(nj)
8. The following computation arose in the research of mathematician P.D.: Let ki (j) = 2n+1 2n for
(i+j )
0 ≤ i, j ≤ n.
n
X
ki (j).
j=0
 2006   2006 
, 2 , . . . , 2006
 
9. How many positive integers appear in the list 1 2006 where bxc represents the
greatest integer that does not exceed x?
10. Evaluate:
∞  
X 1
arctan 2
n=1
n −n+1

1
 Advanced Topics Solutions
2006 Stanford Math Tournament
February 25, 2006
√
2
1. Answer: ± 2
(1 + i)
For an answer in the form z√= a + bi note that
√
z 2 = a2 − b2 + 2abi. The real part is zero, so a = b.
2 2 2
2ab = 2a = 1 so a = b = ± 2 . Thus z = ± 2 (1 + i). One can use polar coordinates and De Moivre’s
theorem to arrive at the same result.
 
0
2. Answer: 1
11
 
2 3 0
A = = 3I Thus A4 = 9I 2 = 9I. We can see A6 = 27I and A8 = 81I. Thus A8 + A6 +
0 3         
4 2 121 0 a 121 0 a 121a
A + A + I = 121I = . Let v = and then = .
0 121 b 0 121 b 121b
 
1 0
Setting 121a = 0 and 121b = 11. This means a = 0 and b = 11 . Thus v = 1 .
11
2007

3. Answer: 11
Imagine a sequence of n numbers, {1, 2, 3 . . . , n+1}. A combination of k +1 elements may be chosen by
first choosing k from the set {1,. . . ,k} and attaching the (k + 1)th number. Then another combination
can be formed by choosing k from the set {1, . . . , k + 1} and attaching the (k + 2)th number. You may
continue in this

fashion
until choosing k from {1, . . . , n}. Therefore the summation that

we
ask for
is
equal to n+1 2007 10 11 12 13


k+1 = 11 . To check, you may examine a smaller sum such as 10 + 10 + 10 = 11

1
4. Answer: 25
C=correct problem; W=wrong problem;
C*=Smartie thinks a problem is correct; W*=Smartie thinks a problem is wrong;
S=problem from Stanford; R=problem from Rice
We are given P (W |W ∗) = 43 , P (W ∗ |R) = 15 , and P (W ∗ |S) = 10
1
. We can solve for P (R) = 13 , P (S) =
2 #correct problems 9·10+16·10
3 , and P (C) = total problems = 10·10+20·10 = 56 .
We want to find P (W ∗ |C):

P (C|W ∗) · P (W ∗)
P (W ∗ |C) = , where
P (C)
P (W ∗) = P (W ∗ |R) · P (R) + P (W ∗ |S) · P (S)
1 1 1 2
= · + · , and
5 3 10 3
P (C|W ∗) = 1 − P (W |W ∗)
3 1
=1− =
4 4

1 2
4 · 15 1
So P (W ∗ |C) = 5 = 25 .
6

√
2− 2
5. Answer: 4

1
 ∞ ∞
X 1 X 1 1
√ √ = p √ √
k=1
k k + 2 + (k + 2) k k=1 k(k + 2) k + k + 2
∞  
X 1 k+2 k
= p −
k(k + 2) 2 2
k=1
∞  
1X 1 1
= √ −√
2 k k+2
k=1
 
1 1 1
= +√
2 1 2
√
2− 2
=
4
The infinite series in its final form is a telescoping sum.
6. Answer: 112
The teams’ scores must sum to 1 + 2 + . . . + 50 = 12 · 50 · 51 = 1275. The winning score must be no
1
larger than 10 · 1275 = 127.5 and is at least 1 + 2 + 3 + 4 + 5 = 15. However, not all scores between
15 and 127 inclusively are possible because all teams must have integer scores and no team can tie the
winning team. If the winning score is s, the sum of all teams’ scores is at least s + 9(s + 1) = 10s + 9,
so solving gives s ≤ 126. Hence, 126 − 15 + 1 = 112 winning scores are possible.
7. Answer: 1200  
0
x+x0 y+y
The midpoint of the segment connecting (x, y) and (x0 , y 0 ) is 2 , 2 . Therefore a and a0 must
have the same parity, as must b and b0 for the midpoint to be a lattice point. We therefore divide the
set into four groups: (even,even), (even,odd), (odd,even), (odd,odd), with the number of points in each
group a, b, c, d. The number of such segments is then
       
a b c d a(a − 1) b(b − 1) c(c − 1) d(d − 1)
+ + + = + + +
2 2 2 2 2 2 2 2
1 2
a − a + b2 − b + c2 − c + d2 − d


=
2
1 2
a + b2 + c2 + d2 − 100


=
2
This is minimized when a = b = c = d, giving a value of 21 (4 · 252 − 100) = 1200.
8. Answer: 1
Expanding ki (j) we have
i+j

 
(n + 1)n!n!(i + j)!(2n − i − j)! i 2n + 1
ki (j) = = 2n−i−j
.
(2n + 1)i!(n − i)!j!(n − j)!(2n)!


n−i
n+1
We claim that
n     
X i+j 2n − i − j 2n + 1
= .
j=0
i n−i n+1
We show this by bijection. If we pick n + 1 items from among 2n + 1 we must choose the i + 1st element
at position i + 1, i + 2, ... , or 2n + 1 − (n − i). For each such choice, we can pick the first i objects
from among the first i + j and the last n − i from among the last 2n − i − j, 0 ≤ j ≤ n. Thus
n     
X i+j 2n − i − j 2n + 1
= .
j=0
i n−i n+1

2
 9. Answer: 88
Let f (n) = 2006n . For sufficiently small n, bf (n)c takes a different value. Consequently, for all sufficiently
small m, there exists at least one value of n for which bf (n)c = m. Note that if a and b are positive
real numbers for which a = bac + a0 and b = bbc + b0 , then bac − bbc = a − b + (b0 − a0 ). Note also that
|b0 − a0 | < 1. Hence, if f (n)−f (n+1) > 1, then bf (n)c > bf (n + 1)c. Also, if f (n)−f (n+ 1) < 1, then
bf (n)c − bf (n + 1)c < 2 (i.e. equals 0 or 1). The equation 2006 2006 2
x − x+1 = 1 implies x + x − 2006 = 0,
√
or x = 21 (5 321 − 1) < 21 (5(18) − 1) = 44.5. Note also that x > 21 (5(17) − 1) = 42. So 42 < x < 45,
implying that if n ≥ 45, f (n) − f (n + 1) < 1 and that if n ≤ 42, f (n) − f (n + 1) > 1. Evaluating bf (n)c
for n = 42, 43, 44, and 45, we see that each are unique. We conclude  that the first 44 terms are unique
integers. The rest of the terms take on the values 1, 2, . . . , 200645 , or 44 additional terms.

10. Answer: π2
We show by induction that
m  
X 1
arctan = arctan(m)
n=1
n2 − n + 1
.  
1
Clearly arctan 1−1+1 = arctan 1.
Pm  
1
If n=1 arctan n2 −n+1 = arctan(m), then

m+1  !   
X 1 1
tan arctan 2
= tan arctan(m) + arctan 2
n=1
n −n+1 (m + 1) − (m + 1) + 1
1
m+ m2 +m+1
= 1
1− m m2 +m+1
2
m(m + m + 1) + 1
=
m2 + m + 1 − m
m3 + m2 + m + 1
=
m2 + 1
(m + 1)(m2 + 1)
=
m2 + 1
m+1  !
X 1
tan arctan =m+1
n=1
n2 − n + 1
m+1  
X 1
arctan = arctan(m + 1)
n=1
n2 − n + 1

π
Thus as m → ∞ the sum goes to arctan(+∞) = 2.

3
 General Test
2006 Stanford Math Tournament
February 25, 2006
1. After a cyclist has gone 23 of his route, he gets a flat tire. Finishing on foot, he spends twice as long
walking as he did riding. How many times as fast does he ride as walk?
2. A customer enters a supermarket. The probability that the customer buys (a) bread is .60, (b) milk
is .50 and (c) both bread and milk is .30. What is the probability that the customer would buy either
bread or milk or both?
3. After a typist has written ten letters and had addressed the ten corresponding envelopes, a careless
mailing clerk inserted the letters in the envelopes at random, one letter per envelope. What is the
probability that exactly nine letters were inserted in the proper envelopes?
4. In a certain tournament bracket, a player must be defeated three times to be eliminated. If 512
contestants enter the tournament, what is the greatest number of games that could be played?
5. A geometric series is one where the ratio between each two consecutive terms is constant (ex. 3, 6, 12, 24, . . .).
The fifth term of a geometric series is 5!, and the sixth term is 6!. What is the fourth term?
6. An alarm clock runs 4 minutes slow every hour. It was set right 3 21 hours ago. Now another clock
which is correct shows noon. In how many minutes, to the nearest minute, will the alarm clock show
noon?
7. An aircraft is equipped with three engines that operate independently. The probability of an engine
failure is .01. What is the probability of a successful flight if only one engine is needed for the successful
operation of the aircraft?
8. Given two 2’s, ”plus” can be changed to ”times” without changing the result: 2+2 = 2·2. The solution
with three numbers is easy too: 1 + 2 + 3 = 1 · 2 · 3. There are three answers for the five-number case.
Which five numbers with this property has the largest sum?
9. If to the numerator and denominator of the fraction 13 you add its denominator 3, the fraction will
double. Find a fraction which will triple when its denominator is added to its numerator and to its
denominator and find one that will quadruple.
10. What is the square root of the sum of the first 2006 positive odd integers?
11. An insurance company believes that people can be divided into 2 classes: those who are accident prone
and those who are not. Their statistics show that an accident prone person will have an accident in a
yearly period with probability 0.4, whereas this probability is 0.2 for the other kind. Given that 30%
of people are accident prone, what is the probability that a new policyholder will have an accident
within a year of purchasing a policy?
12. What is the largest prime factor of 8091?
13. 123456789 = 100. Here is the only way to insert 7 pluses and/or minus signs between the digits on the
left side to make the equation correct: 1 + 2 + 3 − 4 + 5 + 6 + 78 + 9 = 100. Do this with only three
plus or minus signs.
14. Determine the area of the region defined by x2 + y 2 ≤ π 2 and y ≥ sin x.
15. The odometer of a family car shows 15, 951 miles. The driver noticed that this number is palindromic:
it reads the same backward as forwards. ”Curious,” the driver said to himself, ”it will be a long time
before that happens again.” Surprised, he saw his third palindromic odometer reading (not counting
15, 951) exactly five hours later. How many miles per hour was the car traveling in those 5 hours
(assuming speed was constant)?

1
16. Points A1 , A2 , . . . are placed on a circle with center O such that 6 OAn An+1 = 35o and An 6= An+2 for
all positive integers n. What is the smallest n > 1 for which An = A1 ?
17. Car A is traveling at 20 miles per hour. Car B is 1 mile behind, following at 30 miles per hour. A fast
fly can move at 40 miles per hour. The fly begins on the front bumper of car B, and flies back and
forth between the two cars. How may miles will the fly travel before it is crushed in the collision?
18. Alex and Brian take turns shooting free throws until they each shoot twice. Alex and Brian have 80%
and 60% chances of making their free throws, respectively. What is the probability that after each free
throw they take, Alex has made at least as many free throws as Brian if Brian shoots first?
19. When the celebrated German mathematician Karl Gauss (1777-1855) was nine years old, he was asked
to add all the integers from 1 through 100. He quickly added 1 and 100, 2 and 99, and so on for 50
pairs of numbers each adding in 101. His answer was 50 · 101 = 5, 050. Now find the sum of all the
digits in the integers from 1 through 1,000,000 (i.e. all the digits in those numbers, not the numbers
themselves).
20. Given a random string of 33 bits (0 or 1), how many (they can overlap) occurrences of two consecutive
0’s would you expect? (i.e. ”100101” has 1 occurrence, ”0001” has 2 occurrences)
21. How many positive integers less than 2005 are relatively prime to 1001?
22. A certain college student had the night of February 23 to work on a chemistry problem set and a math
problem set (both due on February 24, 2006). If the student worked on his problem sets in the math
library, the probability of him finishing his math problem set that night is 95% and the probability of
him finishing his chemistry problem set that night is 75%. If the student worked on his problem sets
in the chemistry library, the probability of him finishing his chemistry problem set that night is 90%
and the probability of him finishing his math problem set that night is 80%. Since he had no bicycle,
he could only work in one of the libraries on February 23rd. He works in the math library with a
probability of 60%. Given that he finished both problem sets that night, what is the probability that
he worked on the problem sets in the math library?
23. Consider two mirrors placed at a right angle to each other and two points A at (x, y) and B at (a, b).
Suppose a person standing at point A shines a laser pointer so that it hits both mirrors and then hits
a person standing at point B (as shown in the picture). What is the total distance that the light ray
travels, in terms of a, b, x, and y? Assume that x, y, a, and b are positive.

24. The number 555,555,555,555 factors into eight distinct prime factors, each with a multiplicity of 1.
What are the three largest prime factors of 555,555,555,555?
25. P
For positive integers n let D(n) denote the set of positive integers that divide n and let S(n) =
1
k∈D(n) k . What is S(2006)? Answer with a fraction reduced to lowest terms.

2
 General Solutions
2006 Stanford Math Tournament
February 25, 2006
1. Answer: 4
He walks 31 of the way or half as far as he rides, but it takes him twice as long. Therefore, he rides
four times as fast as he walks.
2. Answer: .80 or 4/5
Let B = event that the customer buys bread, M = the event that the customer buys milk. Then,
according to the rule of addition, we have
P (B ∪ M ) = P (B) + P (M ) − P (B ∩ M ) = .60 + .50 − .30 = .80
3. Answer: 0
If nine letters are in the correct envelopes, the tenth must be also, so the probability is zero.

4. Answer: 1535
511 players must be defeated thrice, the winner might be beaten twice.
511 · 3 + 2 = 1535
5. Answer: 20
The ratio between consecutive terms is 6!/5! = 6. So the fourth term is 1/6 of the fifth term. We get
1
6 (5!) = 20.

6. Answer: 15
In 3 12 hours the alarm clock has become 14 minutes slow. When the alarm clock shows noon, it will
fall behind approximately an additonal minute. Its hands will show noon in 15 minutes.
7. Answer: .999999
Let P(S) = the probability of a successful flight.
Let P(S’) = the probability of an unsucessful flight.
Let P(F) = the probability of an engine failure.
Since the flight is unsuccessful only when all three engines fail, then the probability of unsuccessful
flight is
P (S 0 ) = P (F ∩ F ∩ F ) = (.01)(.01)(.01) = (.01)3
But P (S) = 1 − P (S 0 ) = 1 − (.01)3 = 1 − .000001 = .999999.
8. Answer: 1, 1, 1, 2, 5
1 + 1 + 1 + 2 + 5 = 1 · 1 · 1 · 2 · 5, 1 + 1 + 1 + 3 + 3 = 1 · 1 · 1 · 3 · 3, 1 + 1 + 2 + 2 + 2 = 1 · 1 · 2 · 2 · 2
9. Answer: 51 ; 71 .
Any fraction with 1 for numerator and any odd number (2n-1) for the denominator increases to n times
its value when its denominator is added to its numerator and to its denominator. All other answers
simplify to one of these fractions.
10. Answer: 2006
sum of the first odd positive integer: 1 = 1 = 11
sum of the first 2 odd positive integers: 1 + 3 = 4 = 22
sum of the first 3 odd positive integers: 1 + 3 + 5 = 9 = 32
sum of the first 4 odd positive integers: 1 + 3 + 5 + 7 = 16 = 42
sum of the first 5 odd positive integers: 1 + 3 + 5 + 7 + 9 = 25 = 52
sum of the√first 6 odd positive integers: 1 + 3 + 5 + 7 + 9 + 11 = 36 = 62 and so on...
therefore 20062 = 2006

1
11. Answer: 0.26
Let A denote the event that the policyholder has an accident within a year.
Let B denote the event that the policyholder is accident prone.
P (A) = P (A|B)P (B) + P (A|B C )P (B C ) = 0.4 · 0.3 + 0.2 · 0.7 = 0.26
12. Answer: 31
Write 8091 as 8100 − 9 = 902 − 32 = (90 − 3)(90 + 3) = (87)(93) = 32 · 29 · 31. So 31 is the answer.
13. Answer: 123 − 45 − 67 + 89
3
14. Answer: π2
By symmetry, the area is simply half the area of the circle.

15. Answer: 62
The first digit of 15,951 could not change in 5 hours. Therefore, 1 is the first and last digit of the
new number. The second and fourth digits changed to 6. If the middle digit is 0, 1, 2,..., then the car
traveled 110, 210, 310, ..., miles in 5 hours. Therefore the car was traveling 62 miles per hour.
16. Answer: 37 The key is to note that the minor arcs between An and An+1 are all congruent with
measure 110o . Then the solution is the smallest n for which 110(n − 1)o is a multiple of 360o , which
is clearly 37.
17. Answer: 4
The fly will travel for a tenth of an hour at 40 miles per hour, so it will travel a total of 4 miles.
18. Answer: 508 625
The probability that Brian makes more free throws than Alex at any point is the probability that Brian
makes the first and Alex misses the first plus the probability that they made the same number of free
throws after each taking one and then Brian makes the second and Alex misses the second.
So ( 35 )( 15 ) + ( 53 )( 15 )[( 53 )( 45 ) + ( 52 )( 15 )] = 25
3 3
+ ( 25 )( 14 117
25 ) = 625
So the probability that Alex has made at least as many free throws as Brian after each one taken is
1 − 117
625 = 625 .
508

19. Answer: 27, 000, 001

The numbers can be grouped by pairs:
999,999 and 0; 999,998 and 1; 999,997 and 2; and so on.
There are half a million pairs, and the sum of the digits in each pair is 54. The digits in the unpaired
number 1,000,000 add to 1 then
(500, 000 · 54) + 1 = 27, 000, 001
20. Answer: 8
1
a1 = 0; a2 =
4
1 1
an = an−1 + ⇒ an = (n − 1)
4 4
1 1
(33 − 1) = (32) = 8.
4 4

21. Answer: 1442

First, factor 1001 into 7 · 11 · 13. A number relatively prime to 1001 must not be divisible by 7, 11 nor
13. There are 2004 integers less than 2005 and b 2004 2004
7 c = 286 of them are multiples of 7. b 11 c = 182 of
2004
them are multiples of 11 and b 13 c = 154 of them are multiples of 13. We’ve double counted integers
divisible by 7 · 11, 11 · 13 and 7 · 13. We must add back in b 2004 2004 2004
77 c = 26, b 91 c = 22, and b 143 c = 14.
However, we counted multiples of 7, 11 and 13 three times in the first computation and then we took

2
 2004
them out 3 times in the last computation. We need to add them back. There are b 7·11·13 c = 2 of them.
Thus, the number of integers not relatively prime to 1001 is 286 + 182 + 154 − 26 − 22 − 14 + 2 = 562.
Thus 2004 − 562 = 1442 is our answer.
95
22. Answer: 159
Let M = the event that he works in the math library.
Let C = the event that he works in the chemistry library.
Let F = he finishes his problem set
P (M ) = 60%
P (C) = 40%
P (F |M ) = .95 · .75 = .7125
P (F |C) = .90 · .80 = .72

P (F |M )P (M )
P (M |F ) =
P (F |M )P (M ) + P (F |C)P (C)
(0.7125)(0.6) 95
= =
(0.7125)(0.6) + (0.72)(0.4) 159
p
23. Answer: (x + a)2 + (y + b)2

If you consider the mirror image of that light ray’s path for both mirrors you get a second path of
length equal to the actual path, which travels from (x, y) to (−a, −b) (as seen in the second pic-
ture).
p So the total length the light ray travels is the distance from (x, y) to (−a, −b). This would be
(x + a)2 + (y + b)2 .
24. Answer: 37, 101, 9901
555555555555 = 555 · 1001001001
= 555 · 1001 · 1000001
= (5 · 3 · 37)(7 · 11 · 13)(1003 + 1)
= (3 · 5 · 7 · 11 · 13 · 37)(100 + 1)(1002 − 100 + 1)
= 3 · 5 · 7 · 11 · 13 · 37 · 101 · 9901
1620
25. Answer: 1003
If Σ(n) is the sum of the positive factors of n then S(n) = Σ(n)/n.
Since we have the prime factorization 2006 = 2 · 17 · 59, the sum of its factors is (1 + 2)(1 + 17)(1 + 59) =
3240
3 · 18 · 60. Hence S(2006) = 2006 = 1620
1003 in lowest terms.

3
 Team Test
2006 Stanford Math Tournament
February 25, 2006
1. Given 4ABC, where A is at (0, 0), B is at (20, 0), and C is on the positive y-axis. Cone M is formed
when 4ABC is rotated about the x-axis, and cone N is formed when 4ABC is rotated about the
y-axis. If the volume of cone M minus the volume of cone N is 140π, find the length of BC.
n−1
P
2. In a given sequence {S1 , S2 , . . . , Sk }, for terms n ≥ 3, Sn = i · Sn−i . For example, if the first two
i=1
elements are 2 and 3, respectively, the third entry would be 1 · 3 + 2 · 2 = 7, and the fourth would be
1 · 7 + 2 · 3 + 3 · 2 = 19, and so on . Given that a sequence of integers having this form starts with 2,
and the 7th element is 68, what is the second element?

3. A triangle has altitudes of lengths 5 and 7. What is the maximum length of the third altitude?

4. Let x + y = a and xy = b. The expression x6 + y 6 can be written as a polynomial in terms of a and b.

What is this polynomial?
5. There exist two positive numbers x such that sin(arccos(tan(arcsin x))) = x. Find the product of the
two possible x.
6. The expression 16n +4n +1 is equivalent to the expression (2p(n) −1)/(2q(n) −1) for all positive integers
n > 1 where p(n) and q(n) are functions and p(n)
q(n) is constant. Find p(2006) − q(2006).

7. Let S be the set of all 3-tuples (a,b,c) that satisfy a + b + c = 3000 and a, b, c, > 0. If one of these
3-tuples is chosen at random, what’s the probability that a, b, or c is greater than or equal to 2, 500?
2
(n+1)
√1
P
8. Evaluate: lim k
n→∞ k=n2

9. 4ABC has AB = AC. Points M and N are midpoints of AB and AC, respectively. The medians
AB 2


M C and N B intersect at a right angle. Find BC .
10. Find the smallest positive integer m for which there are at least 11 even and 11 odd positive integers
3
n so that nn+2
+m
is an integer.

11. Polynomial P (x) = c2006 x2006 + c2005 x2005 + . . . + c1 x + c0 has roots r1 , r2 , . . . , r2006 . The coefficients
2006
ci cj P ri
satisfy 2i c2006−i = 2j c2006−j for all pairs of integers 0 ≤ i, j ≤ 2006. Given that rj = 42,
i6=j,i=1,j=1
2006
P
determine (r1 + r2 + . . . + r2006 ).
i=1

12. Find the total number of k-tuples (n1 , n2 , . . . , nk ) of positive integers so that ni+1 ≥ ni for each i,
and k regular polygons with numbers of sides n1 , n2 , . . . , nk respectively will fit into a tesselation at a
point. That is, the sum of one interior angle from each of the polygons is 360◦ .
13. A ray is drawn from the origin tangent to the graph of the upper part of the hyperbola y 2 = x2 − x + 1
in the first quadrant. This ray makes an angle of θ with the positive x-axis. Compute cos θ.

14. Find the smallest nonnegative integer n for which 2006

n is divisible by 73 .

1
15. Let ci denote the ith composite integer so that {ci } = 4, 6, 8, 9, .... Compute
∞
Y c2i
.
c2
i=1 i
−1

π2
Pn 1
(Hint: i=1 n2 = 6 ).

2
 Team Solutions
2006 Stanford Math Tournament
February 25, 2006
1. Answer: 29

20x2 π
140π = volume of cone M - volume of cone N = 13 · x2 · 20π − 31 · x · (20)2 π = 3 − 400xπ
3
20x2 − 400x = 420 ⇒ x2 − 20x − 21 = 0 ⇒ (x − 21)(x + 1) = 0 ⇒ x = 21, −1
But x must
q be positive, so x = 21.
2 2 √
BC = AB + AC = 202 + 212 = 29
2. Answer: −2
Let the first element be x, and the second,y. Writing out each element in terms of x and y gives
{x, y, 2x + y, 5x + 3y, 13x + 8y, . . .}, which is apparently the fibonacci sequence with every other ele-
ment as the coefficient of x or y. So the 6th element is 34x + 21y and the seventh, 89x + 55y. Solving
89 · 2 + 55 · y = 68 gives y = −2.
3. Answer: 17.5
Form 4ABC, and set a = BC, b = AC,and c = AB. Let 5 be the altitude from A, 7 be the altitude
from B, and call the third altitude h.
5a = 7b = h · c, so ac < h5 and cb = h7 .
Since a < b + c,
a b h h
= +1⇒ < +1
c c 5 7
 
1 1
h· − <1
5 7
7·5
so h < 7−5 = 17.5

4. Answer: a6 − 6a4 b + 9a2 b2 − 2b3

Note: (xn−1 + y n−1 )(x + y) = xn + y n + yxn−1 + xy n−1 = xn + y n + xy(xn−2 + y n−2 ).
Thus, let f (n) = xn + y n . We see f (n) = af (a − 1) − bf (n − 2).
x0 + y 0 = 2, so f (0) = 2
x1 + y 1 = x + y = a, so f (1) = a
f (2) = a2 − 2b
f (3) = a3 − 3ab
f (4) = a4 − 3a2 b − a2 b + 2b2 = a4 − 4a2 b + 2b2
f (5) = a5 − 4a3 b + 2ab2 − a3 b + 3ab2 = a6 − 6a3 b + 5ab2
f (6) = a6 − 5a4 b + 5a2 b2 − a4 b + 4a2 b2 − 2b3 = a6 − 6a4 b + 9a2 b2 − 2b3

1
5. Answer: 1

sin(arccos(tan(arcsin x))) = x
  
x
sin arccos √ =x
1 − x2
s  2
x
1− √ =x
1 − x2
r
1 − 2x2
=x
1 − x2
1 − 2x2 = x2 − x4
x4 − 3x2 + 1 = 0
√
Solving and restricting x to positive numbers: x2 = 3± 29−4
q √ q √ q
x = 3+2 5 or x = 3−2 5 . Multiplying these together, the answer is 9−5
4 .

6. Answer: 8024
Write the expression as x4 + x2 + 1 where x = 2n . This is equivalent to (x2 + 1)2 − x2 (by adding and
3
−1 x3 +1 6
26n −1
subtracting x2 ). This expression can be written as (x2 +x+1)(x2 −x+1) = xx−1 · x+1 = xx2 −1
−1 = 22n −1 .
Hence p(n) = 6n and q(n) = 2n. It’s not hard to see that this is the only solution by considering the
limit of each expression as n approaches infinity. The highest-order terms predominate: 24n and
2q(n)(p(n)/q(n)−1) . This implies that p and q are linear functions. Exact functions can be determined
by evaluating the expressions at n = 1 and n = 2 and solving for two variables. The answer is 8,024.
1
7. Answer: 12

This is a geometric probability problem. The set of 3-tuples above fits an equilateral triangle on
the plane x + y + z = 3000. We’re going to look at the sections of this triangle where x ≥ 2500.
This is a triangle with vertices (2500, 500, 0), (2500, 0, 500), and (3000, √
0, 0).
√
This is an equilateral
√ √
side2 3 (500 2)2 3 √
triangle with length 500 2. The area of this triangle is 4 = 4 = 125000 3. Since x,
y, or z√can be larger than√ 2500, we need to multiply this by 3 to get the total area that works: √
125000 3 · 3 =
√
375000
√
3. The total possible area is the whole triangle of side length 3000 2:
√
side2 3 (3000 2)2 3 √ √
375000 √3 1
4 = 4 = 4500000 3. So the overall probability is 4500000 3 = 12 .

2
 8. Answer: 2
2
(n+1)
√1
P
Let Sn = k
k=n2

(n+1)2 (n+1)2
X 1 X 1
p < Sn < √
(n + 1)2 n2
k=n2 k=n2

1
1
(n + 1)2 − n2 + 1 < Sn < (n + 1)2 − n2 + 1
n+1 n
2(n + 1) 2(n + 1)
< Sn <
n+1 n
1
2 < Sn < 2 +
n
2
(n+1)
√1
P
Thus lim k
=2
n→∞ k=n2

5
9. Answer: 2
x
Suppose the medians intersect at P . If BC = x, BP = CP = .
By a well-known property of
√
2 √
MP 1 x x 52
centroids, =
MC 3,so M P = 2
√
2
. Using the Pythagorean Theorem, we find that M B = 2 and so
q
AB 2
AB = x 52 . So = 52 .


BC

10. Answer: 638

Notice that n3 + 8 is divisible by n + 2. Therefore, m − 8 must be divisible by n + 2 for the expression
to be an integer. If f is a factor of m − 8, n = f − 2 is a corresponding suitable n; we then need
f ≥ 3 to make n > 0. Thus m − 8 must have twelve each odd and even factors including 1 and 2. To
make the number of odd and even factors equal in order to minimize m, the power of 2 in the prime
factorization of m − 8 must be 1. Suppose the prime factorization of m − 8 is then 21 · 3a · 5b · 7c · 11d
(larger prime factors will clearly not minimize m). Then (a + 1)(b + 1)(c + 1)(d + 1) ≥ 12. To minimize
m, a ≥ b ≥ c ≥ d. We then examine values of m−8 2 to determine the best (a, b, c, d). 3 · 5 · 7 · 11 = 1155,
2 2
3 · 5 · 7 = 315. Moving any more factors into smaller primes involves multiplying by 37 or 35 (or
2

subsequent larger powers of 3), which increases the value. Therefore m − 8 = 2 · 32 · 5 · 7, so m = 638.
11. Answer: 64
Using the first condition with j = 1003 we get ci = 2(1003 − i)c2006−i . Replace the coefficients of
2
P(x )
P in this manner and notice that x2006 2006 = P (x). Therefore if r is a solution of P (x) = 0 then
P (2/r) = 0. Then:
2006 2006
!2
X ri X 2006 X 1 1 X
2006
= ri − 2006 = ri − 2006 = 42
rj i=1 i=1 i
r 2 i=1
i6=j,i=1,j=1

Solving for the desired sum gives 64.

12. Answer: 17
k   k
180 − 360 1
P P
ni = 0, so k/2 − 1 = ni . Clearly, 3 ≤ k ≤ 6, since the interior angles are less than
i=1 i=1
◦
180 , and six equilateral triangles maximize k. For each k, bounds can be established on the smallest or
largest ni . From then, we can fix all but two of the ni , solve algebraically, then use reasonable guesswork
to find all integer solutions. For k = 3, fix n1 at 3, 4, 5, or 6 and then solve 32 − 1 = n11 + n12 + n13 .
This yields 10 solutions. For k = 4, n4 = 3 or 4; there are 4 solutions. For k = 5, n5 = n4 = n3 = 3,
giving two solutions. Finally there is of course only one solution for k = 6. 10 + 4 + 2 + 1 = 17

3
 √
13. Answer: 2 7 7
It is clear from drawing the graph that we want to find the cosine of the smallest angle θ(0 < θ < π2 )
such that a ray leaving the origin at angle θ will hit the graph of the hyperbola in the first quadrant.
Since cos θ is a decreasing function on this interval, we want the largest possible value of cos θ.
We begin by writing the hyperbola in polar coordinates: r2 sin2 θ = r2 cos2 θ − r cos θ + 1.
Using sin2 θ = 1 − cos2 θ and collecting like terms, we get: (2 cos2 θ − 1)r2 − (cos θ)r + 1 = 0.
Now we can use the quadratic formula to solve for r:
p
cos θ ± cos2 θ − 4(2 cos2 θ − 1)
r=
4 cos2 θ − 2
If there are any solutions for r, the quantity under the square root must be nonnegative:
cos2 θ ≥ 8 cos2 θ − 4
7 cos2 θ ≤ 4
√
2 7
cos θ ≤
7
So the angle we are looking for has √
2 7
cos θ =
7
14. Answer: 292
First we find the largest power of an integer d that divides k!. Notice that kd of the integers 1, 2, . . . , k
 
k
are divisible by d, d2 are divisible by d2 , and so on. The largest power we are looking for is then
2006
k  k   k 
2006!
d +  d2 + d3 + . . .. Now let m = 2006 − n, so that n = n!m! ; the largest
 0powermof 7divisor is
2006
− 72 − 72 +. . . Note that if d = d +n and d = d +m0 ,
  n   m 
 2006   n   m 
n n m

then 7 − 7 − 7 + 72
then d = d leaves a remainder of r = n0 + m0 or n0 + m0 − d, whichever satisfies 0 ≤ r < d.
2006 n+m

Therefore 2006 − d − d = 0 or 1. To make this 1 in order to get large divisors of 2006

  m n

d n , we need
m0 , n0 > r. We therefore find the remainders when 2006 is divided by 7, 72 , and 73 : 4, 46, and 291.
Therefore n must leave a remainder of at least 292 when divided by 343, so we try n = 292, which has
remainders of 5 and 47 when divided by 7 and 49.
12
15. Answer: π2
Write
∞ ∞
Y p2 Y c2 Y n2 Y n n
= =
p prime
p2 − 1 c composite
2 2
c − 1 n=2 n − 1 n=2 n − 1 n + 1
which telescopes and evaluates to 2. Meanwhile we can write
Y p2 Y 1
= .
p prime
p − 1 p prime 1 − p12
2

The latter is equivalently rewritten:

Y X ∞ 
Y 1 1 1
1 + 2 + 4 + ... = .
p prime
p p p prime n=0
p2n

When we distribute the infinite product over the infinite sum, we get a sum of terms. Each term is of
P∞ 2
the form m12 for integer m. Each m appears exactly once, so the product is equal to n=1 n12 = π6 .
Hence
Y c2 2 12
2−1
= π2 = 2 .
c composite
c 6
π

4
 Algebra Test
2007 Stanford Math Tournament
March 4, 2007
1. Find all real roots of f if f (x1/9 ) = x2 − 3x − 4.
2. Given that x1 > 0 and x2 = 4x1 are solutions to ax2 + bx + c and that 3a = 2(c − b), what is x1 ?
3. Let a, b, c be the roots of x3 − 7x2 − 6x + 5 = 0. Compute (a + b)(a + c)(b + c).

4. How many positive integers n, with n ≤ 2007, yield a solution for x (where x is real) in the equation
bxc + b2xc + b3xc = n?
5. The polynomial −400x5 + 2660x4 − 3602x3 + 1510x2 + 18x − 90 has five rational roots. Suppose you
guess a rational number which could possibly be a root (according to the rational root theorem). What
is the probability that it actually is a root?

6. What is the largest prime factor of 49 + 94 ?

1 1 1
7. Find the minimum value of xy + x + y + xy + x + y for x, y > 0 real.

8. If r + s + t = 3, r2 + s2 + t2 = 1, and r3 + s3 + t3 = 3, compute rst.

9. Find a2 + b2 given that a, b are real and satisfy
1 1
a=b+ 1 ; b=a− 1
a+ 1
b+ a+···
b+ 1
a− b+···

10. Evaluate
2007
X
(−1)k k 2
k=1
 Algebra Solutions
2007 Stanford Math Tournament
March 4, 2007
1. Answer: 41/9 , −1
f (x1/9 ) = (x − 4)(x + 1) so f = 0 means x = 4 or x = −1, so f (41/9 ) = f (−1) = 0.
1
2. Answer: 4
Since (x − x1 )(x − 4x1 ) = x2 − 5x1 x + 4x21 , we know 3 = 2(4x21 + 5x1 ) so x1 = 41 . (The other root is
negative.)
3. Answer: −37
Simply note that (a + b)(a + c)(b + c) = (ab + ac + bc)(a + b + c) − abc = −6 · 7 + 5 = −37.
4. Answer: 1339
Let k be a nonnegative integer. Let f (x) = bxc + b2xc + b3xc. If k ≤ x < k + 31 , then f (x) = 6k. If
k + 31 ≤ x < k + 12 , then f (x) = 6k + 1. If k + 12 ≤ x < k + 32 , then f (x) = 6k + 2. If k + 23 ≤ x < k + 1,
then f (x) = 6k + 3. There is therefore only a solution if n is 0, 1, 2, or 3 mod 6; there are 2004 · 46 + 3
of these.
5
5. Answer: 144
There are clearly five correct guesses; counting the number of possible guesses is the difficult part. A
possible guess q is ±1 times a divisor of 90 divided by a divisor of 400. We count these by extending
the idea of prime factorization: from the factorizations of 90 and 400: we have q = 2i 3j 5k where
−4 ≤ i ≤ 1, 0 ≤ j ≤ 2, and −2 ≤ k ≤ 1. There are thus 6 · 3 · 4 = 72 possible fractions making 144
possible guesses.
6. Answer: 881
We can factor 4x4 + y 4 = (4x4 + 4x2 y 2 + y 4 ) − 4x2 y 2 = (2x2 + y 2 )2 − (2xy)2 = (2x2 + 2xy + y 2 )(2x2 −
2xy + y 2 ). Since 49 + 94 = 4(16)4 + 94 , we plug in to obtain the factoring 881 · 305. Quick checking
(up to 29) shows 881 to be prime.
7. Answer: 6
The expression is 6 times the arithmetic mean of the terms, which is is always greater than or equal
to the geometric mean, which is xy · x · y · x1 · y1 · xy
1
= 1. The minimum is achieved when all terms are
equal, i.e. x = y = 1.
8. Answer: 4
(r + s + t)3 − 3(r + s + t)(r2 + s2 + t2 ) + 2(r3 + s3 + t3 ) = 6rst - just plug in!
√
9. Answer: 5
1 1
Note that the equations reduce by substitution to a = b + a+1/a and b = a − b+1/b . Solving the second
4 2
for a, substituting into the first, and reducing
√
yields b + b − 1 = 0; solving this as a quadratic
√
in b2
2 5−1 2 5+1
yields only one positive value for b = 2 . Plugging back in and solving for a gives a = 2 .
10. Answer: −2015028
Note that (x + 1)2 − x2 = 2x + 1 so:
2007
X 1003
X
(−1)k k 2 = −20072 + (2(2k − 1) + 1)
k=1 k=1
1003 · 1003
= −20072 + 4 + 1003
2
= −20072 + 1003 · 2007 = 2007(1003 − 2007)
 Geometry Test
2007 Stanford Math Tournament
March 4, 2007
1. An equilateral triangle has perimeter numerically equal to its area, which is not zero. Find its side
length.
2. Two spheres of radius 2 pass through each other’s center. Find the surface area of the regular octahe-
dron inscribed within the space enclosed by both spheres.

3. Cumulation of a polyhedron means replacing each face with a pyramid of height h using the face as a
base. There is a cumulation of the cube of side length s which (after removing unecessary edges) has
twelve sides, each a congruent rhombus. What is the height h used in this cumulation?
4. Nathan is standing on vertex A of triangle ABC, with AB = 3, BC = 5, and CA = 4. Nathan walks
according to the following plan: He moves along the altitude-to-the-hypotenuse until he reaches the
hypotenuse. He has now cut the original triangle into two triangles; he now walks along the altitude
to the hypotenuse of the larger one. He repeats this process forever. What is the total distance that
Nathan walks?
5. Given an octahedron with every edge of length s, what is the radius of the largest sphere that will fit
in this octahedron?
6. Let T IN A be a quadrilateral with IA = 8, IN = 4, m6 T = 30◦ , m6 N AT = 60◦ , and m6 T IA =
m6 IN A. Find N A.
7. Two regular tetrahedra of side length 2 are positioned such that the midpoint of each side of one
coincides with the midpoint of a side of the other, and the tetrahedra themselves do not coincide. Find
the volume of the region in which they overlap.
8. ∆ABC has AB = AC. Points M and N are midpoints of AB and AC, respectively. The medians
AB 2
M C and N B intersect at a right angle. Find ( BC ) .

√S, T lie in the pane with S on P R and R on QT . If P Q = 5, P S = 3, P R = 5, QS = 3,

9. Points P, Q, R,
and RT = 4/ 3, what is ST ?
10. A car starts moving at constant speed at the origin facing in the positive y-direction. Its minimum
turning radius is such that it the soonest it can return to the x-axis is after driving a distance d. Let
Γ be the boundary
√ 
of the region the car can reach by driving at most a distance d; find an x > 0 so
d d 3
that x, 3 + 2π is on Γ.
 Geometry Solutions
2007 Stanford Math Tournament
March 4, 2007
√
1. Answer: 4 3
√
3s2
√ 12
√
Let the side length be s. 3s = 4 ⇒ 12s = 3s2 ⇒ s = √
3
=4 3
√
2. Answer: 8 3
Since each side of the octahedron is a radius of a sphere, the surface area is the area of 8 equilateral
√ √
triangles with side length 2. 2·2 3 · 8 = 8 3.
s
3. Answer: 2
Let A, B be the apexes of two pyramids over adjacent faces, and let O lie at the center of the cube.
For the cumulation to form rhombuses instead of tetrahedrons, the hypotenuse of the right triangle
OAB
√ must intersect the shared edge of the two faces at a point P . OP is then an altitude, with length
s 2/2, and so OA is s, with s/2 lying inside the cube, leaving h = s/2.
4. Answer: 12
Let M be the endpoint of the altitude on the hypotenuse. Since we are dealing with right triangles,
∆M AC ∼ ∆ABC, so AM = 12/5. Let N be the endpoint he reaches on side AC. ∆M AC ∼ ∆N AM ,
so M N
AM = 4/5. This means
that each altitude that he walks gets shorter by a factor of 4/5. The total
distance is thus 12 4
5 / 1 − 5 = 12.
√
s 6
5. Answer: 6
Let ABC be the vertices on one face of the octahedron, and O be the center of the octahedron (and
the circle). Let M be the midpoint of AB and P be the center of ∆ABC. Then OM = 2s , and
√ q √
M P = 31 · s 2 3 , and OP ⊥ M P . So OP = s ( 2l )2 − ( s 2 3 )2 = √s6 .
√
12 5
6. Answer: 5
Extend IT and N A to point X. Clearly, m6 X = 90. Note that ∆XIA ∼ ∆XN I. Since (XC)/(XI) =
4/8 = 1/2, we can use the Pythagorean Theorem to solve for XC and XI to get XN = √45 and
√ √
XI = √85 . Since XN/XI = XI/XA, XA = 16 5/5, and N A = XA − XN = 125 5 .
√
2
7. Answer: 3
The regions in which they do NOT overlap are the smaller tetrahedra of side length 1 positioned at 3
vertices of both larger tetrahedra. The volume of the overlap region is the volume of one tetrahedron
of side length 2, minus the volume of four tetrahedra of side length 1.
In general, the volume of a tetrahedron of side length √s2 can be found by noting that a cube consists
of regular tetrahedron and 4 triangular pyramids (formed from a vertex of the cube and 3 adjacent
s2
·s
vertices). The cube has side length s, so the volume of the tetrahedron is s3 − 4 · 23 = 31 s3 , or 1/3 the
√
volume of a cube of side length s. So the volume of a tetrahedron of side length 1 is 13 · √18 · (1)3 = 122 .
√ √ √
2 4 2 2
So the total volume of the region we are looking for is 12 · 23 − 12 = 3 .
5
8. Answer: 2
x
Suppose the medians intersect at P . If BC = x, BP = CP = .
By a well-known property of
√
2 √
MP 1 x x 5/2
centroids, M = , so M P = √ . Using the Pythagorean Theorem, we find that M B = ⇒
qC 3 2 2 2
2
AB = x · 52 . So BC
AB
= 52 .

1
 √
9. Answer: 2 3
Stewart’s Theorem states that (P Q)2 (SR)+(QR)2 (P S) = (QS)2 (P R)+(P S)(P R)(SR). (This can√be
derived by applying the law of cosines to 6 P SQ and 6 RSQ.) Plugging
√ in and solving gives QR = 5/ 3,
and applying the theorem again to triangle QST gives ST = 2 3.
√
d 3 d
10. Answer: 3
+ 2π
The car can drive the farthest (reaching Γ) by turning as sharply as possible for some distance then
driving straight. Suppose it turns while driving a length s. Noting the correspondence between angle
at the center of a circle and arclength, it has then reached the point πd , 0 + πd − cos sπ sπ





d , sin d .
It now
has d − s

left to drive in the direction (since the tangent is perpendicular to the radius)
sin sπ
d , cos d
sπ
. Adding up, we have a parametric equation for the boundary:
 
d d  sπ   sπ    sπ   sπ 
Γ(s) = ,0 + − cos , sin + (d − s) sin , cos
π π d d d d

Solving for s would be very difficult, but noting that one term has a π in the denominator and one
doesn’t, we can make the educated guess s = d3 ; this indeed works.

2
 Calculus Test
2007 Stanford Math Tournament
March 4, 2007
1. Find
−1 + cos x
lim
x→0 3x2 + 4x3

2. A line through the origin is tangent to y = x3 + 3x + 1 at the point (a, b). What is a?
3. A boat springs a leak at time t = 0, with water coming in at constant rate. At a time t = τ > 0 hours,
someone notices that there is a leak and starts to record distance the boat travels. The boat’s speed is
inversely related to the amount of water in the boat. If the boat travels twice as far in the first hour
as in the second hour, what is τ ?
Rπ ∞
I(5n ).
P
4. Let I(n) = 0
sin(nx)dx. Find
n=0

d
5. Let Θk (x) be 0 for x < k and 1 for x ≥ k. The Dirac delta “function” is defined to be δk (x) = dx Θk (x).
d2
(It’s really called a distribution, and we promise it makes sense.) Suppose dx2 f (x) = δ1 (x) + δ2 (x) and
f (0) = f 0 (0) = 0. What is f (5)?
6. Point A is chosen randomly from the circumference of the unit circle, while point B is chosen randomly
in the interior. A rectangle is then constructed using A, B as opposite vertices, with sides parallel or
perpendicular to the coordinate axes. What is the probability that the rectangle lies entirely inside
the circle?
√
7. A balloon in cross-section has the equation y = ± 2x − x2 e−x/2 , with the x-axis beginning at the top
of the balloon pointing toward the knot at the bottom. What is its volume?
8. Silas does nothing but sleep, drink coffee, and prove theorems, and he never more than one at a time.
It takes 5 minutes to drink a cup of coffee. When doing math, Silas proves s + ln c theorems per hour,
where c is the number of cups of coffee he drinks per day, and s is the number of hours he sleeps per
day. How much coffee should Silas get in a day to prove the most theorems?
2n
1
P
9. Evaluate lim .
n→∞ k=n+1 k

x
10. Find the 10th nonzero term of the power series for f (x) = (x2 −1)2 (expanding about x = 0).
 Calculus Solutions
2007 Stanford Math Tournament
March 4, 2007
1. Answer: − 16
Use l’Hopital’s rule:
−1 + cos x − sin x − cos x
lim = lim = lim
x→0 3x2 + 4x3 x→0 6x + 12x2 x→0 6 + 12x
√
3
4
2. Answer: 2

a3 + 3a + 1 y
y 0 = 3a2 + 3 = =
a x
2a3 − 1
=0
a
1
a3 =
2

√
5−1
3. Answer: 2
The speed will cancel out so assume it is 1. We then have:
Z t+1 Z τ +2
1 1
dt = 2 dt
τ t τ +1 t
τ +1 τ +2
ln = 2 ln
τ τ +1
 2
τ +1 τ +2
=
τ τ +1
√
−1 ± 5
τ=
2
5
4. Answer: 2
π ∞ ∞
cos(nx) 
I(5n ) = 2/5n = 5/2
P P
For odd n, I(n) = − n  = 2/n, so
0 n=0 n=0

5. Answer: 7
We have f 0 (x) = (δ1 (x) + δ2 (x)) dx = Θ1 (x) + Θ2 (x) + C, and f 0 (0) = 0 so C = 0. Integrating up to
R

f is most easily accomplished graphically; the region under the curve from 0 to 5 is a 1 × 4 rectangle
from x = 1 to x = 5 with a 1 × 3 rectangle from x = 2 to x = 5 on top.
4
6. Answer: π2
Suppose A lies at polar coordinate 0 < θ < π/2. For the rectangle to lie within the circle, B must lie in
the rectangle with vertices at A, A reflected over the x-axis, A reflected over the y-axis, and A reflected
over both axes. Thus for this fixed A, the probability is (2 sin θ)(2 cos θ)/π = 2 sin(2θ)/π. The total
R π/2
probability is then π2 0 π2 sin(2θ)dθ. (Integrating over the circle requires taking the absolute value
of the expression for area, which then splits up into four sections identical to the one considered here.)

1
 4π
7. Answer: e2

Z 2 p 2 Z 2 2
V =π 2x − x2 e−x/2 dx = π (2x − x2 )e−x dx = πx2 e−x 0
0 0

8. Answer: 12 cups of coffee

The number of theorems proven is (s + ln c)(24 − s − c/12). Differentiating with respect to s gives
c c
24 − 12 − 2s − ln c = 0, so s = 12 − 24 − 12 ln c. This is a maximum in s since the second derivative
c
is −2. Plugging this back in and simplifying gives (12 − 24 + ln2c )2 = f (c)2 theorems proven. This
differentiates to 2f (c)f (c), so the derivative will be zero when either f (c) or f 0 (c) is zero. f (c) = 0
0

is difficult to solve, involving both a logarithm and a binomial, but f 0 (c) = 2c 1 1

− 24 , so c = 12 is a
solution. It is a maximum in c since the second derivative is 2f (c) + 2f (c)f (c), with f 00 (12) < 0,
0 2 00

f (12) > 0, and f 0 (12) = 0.

9. Answer: ln 2

2n n n
X 1 nX 1 X 1 1
= = k
k n k+n n1+ n
k=n+1 k=1 k=1
R2 1
This is a Riemann sum: 1 x
dx = ln 2.
10. Answer: 10x19  
1 1 1 1
R
Note that f (x)dx = 2(1−x2 ) = 4 1+x + 1−x . These are geometric sums, so we have

∞ ∞
Z !
1 X X
f (x)dx = xk + (−x)k
4
k=0 k=0
∞
1 X 2k
= x
2
k=0
∞
X
f (x) = kx2k−1
k=0

2
 Advanced Topics Test
2007 Stanford Math Tournament
March 4, 2007
    
1 2 2 x 0
1. The equation  1 3 4   y  =  0  has a solution for (x, y, z) besides (0, 0, 0)t . Find the
3 4 k z 0
value of k.
2. If log3 27 · logx 7 = log27 x · log7 3, find the least possible x.

3. Scott had too much to drink at the math building last night. He came out of the math building at
(0, 0) facing in the positive y-direction. He walked to his home at (4, 5) one block at a time, picking
either the +y or +x direction. What is the probability that Scott passed Paula’s house at (2, 3) on
the way home, assuming that all possible paths are equally likely? (Coordinates of points are given in
units of blocks.)

4. A village has n residents, named P1 , P2 , · · · , Pn . Each either tells the truth or lies all the time. For
each k: If k is a perfect square, Pk says that Pk+1 is lying. Otherwise, Pk says that Pk+1 is telling the
truth. (Pn talks about P1 .) What is the minimum number of residents, given that n > 1000?
5. The Tower of Hanoi game consists of three pegs, upon one of which are stacked n disks of radii 1, 2, . . . , n
from largest to smallest, bottom to top. The object is to move the stack to another peg by moving one
disk at a time, never placing a disk on top of a smaller one. What is the minimum number of moves
required to complete the game?
6. A graph is defined in polar coordinates by r(θ) = cos θ + 12 . Find the smallest x-coordinate of any
point on this graph.
7. A permutation a1
, a2 , . . . , an of the digits 1, 2, . . . , n is said to have k ascents if for exactly k of the ai ,
ai < ai + 1. Let nk denote the number of perumtations of length n with k ascents. Find the ordered
pair (a, b) such that nk = a n−1 + b n−1




k k−1 .

8. A board has 2n + 1 holes in a line, with n red pegs starting in the first n holes, followed by a gap, then
n blue pegs in the remaining n holes. The pegs may be moved either one hole forward into an empty
hole or a peg of one color may jump over a peg of the other color into an empty hole on the other side.
What is the fewest number of moves it takes to completely interchange the red and blue pegs?
9. Let p be an odd prime, and let  square of a non-zero element mod p. Find the number of 2 × 2
 a be the
a 0
matrices X such that X 2 ≡ mod p.
0 a
10. Let S be the set 1, 2, . . . , 2007, and let P be any polynomial of degree 2007 with positive integer
coefficients. A subset T of S is chosen so that the set {|P (a) − P (b)| : a 6= b, a, b ∈ T } cannot contain
more than 6 prime numbers. What is the largest possible number of elements T can contain?
 Advanced Topics Solutions
2007 Stanford Math Tournament
March 4, 2007
1. Answer: 2
This gives three equations: x + 2y + 2z = 0, x + 3y + 4z = 0, and 3x + 4y + kz = 0. To have a nonzero
solution, the first two equations must agree with the last one, so we add 5 times the first to −2 times
the second to get 3x + 4y + 2z. Thus if k = 2, we actually only have two equations and can get a
nonzero solution.
1
2. Answer: 343
1
We see that log3 27 = log3 33 = 3 log3 3 = 3, and logx 7 = . Thus we have 3 = log27 x·log7 3·log7 x.
log7 x
1 1 1
We also see that log27 x = = = log3 x, giving us 9 = log3 x · log7 3 · log7 x. Now
logx 33 3 logx 3 3
we note that log3 x · log7 3 = log7 3 · log3 x = log7 x, so we have 9 = (log7 x)2 , so log7 x = ±3, so
x = 7±3 = 343, 343
1

10
3. Answer: 21
By writing the number of ways to reach each vertex of the path, we see that we get Pascal’s triangle,
(5)(4)
so the answer is 3 9 2 = 10·6 10
126 = 21
(5)
4. Answer: 1024
By writing out the trustworthiness of the first several people, we see that it flips after each square; that
is, if T represents a truth-teller and L represents a liar, we either have L, T, T, T, L, L, L, L, L, T, . . . or
T, L, L, L, T, T, T, T, T, L, . . .. For this to end up so that Pn is right about P1 , we must have n between
an even square below and an odd one above, i.e. (2k)2 ≤ n < (2k + 1)2 for some k, so the least such n
is 1024.
5. Answer: 2n − 1
Let g(n) denote the solution to the problem. Note that to move the bottom disk we must first move
the top n − 1, requiring g(n − 1) moves. We then can move the last disk to the desired peg, and then
must move the other n − 1 disks on top of it, taking another g(n − 1) moves. Clearly g(1) = 1, and we
have g(n) = 2g(n − 1) + 1, so g(n) = 2n − 1. (We can check that this satisfies the recursive relation;
the easiest way to guess it is simply to write out the first several values.)
1
6. Answer: − 16
We have x = r cos θ.
1 1
x = (cos θ + ) cos θ = cos2 θ + cos θ
2 2
1 1 1
= cos2 θ + cos θ + −
2 16 16
 2
1 1 1
= cos θ + − ≥ −
4 16 16

7. Answer: (k + 1, n − k)
We
consider adding n to a permutation of length n − 1. Suppose we already have k ascents (there
are n−1

k such permutations); to add no ascents, n must either be placed at the beginning or in the
middle

n−1 of one of the k ascents, giving a = k + 1. Now suppose we have k − 1 ascents (there are
k−1 such permutations); to add one ascent we must place n either at the end or after one of the
(n − 1 − 1) − (k − 1) descents, giving b = n − k.

1
8. Answer: n2 + 2n
We first note that this is clearly a lower bound, since if we are able to do this without ever moving
backward, we must move each of the red pegs n + 1 and each of the blue pegs n + 1, but we save
moves since each peg will jump or be jumped by the n pegs of the other color, making a total of
2(n + 1)n − n2 = n2 + 2n moves. We see that we can achieve this lower bound by the following process.
First we move the red pegs as far as we can without moving one next to another one, then the same for
the blue pegs, then the red pegs again, and so on. (The first two steps move one red peg, then jump
with a blue and move the next blue up). After doing this n times (taking 1+2+· · ·+n moves), the pegs
will all be alternating, with the empty spot at one end. We can then make n consecutive jumps, ending
with the empty spot at the other end, then repeat the first algorithm for another n + (n − 1) + · · · + 1
moves to be done. The total is then 2(n + 1)n/2 + n = n2 + 2n.

9. Answer: p2 + p + 2
 
u v
Let a = b2 and let X = . The resulting equations are u2 + vy = yv + z 2 = a and (u + z)v =
y z
(u + z)y = 0. First suppose u + z = 0. The resulting solution is

(u, v, y) = (±b, 0, 0); (±b, 0, t), (w, t, t−1 (b2 − u2 ))

where t 6≡ 0 mod p and w 6≡ ±b. There are 2 + 2(p − 1) + 2(p − 1) + (p − 2)(p − 1) = p2 + p solutions
here. Otherwise u + z 6= 0 so v = y = 0 and u2 = z 2 = a, giving u, v = ±b, two more solutions.
10. Answer: 1007
First note that P (a) − P (b) is the sum of terms in the form Cn · (an − bn ) = Cn · (a − b)(an−1 +
an−2 · b + · · · + bn−1 ). So the sum is divisible by (a − b) and by another complex sum that is certainly
greater than 1 given that each C is a positive integer. So P (a) − P (b) can only be prime if a and b
differ by 1. Therefore, we cannot have more than 6 such pairs of elements in T . Clearly we could start
with 1, 3, 5, . . . , 2007 and add in any three even numbers since each would produce 2 pairs of elements
differing by 1.

2
 General Test
2007 Stanford Math Tournament
March 4, 2007
1. There are three bins: one with 30 apples, one with 30 oranges, and one with 15 of each. Each is labeled
“apples,” “oranges,” or “mixed.” Given that all three labels are wrong, how many pieces of fruit must
you look at to determine the correct labels?
2. Aliens from Lumix have one head and four legs, while those from Obscra have two heads and only one
leg. If 60 aliens attend a joint Lumix and Obscra interworld conference, and there are 129 legs present,
how many heads are there?
3. Mary puts one red and one blue marble into a box. In another box she places two red marbles. She
then forgets which box is which and randomly reaches into one of the boxes and takes out a red marble.
What is the probability that the other marble in that box is blue?
4. Evaluate (tan 10◦ )(tan 20◦ )(tan 30◦ )(tan 40◦ )(tan 50◦ )(tan 60◦ )(tan 70◦ )(tan 80◦ ).
5. Two disks of radius 1 are drawn so that each disk’s circumference passes through the center of the
other disk. What is the circumference of the region in which they overlap?
6. Team Stanford has a 13 chance of winning any given math contest. If Stanford competes in 4 contests
this quarter, what is the probability that the team will win at least once?
7. A boat is traveling upstream at 5 mph relative to the current flowing against it at 1 mph. If a tree
branch 10 miles upstream from the boat falls into the current of the river, how many hours does it
take to reach the boat?
8. Tina writes four letters to her friends Silas, Jessica, Katie, and Lekan. She prepares an envelope for
Silas, an envelope for Jessica, an envelope for Katie, and an envelope for Lekan. However, she puts
each letter into a random envelope. What is the probability that no one receives the letter they are
supposed to receive?
9. Peter Pan and Crocodile are each getting hired for a job. Peter wants to get paid 6.4 dollars daily, but
Crocodile demands to be paid 10 cents on day 1, 20 cents on day 2, 40 cents on day 3, 80 cents on day
4, and so on. After how many whole days will Crocodile’s total earnings exceed that of Peter’s?
10. Al, Betty and Clara are in the same class of 50 students total, but are not friends with each other. Al
is friends with 24 students, Betty is friends with 39, and Clara is friends with 20. What is the greatest
number of students that could be friends with all 3 of them?
11. Jonathan finds out that his ideal match is Sara Lark, but to improve his odds of finding a girlfriend,
he is willing to date any girl whose name is an anagram of “Sara Lark”, and whose name consists of
both a first and last name of at least one letter. How many such anagrams are there?
12. Pete has some trouble slicing a 20-inch (diameter) pizza. His first two cuts (from center to circumference
of the pizza) make a 30◦ slice. He continues making cuts until he has gone around the whole pizza,
each time trying to copy the angle of the previous slice but in fact adding 2◦ each time. That is, he
makes adjacent slices of 30◦ , 32◦ , 34◦ , and so on. What is the area of the smallest slice?
13. A rope of length 10 m is tied tautly from the top of a flagpole to the ground 6 m away from the base of
the pole. An ant crawls up the rope and its shadow moves at a rate of 30 cm/min. How many meters
above the ground is the ant after 5 minutes? (This takes place on the summer solstice on the Tropic
of Cancer so that the sun is directly overhead.)
14. Let there be 50 natural numbers ai such that 0 < a1 < a2 < · · · < a50 < 150. What is the greatest
possible sum of the differences dj , where each dj = aj+1 − aj ?

1
15. A number x is uniformly chosen on the interval [0, 1], and y is uniformly randomly chosen on [−1, 1].
Find the probability that x > y.
16. Find the area of a square inscribed in an equilateral triangle,
√ with one edge of the square on an edge
of the triangle, if the side length of the triangle is 2 + 3.
17. There is a test for the dangerous bifurcation virus that is 99% accurate. In other words, if someone has
the virus, there is a 99% chance that the test will be positive, and if someone does not have it, then
there is a 99% chance the test will be negative. Assume that exactly 1% of the general population has
the virus. Given an individual that has tested positive from this test, what is the probability that he
or she actually has the disease? Express your answer as a percentage.

18. A farmer wants to build a rectangular region, using a river as one side and some fencing as the other
three sides. He has 1200 feet of fence which he can arrange to different dimensions. He creates the
rectangular region with length L and width W to enclose the greatest area. Find L + W .
10
19. Arrange the following four numbers from smallest to largest: a = (10100 )10 , b = 10(10 )
, c = 1000000!,
d = (100!)10
20. Let there be 4n + 2 distinct paths in space with exactly 2n2 + 6n + 1 points at which exactly two of
the paths intersect. (A path never intersects itself.) What is the maximum number of points where
exactly three paths intersect?
21. Convert the following decimal to a common fraction in lowest terms: 0.92007200720072007 · · · (or
0.92007).
22. Katie begins juggling five balls. After every second elapses, there is a chance she will drop a ball. If
k
she is currently juggling k balls, this probability is 10 . Find the expected number of seconds until she
has dropped all the balls.
23. A quadrilateral has side lengths 3, 3, x, and y, where x and y are integers. We are allowed to choose x
arbitrarily, then we choose y. Let N be the number of possible integer values for y after x is chosen.
Find the greatest number of possible values for N .
24. Andy and Bob are playing a ping-pong match. Right now, Andy has 17 points and Bob has 18 points.
On any particular play, Andy has a 50% chance of winning, which gains him 1 point. If Andy loses,
Bob gains 1 point. The game ends when one player wins by getting 21 points. What is the probability
that Andy will win?
25. Nathan is standing on vertex A of triangle ABC, with AB = 3, BC = 5, and CA = 4. Nathan walks
according to the following plan: He moves along the altitude-to-the-hypotenuse until he reaches the
hypotenuse. He has now cut the original triangle into two triangles; he now walks along the altitude
to the hypotenuse of the larger one. He repeats this process forever. What is the total distance that
Nathan walks?

2
 General Solutions
2007 Stanford Math Tournament
March 4, 2007
1. Answer: 1
Pick up a piece of fruit in the “mixed” bin; say it is an apple. Then you know that this bin should be
labeled “apples,” and the bin currently labeled “apples” must be “oranges” (since you know that the
oranges label was incorrect). And so the current “oranges” should be “mixed.”

2. Answer: 97
If there are x Lumixians and y Obscrans, x + y = 60 and 4x + y = 129; solving gives x = 23 and
y = 37, so x + 27 = 97.
1
3. Answer: 3
Since there are three red marbles, drawing a red marble means that 2/3 of the time she has drawn
from the box with two red marbles, so only 1/3 of the time is the other marble blue.
4. Answer: 1
Note that sin(90◦ − θ) = cos θ and cos(90◦ − θ) = sin θ; writing out the tangents as sines divided by
cosines, we see that 1 = tan 10◦ tan 80◦ = tan 20◦ tan 70◦ = · · · .
4π
5. Answer: 3
The arc forming half the boundary of the intersection is of measure 120◦ , so the circumference is
2 · 120
360 · 2π.
65
6. Answer: 81
P (lose) = 1 − P (lose all 5) = 1 − ( 23 )4 = 65
81 .
5
7. Answer: 2
or 2.5
Essentially, the boat is traveling at 4 mph toward a stationary branch 10 miles away. It will take
10/4=2.5 hours to reach it.
3
8. Answer: 8
Let Silas receive a letter from Jessica: then there are 3 ways to arrange the remaining envelopes and
letters by letting Jessica receive letters from either Silas, Katie, or Lekan. The same is true if Silas
receives the letter from Katie or Lekan, so there are 3 · 3 = 9 total ways for no one to receive the correct
letter. There are a total of 4! ways to receive letters, so the probability we want is 9/4! = 3/8.

9. Answer: 10
0.1(2n −1)
After n days, Peter will have earned 6.4n dollars, while Crocodile will have earned 2−1 dollars.
0.1(2n −1) n−6 1 9−6
We want the n to satisfy 6.4n < 2−1 .
Simplifying: n < 2 − 64 . Since 9 > 2 = 8 but
10 < 210−6 = 16, so the smallest such n is 10.
10. Answer: 18
Draw a Venn diagram - we can put the most in the center if none are friends with just two of them.
Let x be the number of students friends with all 3; then 24 − x, 20 − x, and 39 − x are friends with
just Al, Betty, and Clara respectively. Adding up, x + (24 − x) + (20 − x) + (39 − x) = 47, since they
are 3 members of the class of 50, so x = 18.

1
11. Answer: 23520
There are 8! arrangements of the letters, but we divide by 3! and 2! since switching the a’s or the
r’s has no effect. There are then seven ways to split each of these into a first and last name, giving
8! · 7/(2!3!) = 8 · 7 · 7 · 5 · 4 · 3 possible names.
12. Answer: 5π
After nine slices, he has cut 9 · 30◦ + 0◦ + 2◦ + · · · + 16◦ = 270◦ + 18◦ · 4 = 342◦ , so the leftover slice is
18◦ , with area 100π · 18◦ /360◦ = 5π.
13. Answer: 2
The ratio of distances along the rope, ground (shadow), and above the ground is 10 : 6 : 8, so having
travelled 30 · 5 cm = 1.5m along the ground, it has travelled 8 · 1.5/6 m = 2 m above the ground.
14. Answer: 148

49
X
(ai+1 − ai ) = a50 − a1 ≤ 149 − 1
j=1

3
15. Answer: 4
We draw the rectangle from 0 to 1 on the x-axis and −1 to 1 on the y-axis; x > y is the region below
a line from the origin to (1, 1). This region has area 23 out of the total area of 2.
16. Answer: 3
Let s be the side length of the square. Looking at one of the 30-60-90 triangles outside the square
√ √ √
would give 2+ 23−s · 3 = s, so s = 3.
17. Answer: 50%
Let T +/− indicate the test result and B +/− indicate whether the person actually does or does not
have bifurcation virus. The probability that someone has the virus, given that their test is positive,
is equal to the probability that a given person tests positive and has it over the total probability of
testing positive. In statistical notation:
P (T + |B + ) · P (B + )
P (B + |T + ) =
· P (B + ) + P (T + |B − ) · P (B − )
P (T + |B + )
0.99 · 0.01
=
0.99 · 0.01 + (1 − 0.99) · (1 − 0.01)
= 1/2.

18. Answer: 900

Since L = 1200 − 2W , Area=W · L = W (1200 − 2W ) = −2W 2 + 1200W . This is a parabola, so its
vertex is at W = 1200/(2 · · · 2) = 300, giving L = 600.
19. Answer: a, d, c, b
We know d > a: 100! > 101 00 since each is a product 100 numbers but the first has larger ones. We
know c > d since 1000000! > (100000!)10 > (100!)1 0. Finally, b > c since 1000000! < 10000001000000 =
6 6
(106 )10 = 106·10 .
20. Answer: 2n2
There are a total of 4n+2


2 = 8n2 + 6n + 1 pairs of paths; subtracting the pairs which represent
intersections of only two lines, there are 6n2 pairs of paths left. Each point at which three paths
2
intersect accounts for three pairs, so there are at most 6n3 points at which exactly three paths intersect.

2
 5111
21. Answer: 5555

105 x = 92007.2007 and 10x = 9.2007

105 x − 10x = 92007.2007 − 9.2007 = 91998
91998 5111
x= =
99990 5555
137
22. Answer: 6
If an event happens with probablity p each second, the expected number of seconds for it to occur is
1/p (this is a definition of probability!). Thus adding up for each dropped ball, the expected time until
no balls are left is 10 10 10 10 10
5 + 4 + 3 + 2 + 1 = 10(
12+15+20+30+60
60 ) = 16 · 137 = 137
6 seconds.

23. Answer: 11
Note that in a quadrilateral ABCD we have AB + BC + CD > AD, since AB + BC > AC and
AC + CD > AD by the triangle inequality. This gives us three possible equations: 3 + x + y > 3 (this
will always be true), x + 6 > y, and y + 6 > x ⇒ y > x − 6. Thus x + 6 > y > x − 6, giving 11 possible
values from x − 5 to x + 5.
11
24. Answer: 32
For Andy (A) to win, he must win 4 points before Bob wins 3.

4
P (A wins on 4th point) = 12
1 4 1



P (A wins on 5th point) = (Ways to happen) · P (A wins 4) · P (A loses 1) = 4 · 2 · 2

4
2
P (A wins on 6th point) = 52 · 12 · 21



4
4
4
2
Sum: 12 + 4 · 12 · 12 + 52 · 12 · 12 = 22 11



64 = 32 .

25. Answer: 12
Let M be the endpoint of the altitude on the hypotenuse. Since we are dealing with right triangles,
∆M AC ∼ ∆ABC, so AM = 12/5. Let N be the endpoint he reaches on side AC. ∆M AC ∼ ∆N AM ,
so M N
AM = 4/5. This means
that each altitude that he walks gets shorter by a factor of 4/5. The total
distance is thus 12 4
5 / 1 − 5 = 12.

3
 Power Test
2007 Stanford Math Tournament
March 4, 2007
Definitions:
• Floor: bxc is the greatest integer less than or equal to x.
• Ceiling: dxe is the least integer greater than or equal to x.
• Fractional part: {x} = x − bxc.
• Intervals:
– Open: (α, β) = {α < x < β}
– Closed: [α, β] = {α ≤ x ≤ β}
– Half-open: [α, β) = {α ≤ x < β} and (α, β] = {α < x ≤ β} (called half-closed by pessimists)
In all problems, assume that x, y, α, β are real and m, n are integers. (If you define new variables in your
proofs please try to keep to this convention!) Note that for i < j you may use the result of problem i for
problem j even if you have not solved it.

1. Show that bxc = n if and only if n ≤ x < n + 1 and if and only if x − 1 < n ≤ x. Write similar
statements for ceilings (you needn’t prove them separately).
2. Show that b−xc = − dxe.
3. Show that x < n if and only if bxc < n. Write similar statements for n < x, x ≤ n, and n ≤ x (you
needn’t prove them separately).
4. Show that bn + xc = n + bxc, and write a similar statement for dn + xe (again, you needn’t prove it
separately).
5. Determine, with proof, under what conditions bnxc = n bxc.
6. How can we round, that is, find the nearest integer to x? We usually round up ties (when x is halfway
between integers), so give two formulas, one which rounds ties up and one which rounds them down.
7. Show that 2x+1
   2x+1   2x+1 
2 + 4 − 4 is either bxc or dxe, and when each is true.
 n   n+m−1 
8. Show that m = m when m > 0.
9. Find, with proof, forumulas for the number of integers contained in the half-open intervals [α, β) and
(α, β], assuming α ≤ β.
10. Show that bbmαc n/αc = mn − 1 where m, n > 0 and α > n is irrational.
11. Suppose f (x) is a continuous and increasing function such that if f (x) is an integer, x is an integer.
Show that bf (bxc)c = bf (x)c. What is a similar statement bf (x)c = ? if f is decreasing instead of
increasing? (The relevant property of continuous functions is that if f (x1 ) = y1 and f (x2 ) = y2 , then
f passes through all y-values between y1 and y2 at some point as x goes from x1 to x2 .)
12. The spectrum of a real number x is the sequence of integers Spec (x) = {bxc , b2xc , b3xc , . . .}. Show
that spectra are unique, i.e. that Spec (α) = Spec (β) if and only if α = β.
13. A casino has a roulette wheel with one thousand √ slots, numbered 1 to N 3 . If the number n that comes
up is divisible by the floor of its cube root (b nc | n), it’s a winner. Determine with proof the number
3

of winners.

1
 n
j 2 = 16 n(n + 1)(2n + 1)
P
14. Show that
j=0

√ P j√ k
n−1
15. Show that, if a = b nc, k = na − 31 a3 − 21 a2 − 16 a.
k=0

16. A circle, 2r = 2n − 1 units in diameter, is drawn centered at the center of a 2n × 2nPsquare grid. Show
n−1
that the circle passes through 8r cells of the grid, determine an f (n, k) such that k=1 f (n, k) is the
number of cells entirely contained inside the circle.

2
 Power Solutions
2007 Stanford Math Tournament
March 4, 2007
1. n ≤ x < n + 1 if and only if n is the greatest integer less than or equal to x. The second condition is
equivalent to the first since x − 1 < n ⇒ x < n + 1. The corresponding statements are dxe = n ⇐⇒
n − 1 < x ≤ n ⇐⇒ x ≤ n < x + 1.
2. From the first problem, we have b−xc = n ⇒ n ≤ −x < n + 1 ⇒ −n − 1 < x ≤ −n ⇒ −n = dxe.

3. Assume first that x < n. Then by problem 1, bxc ≤ x < n. Now assume bxc < n; by problem
1 we know x < bxc + 1, and since both are integers, bxc ≤ n. Similarly, n < x ⇐⇒ n < dxe,
x ≤ n ⇐⇒ dxe ≤ n, and n ≤ x ⇐⇒ n ≤ bxc.
4. Let m = bn + xc. Then m ≤ n + x < m + 1, so m − n ≤ x < m − n + 1, so bxc = m − n and thus
m = n + bxc. Similarly dn + xe = n + dxe.
5. We split x into floor and fractional part: bnxc = bn bxc + n {x}c = n bxc + bn {x}c. Thus for the two
to be equal, bn {x}c = 0 so 0 ≤ n {x} < 1, so {x} < 1/n.
6. To round up, take x + 21 . We see this works by splitting the inside into a floor and a fractional part;
 

if {x} < 1/2, adding 1/2 doesn’t change the floor, but if {x} ≥ 1/2, adding 1/2 increases the floor by
1. A similar argument gives x − 21 for rounding down.
 

2x+1
7. 2 = x + 21 , so the first term rounds looks like our rounding formula, except the result is always
one too high except when x + 1/2 is an integer, in which case it correcly rounds up. Now notice that
dαe − bαc is 0 if α is an integer and 1 otherwise, so the next two terms subtract 1 if 2x+1
4 = x+1/2
2 is
not an integer. Thus the other terms correct the first term to the correctly rounded value when x + 1/2
is not an integer. When x + 1/2 is an integer, the other terms leave the first term alone if it’s an even
one, but subtract one if it’s odd. Thus the formula always rounds x to the nearest integer, rounding
halves up or down when x + 1/2 is even or odd.
n n
8. Let k = m . We have k − 1 < m ≤ k. Since m−1
m < 1,
n+m−1
m <k+ 1. Since
 n, m are integers, and
n n 1 n+m−1
 n+m−1
m > k − 1, we know that m ≥ k − 1 + m , so m > k. Thus k = m .
9. First note that if α and β are integers, the answer in both cases is β − α. Let n be an integer in [α, β);
by problem 3 we have that dαe ≤ n < dβe, so the number of integers in the interval is dβe − dαe.
Similarly, n ∈ (α, β] implies bαc < n ≤ bβc, giving bβc − bαc.

10. Since α is irrational, we know 0 < {mα} < 1, and also n/α < 1. Plugging in bmαc = mα − {mα}, we
obtain bmαn/α − {mα} n/αc = bmn − {mα} n/αc = mn − 1.
11. If bxc = x, we are done; otherwise, bxc < x. Thus f (bxc) < f (x) since f is increasing, and so
bf (bxc)c ≤ bf (x)c. If bf (bxc)c < bf (x)c, since f is continuous there must be a number y such that
bxc ≤ y < x and f (y) = bf (x)c. By the special property of f , this means y is an integer, but there can
be no integer between x and its floor! Thus we must have bf (bxc)c = bf (x)c. Similarly, for decreasing
f , bf (x)c = bf (dxe)c.
12. (Proof by contrapositive) Suppose α 6= β, and assume without loss of generality that α < β. Then
there must be a positive integer m such that m(β − α) ≥ 1. Thus mβ − mα ≥ 1 so bmβc > bmαc, so
the mth elements of the spectra are different.
√
13. Suppose n is a winner; let k = b 3 nc. Then k 3 ≤ n < (k + 1)3 and n = km for some m. Note that N 3
is a winner; let’s assume n < N 3 , so that 1 ≤ k < N . Now substituting km for n, k 3 ≤ km < (k + 1)3
so k 2 ≤ m < (k + 1)3 /m. Using our formula for the number of integers in a half-open interval,

1
      
there are (k + 1)3 /k − k 2 = k 2 + 3k + 3 + 1/k − k 2 = 3k + 4 of these. We then simply sum
this for the possible values of k (it’s an arithmetic series), and add back in the n = 1000 case to get
1 + 4(N − 1) + 23 (N − 1)N = 21 (3N 2 + 5N − 6).
14. A proof by induction is quickest (though not the most general or elegant). The statement is true for
n = 0, and starting from n and moving up to n + 1:
 2 
1 2 n n
n(n + 1)(2n + 1) + (n + 1) = (n + 1) + +n+1
6 3 6
1 2
= (n + 1)(2n + 7n + 6)
6
1
= (n + 1)(n + 2)(2n + 3)
6

15. Note that the terms for a2 ≤ k < n are all equal to a, so they √ 2
contribute (n − a√ )a to the sum.
2
We now consider the rest of the sum, 0 ≤ k < a . Let m = b kc; then m ≤ k < m + 1 so
m2 ≤ k < (m + 1)2 ≤ a2 . We sum over k first instead of m; there are (m + 1)2 − m2 possible values of
k, so our new sum is:

a−1 a−1
X X 1 1
m((m + 1)2 − m2 = m(2m + 1) = 2 (a − 1)a(2a − 1) + a(a − 1)
m=0 m=0
6 2

2a3 a2
Expanding, we have 3 − 2 − a6 ; adding in the k ≥ a2 terms, we obtain the desired result.
16. There are 2n − 1 each of horizontal lines vertical lines between cells of the grid, and the circle crosses
each one twice. Since r2 is not an integer, the circle cannot pass through the corner of any cell, by
the Pythagorean theorem. Thus the circle passes through a cell for each time it crosses a line, giving
4(2n − 1) = 8n − 4 = 8r cells. f (n, k) = 4 r2 − k 2 : consider f (n, k)/4; placing the x, y axes along the
grid with origin at the center we can easily see from the equation of a circle that this is the number of
cells above x = k within the circle.

2
 Team Test
2007 Stanford Math Tournament
March 4, 2007
1. How many rational solutions for x are there to the equation x4 +(2−p)x3 +(2−2p)x2 +(1−2p)x−p = 0
if p is a prime number?
2. If a and b are each randomly and independently chosen in the interval [−1, 1], what is the probability
that |a| + |b| < 1?
3. A clock currently shows the time 10:10. The obtuse angle between the hands measures x degrees.
What is the next time that the angle between the hands will be x degrees? Round your answer to the
nearest minute.
4. What is the area of the smallest triangle with all side lengths rational and all vertices lattice points?
5. How many five-letter “words” can you spell using the letters S, I, and T, if a “word” is defined as any
sequence of letters that does not contain three consecutive consonants?
2007 
P
6. x ≡ k mod 2016, where 0 ≤ x ≤ 2015. Solve for x.
k=1

7. Daniel counts the number of ways he can form a positive integer using the digits 1, 2, 2, 3, and 4
(each digit at most once). Edward counts the number of ways you can use the letters in the word
“BANANAS” to form a six-letter word (it doesn’t have to be a real English word). Fernando counts
the number of ways you can distribute nine identical pieces of candy to three children. By their powers
combined, they add each of their values to form the number that represents the meaning of life. What
is the meaning of life? (Hint: it’s not 42.)
8. A 13-foot-tall extraterrestrial is standing on a very small spherical planet with radius 156 feet. It sees
an ant crawling along the horizon. If the ant circles the extraterrestrial once, always staying on the
horizon, how far will it travel (in feet)?
9. Let dn denote the number of derangements of the integers 1, 2, . . . , n so that no integer i is in the ith
position. It is possible to write a recurrence relation dn = f (n)dn−1 + g(n)dn−2 ; what is f (n) + g(n)?
10. A nondegenerate rhombus has side length l, and its area is twice that of its inscribed circle. Find the
radius of the inscribed circle.
11. The polynomial R(x) is the remainder upon dividing x2007 by x2 − 5x + 6. R(0) can be expressed as
ab(ac − bc ). Find a + c − b.
12. Brownian motion (for example, pollen grains in water randomly pushed by collisions from water mole-
cules) simplified to one dimension and beginning at the origin has several interesting properties. If
B(t) denotes the position of the particle at time t, the average of B(t) is x = 0, but the average of
B(t)2 is t, and these properties of course still hold if we move the space and time origins (x = 0 and
t = 0) to a later position and time of the particle (past and future are independent). What is the
average of the product B(t)B(s)?
13. Mary Jane and Rachel are playing ping pong. Rachel has a 7/8 chance of returning any shot, and
Mary Jane has a 5/8 chance. Mary Jane serves to Rachel (and doesn’t mess up the serve). What is
the average number of returns made?
p
14. Let p, q be positive integers and let x0 = 0. Suppose xn+1 = xn + p + q 2 + 4pxn . Find an explicit
formula for xn .
R∞ tan−1 (πx)−tan−1 x
15. x dx
0
 Team Solutions
2007 Stanford Math Tournament
March 4, 2007
1. Answer: 2
By the Rational Roots Theorem, the roots must be ±1 or ±p. Check all of these; only −1 and p work.
1
2. Answer: 2
In the ab plane this is the square with vertices (1, 0), (0, 1), (−1, 0), (0, −1), which has area 2, while the
total range of possible values is a 2 × 2 square.
3. Answer: 10:34
Each angle between consecutive numbers on a standard clock is 30◦ . So each minute, the hour hand
moves 30/60 = (1/2)◦ , and the minute hand moves 30/5 = 6◦ . Letting the 12 o’clock position be 0, at
10:10, the small angle difference between the two hands is 360−(300+1/2·10)+60 = 115. In m minutes,
the angle between the hand positions will be 115 + 6m − 21 m. The angle will actually be the same when
this is 360 − 115 = 245; solving for m gives Set those equal, and solve for m = 260/11 = 23 + 7/11;
adding back up gives the answer.
4. Answer: 6
The triangle must have at least one side not in the direction of axes, and since {3, 4, 5} is the smallest
Pythagorean triplet, the smallest length for this side is 5, and we can minimize the area by making the
remaining two sides 3 and 4.
5. Answer: 123
Look at how many five-letter non-words you can make. Let C be a consonant and let V be a vowel. The
possible patterns are: CCCV C, CV CCC, CCCCV , V CCCC (16 words each); CCCV V , V CCCV ,
V V CCC (8 words each); CCCCC (32 words). Thus there are 120 non-words. There are 35 = 243
sequences, so there are 243 − 120 = 123 words.
6. Answer: 1044
2015 
P
First consider y ≡ k mod 2016: the sum of the first and last term, of the second and next to
k=1
last, and so on is always 2016. However, there will be one term left in the middle, 1008. Thus y ≡ 1008
mod 2016. We have added in 2015 + 2014 + · · · + 2008 ≡ (−1) + (−2) + · · · + (−8) ≡ −36 mod 2016,
so finally x ≡ 1008 − (−36) ≡ 1044 mod 2016.
7. Answer: 645
5!
Daniel: There are 2! integers using all the digits. Dropping one, in 3 cases (like 1,2,2,3) there are 4!
2!
integers and (dropping a 2) there are 4!. Dropping 2, there 3! integers in 4 cases (like 1,2,3) and 3! 2!
in three cases (like 1,2,2). Dropping 3, there are 2! integers in 4 cases (like 1,2) and also 22. Finally,
6!
there are 4 one-digit numbers. Adding up gives 170. For Edward, there are 2!3! words if he drops a B
6! 6!
or S, 3! if he drops an N, and 2!2! if he drops an A, totalling to 420. Finally, Fernando can distribute
candy
by putting the nine pieces in a line and drawing two lines to divide it into three piles. There
are 102 ways to do this without drawing the two lines in the same place, and another 10 with the two
lines in the same place.
8. Answer: 120π
p √
First we have that the distance to the horizon is (13 + 156)2 − 1562 = 13 132 − 122 = 65. Let r be
the length of an altitude to hypotenuse of 65-156-169 triangle; this is the radius of the ant’s circular
path. By similar triangles we have r = 65·156 5·13 12·13
169 = 13 · 13 = 5 · 12. Path length = 2πr = 120π.

1
9. Answer: 2n − 2
To write a derangement of length n, we first pick i from among the n − 1 possibilities for the first
element. If we place 1 in the ith position, there are dn−2 ways left to derange the remaining n − 2
integers, since their positions are all left. There are dn−1 ways to place 1 somewhere other than the
ith position and derange the remaining, since we might as well rename 1 to i in this case. Thus the
recurrence relation is dn = (n − 1)(dn−1 + dn−2 ).
l
10. Answer: π
For any polygon with an inscribed circle of radius r, the area is A = 12 rP , where P is the perimeter.
This is true since we can cut the polygon up into right triangles, each with one leg of length r and the
other leg a part of the perimeter; adding up the areas of all the triangles and factoring out the latter
lengths we obtain A = 12 rP . Thus A = 12 r · 4l = 2πr2 ⇒ rl = πr2 ⇒ r = πl .
11. Answer: 2005
The degree of R cannot be greater than the degree of the denominator polynomial, so we can write
R(x) = ax + b. x2007 = Q(x) · (x2 − 5x + 6) + rx + s, where Q is some polynomial. Set x = 2 and x = 3
successively to get x2006 = 2r+s and 32006 = 3r+s. Solving for s, we get R(0) = s = 2·3·(22006 −32006 ).
12. Answer: min{t, s}
Suppose without loss of generality s < t. Then by independence of past and future, the average of
(B(t) − B(s))2 = t − s. But (B(t) − B(s))2 = B(t)2 − 2B(t)B(s) + B(s)2 . Averaging both sides and
plugging in for the averages of squares, we have t − s = t − 2B(t)B(s) + s, so the desired average is s.
This is of course true if t = s as well.
91
13. Answer: 29
7
Let p = 8 and q = 58 . For the volley to end after an odd number of returns, the probability is
pqpq · · · pqp(1 − q) = (pq)(n−1)/2 p(1 − q), and for it to end after an even number of returns it is
pqpq · · · pq(1 − p)
P∞ = (pq)n/2 (1 − p). Writing
P∞ n = 2k + 1 and n = 2k for odd and even, the desired
average is then k=0 2k(1 − p)(pq)k + k=0 (2k + 1)p(1 − q)(pq)k . To evaluate these, we try:

∞
X
npn−1 = 1 + 2p + 3p2 + · · ·
n=0
= (1 + p + p2 + · · · ) + (p + p2 + · · · ) + (p2 + · · · ) + · · ·
1
= (1 + p + p2 + · · · )(1 + p + p2 + · · · ) =
(1 − p)2
p(1+q)
Applying this, multiplying back in the other constants, and simplifying gives 1−pq , and then we just
plug in for p, q.
Answer: pn2 + qn
p
First we solve the precurrence relation for xn , obtaining xn = xn+1 + p − q 2 + 4pxn+1 . We rewrite this
as xn−1 = xn +p− q 2 + 4pxn and add to the original recurrence relation to get xn +1−2xn +xn−1 = 2p.
Let x0n = xn+1 − xn . Thus x0n − x0n−1 = 2p, and since x00 = x1 − x0 = p + q, x0n = 2pn + (p + q).
Since the first difference of the sequence is linear, the next is constant and so we have a quadratic. The
second difference 2p is twice the coefficient of n2 , and considering x0 and x1 yields the coefficient of n
and the constant, giving the desired solution.
π
14. Answer: 2
log π
R ∞ tan−1 (ax)−tan−1 x R∞ ∞
dI 1 1
tan−1 x0 = π dI
R
Let I(a) = 0 x dx. Then da = 0 1+(ax)2 dx = a 2a . I(a) = da da =
π
2 log a + C, but clearly I(1) = 0 so C = 0.

2
SMT 2008 Algebra Test February 23, 2008

1. Reid is twice as old as Gabe. Four years ago, Gabe was twice as old as Dani. In 10 years, Reid will be
twice as old as Dani. How many years old is Reid now?
2. Let P (x) = x6 + ax5 + bx4 + x3 + bx2 + ax + 1. Given that 1 is a root of P (x) = 0 and −1 is not, what
is the maximum number of distinct real roots that P could have?

3. If a, b, c ∈ C and a + b + c = ab + bc + ac = abc = 1, find a, b, c. (The order in which you write your

answers does not matter.)
4. Find x4 + y 4 + z 4 , given that (
0=x+y+z
1 = x2 + y z + z 2

5. The product of a 13x5 matrix and a 5x13 matrix contains the entry x in exactly two places. If D(x) is
the determinant of the matrix product, D(x = 0) = 2008, D(x = −1) = 1950, and D(x = 2) = 2142.
Find D(x).
6. For how many integers k, with 0 ≤ k ≤ 2008, does x2 − x − k = 0 have integer solutions for x?
7. Find all ordered pairs of positive integers (p, q) such that 2p2 + q 2 = 4608.

8. How many monic polynomials P (x) are there with P (x)Q(x) = x4 − 1 for some other polynomial Q(x),
where the coefficients of P and Q are in C?
9. Find the number of distinct ordered integer pairs (x,y) with x + y − xy = 43.
10. Evaluate
∞
X k
.
5k
k=1
SMT 2008 Algebra Solutions February 23, 2008

1. Answer: 28

 r = 2g
g − 4 = 2(d − 4)
r + 10 = 2(d + 10)


So r = 2g = 2(2(d − 4) + 4) = 4d − 8 = 4(r/2 − 5) − 8 = 2r − 28 ⇒ r = 28.

2. Answer: 5
Divide the equation P (x) = 0 by x3 to get x3 + ax2 + bx + 1 + b x1 + a x12 + x13 = 0. In this equation,
replacing x by x1 doesn’t change anything, so anytime x is a root of P (x) = 0, x1 is also a root. Any
root other than ±1 (since 0 isn’t a root) must be paired with another, namely its reciprocal. And 1
is a root, while −1 is not. So the total number of real roots must be odd. Note that having an odd
number of distinct real roots requires that 1 be a double root. This makes the maximum number of
real roots 5.
3. Answer: {i, −i, 1}
First, notice that the three solutions are symmetric. We write our conditions as a system of equations:

 a+b+c=1 (1)
ab + bc + ca = 1 (2)
abc = 1 (3)


(3) can be rewritten c = 1/ab. Substituting that in (2), we get

b a
ab + ab + ab = 1
ab + 1/a + 1/b = 1
a2 b + 1 + a/b = 1
a(ab + 1/b) = 0
Because we know from (3) that a = 0 cannot be a solution, we throw it out:
ab + 1/b = 0
a = −1/b2

Substituting this as well as our expression for c in (1), we get:

−1 1
b2 +b+ −1/b2 (b) = 1
−1
b2 +b−b = 1
−1
b2 = 1
b = ±i

Letting any two variables be −i and i, we easily find using any of our three equations that the third
must equal 1.
1
4. Answer: 2

0 = (x + y + z)2 = x2 + y 2 + z 2 + 2(xy + xz + yz)

−1
= xy + yz + xz
2
1
= (xy + yz + xz)2 = x2 y 2 + x2 z 2 + y 2 z 2 + 2(x2 yz + xy 2 z + xyz 2 )
4
= x2 y 2 + x2 z 2 + y 2 z 2 + 2xyz(x + y + z) = x2 y 2 + x2 z 2 + y 2 z 2
1 = (x2 + y 2 + z 2 )2 = x4 + y 4 + z 4 + 2(x2 y + x2 z 2 + y 2 z 2 )
1
1 = x4 + y 4 + z 4 + 2 ·
4
SMT 2008 Algebra Solutions February 23, 2008

5. Answer: 3x2 + 61x + 2008

The highest power of x that can occur in the determinant is x2 , so D(x) must be quadratic; let it be
ax2 + bx + c. The constant term is c = D(0) = 2008, so we have D(−1) − 2008 = −58 = a − b and
D(2) − 2008 = 134 = 4a + 2b. Solving the pair of linear equations gives a = 3 and b = 61.

6. Answer: 45
By the quadratic formula, the solutions to x2 − x − k = 0 are precisely
√
1 ± 1 + 4k
.
2
√ √
√ when 1 ± 1 + 4k is an even integer, i.e. when 1 + 4k is an odd
These solutions are integers precisely
integer. Since 1 + 4k is itself odd, 1 + 4k is an odd integer precisely when 1 + 4k is a perfect square.
Thus, we are interested in how many (nonnegative, to avoid double counting) integers a give an integer
solution for k with 0 ≤ k ≤ 2008 in 1 + 4k = a2 , or equivalently to 4k = a2 − 1. Notice that a2 − 1 is
divisible
√ by 4 precisely when a is odd. The only other restriction on a is that 4 · 2008 ≥ a2 − 1. Since
90
89 < 4 · 2008 + 1 < 90, there are 2 = 45 values for a such that 4k = a2 − 1 has an integer solution
for k with 0 ≤ k ≤ 2008. Consequently, there are 45 values for k such that x2 − x − k = 0 has integer
solutions for k.
7. Answer: (16, 64)

2
Since 2p2 + q 2 is even, q must be even, so we divide through by 2 to obtain p2 + 2 2q = 2304. Now,
p must be even, so we divide through by 2 again. Repeating until the number on the right is no longer
p 2 q 2 p q



even, we find that 16 + 2 32 = 9, where 16 and 32 are integers. This has the obvious solution
12 + 2 · 22 = 9, which gives (p, q) = (16, 64).

8. Answer: 16

P (x)Q(x) = x4 − 1
= (x2 − 1)(x2 + 1)
= (x − 1)(x + 1)(x − i)(x + i)
4 4 4 4 4






We have four distinct monomial factors, so the number of possible P (x) is 0 + 1 + 2 + 3 + 4 = 16.
9. Answer: 16
Solving the equation for y gives y = x−43
x−1 . Essentially, we are now finding all z = x − 1 s.t. z|z − 42 ⇒
z|42, since z|z. The possible values for z are {±1, ±2, ±3, ±6, ±7, ±14, ±21, ±42}. There is one pair
(x, y) for each of these.
5
10. Answer: 16
P∞ k
Let S = k=1 5k
. Then,
∞ ∞ ∞ ∞ ∞
X k X k X k+1 X k X 1
5S = 5 = = = +
5k 5k−1 5k 5k 5k
k=1 k=1 k=0 k=0 k=0
∞ ∞ ∞
X k X 1 X 1
=0+ k
+ k
=S+
5 5 5k
k=1 k=0 k=0
∞
X 1 1 5
4S = = 1 =
5k 1 − 5
4
k=0
5
S=
16
SMT 2008 Geometry Test February 23, 2008

1. A regular polygon of side length 1 has the property that if regular pentagons of side length 1 are placed
on each side, then each pentagon shares a side with the two adjacent ones. How many sides does such
a polygon have?
2. John stands against one wall of a square room with walls of length 4 meters each. He kicks a frictionless,
perfectly elastic ball in such a way that it bounces off the three other walls once each and returns to
him (diagram not geometrically accurate). How many meters does the ball travel?

3. A cube is inscribed in a sphere of radius r. Find the ratio of the volume of the cube to that of the
sphere.
4. A circle of radius 144 has three smaller circles inside it, all congruent. Each small circle is tangent to
the other two and to the large circle. Find the radius of one of the smaller circles.
√ √
5. In 4ABC, ∠C is right, AC = 2 − 3 + x and BC = 1 − 2x + x 3. Find m∠B.
6. Points A, B, C lie on sides DE, EF , and F E of 4DEF , respectively. If DA = 3, AE = 2, EB = 2,
BF = 11, F C = 11, and CD = 1, find the area of 4ABC.
7. What is the area of the incircle of a triangle with side lengths 10040, 6024, and 8032?

8. Rhombus ABCD has side length l, with cos(m∠B) = − 23 . The circle through points A, B, and D has
radius 1. Find l.
9. A trapezoid has bases of length 10 and 15. Find the length of the segment that stretches from one leg
of the trapezoid to the other, parallel to the bases, through the intersection point of the diagonals.

10. A regular polygon with 40 sides, all of length 1, is divided into triangles, with each vertex of each
triangle being a vertex of the original polygon. Let A be the area of the smallest triangle. What is the
minimum number of square root signs needed to express the exact value of A?
SMT 2008 Geometry Solutions February 23, 2008

1. Answer: 10
Label three consecutive vertices of the polygon A, B, and C. Let BP be the common side to the
pentagons placed on sides AB and BC. Then m∠ABP = m∠P BC = 108◦ . Since m∠ABP +
m∠P BC + m∠ABC = 360◦ , this gives m∠ABC = 144◦ . So the exterior angle of this polygon is 36,
and thus it has 10 sides.
√
2. Answer: 8 2
Notice that the ball travels the length of the room twice and the width of the room twice, so it’s
traveled a total of 8 meters in the horizontal direction and 8 meters in the vertical direction. Because
the ball is bouncing (and thus its path after a bounce is the same as its path before the bounce, but
reflected), we can rearrange the four segments of its path into a straight line by only reflection and
translation. This line travels 8 meters
√ horizontally and 8 meters vertically, so its length, which is the
total length of the ball’s path, is 8 2.
√
3π
3. Answer: 2
√
Since the space diagonal of the cube is a diameter of the sphere, we have s 3 = 2r. The ratio is then
4 3
√
3 πr 3π
 3 =
2r
√
2
3

√
4. Answer: 288 3 − 432
Let r be the radius of a small circle. The centers of the small√circles form an equilateral triangle of
side length 2r. The length of the median of such a triangle is 3r, so the √
distance from the center of
the triangle (which is also the center of the large circle) to a vertex is 2 3 3 r. Since each vertex of the
√
triangle is distance r from the edge of the large circle, the radius of the large circle is 2 3 3 r + r = 144.
√ √
2√3−3
√ √
This gives (2 3 + 3)r = 432, so r = 2√432 ·
3+3 2 3−3
= 144(2 3 − 3) = 288 3 − 432.

5. Answer: 15◦ + tan−1 x or π

12
+ tan−1 x
√
2− 3+x
From basic trigonometry, we have tan(m∠B) = 1−(2− √
3)x
. This is the tangent angle addition identity,
√ ◦
√
for angles with tangents x and 2 − 3. Since tan(15 ) = 2 − 3, m∠B, the inverse tangent, is therefore
15◦ + tan−1 x.
44
6. Answer: 13
5
Let α = m∠E and β = m∠F . Note that D is a right angle. Therefore, sin α = 13 . [CBF ] =
1 1 5 1 1 1 12
2 · 11 2
sin α = 2 · 121 · 13 . Similarly, [ABE] = 2 · 22
sin β = 2 · 4 · cos α = 2 · 4 · 13 . Finally, [ACD] = 3·1
2 .
Subtracting these three areas from that of 4DEF gives the result.
7. Answer: 20082 π or 4032064π
Note that we can scale the triangle down by a factor of 2008 to a 3,4,5 right triangle. Let AB, AC be
the legs of the triangle. The incircle splits AB into two segments of lengths x and y. It similarly splits
AC into segments of lengths x and z and BC into segments of lengths y and z. Thus, we get:

x+y =3
x+z =4
y+z =5

Thus, x = 1, y = 2, z = 3. Thus, the incircle has a radius of 1, and so an area of π. Scaling back up
will increase the incircle’s radius by a factor of 2008, giving us an area of 20082 π.
SMT 2008 Geometry Solutions February 23, 2008

√
30
8. Answer: 3
Let O be the center of the circle, and
X be the center of the rhombus (the intersection of AC and
BD). Let m∠ABC = θ = cos−1 − 23 . Considering 4OBX and 4ABX, using triangle angle sums
and the fact that an inscribed angle has half the measure of the intercepted arc, we have OX =
cos(π − θ), so AX = 1 +q cos(π − θ). Also, BX = sin(π − θ). The Pythagorean theorem then gives
p
l = 2(1 + cos(π − θ) = 2(1 + 32 ).

9. Answer: 12
Let the trapezoid be ABCD with AB = 10, CD = 15. Let P be the intersection of the diagonals,
and let XY be the segment through P parallel to the bases with X on AD and Y on BC. Note that
4P Y C ∼ 4ABC, so PAB Y
= YBC
C
. Also, 4P Y B ∼ 4DCB, so CD PY
= BY
BC . Adding these equations gives
PY PY BY +Y C 1 1 1
AB + CD = BC = 1, so P Y ( 10 + 15 ) = P Y · 6 = 1, hence P Y = 6.
The same argument shows that P X = 6, so XY = 12.
10. Answer: 3
The polygon has angles of 171◦ , and the smallest triangle has two adjacent sides of the original polygon
as two of its sides. The area of this triangle is 21 · 1 · 1 · sin(171) = 21 sin(9). So the question is,
√
how many square roots do we need to express sin(9)? Conveniently enough, sin(18) = 5−1 4 , so
q
2
cos(18) = 1 − sin (18), which requires two square roots to express. Then by the half-angle formula,
q
sin(9) = 1−cos(18)
2 , which requires three square roots.
SMT 2008 Calculus Test February 23, 2008

Z π/2
1. Compute sin x cos x dx.
0

2. Evaluate:
10x2 − 12 x3
lim 1 2
x→0 e3x − 1
3. Find the area enclosed by the graph given by the parametric equations

y = sin(2t)
x = sin(t)

4. Find the value of the nth derivative of f (x) = sinn (x) at x = 0.

5. Water flows into a tank at 3 gallons per minute. The tank initially contains 100 gallons of water, with
50 pounds of salt. The tank is well-mixed, and drains at a rate of 2 gallons per minute. How many
pounds of salt are left after one hour?
6. Evaluate e3x sin(x)dx.
R

∞
X 2n−1
7. Compute .
n=0
n!

h2 3
8. Find f (x) such that lim = − x2 − x − 1
2x .
h→0 f (x+2x)−2f (x+h)+f (x)

9. Suppose x00 (t) + x0 (t) = t5 x(t). Let the power series representation of x be x(t) = an tn . Find an in
P
terms of an−1 and an−7 , where n > 7.
10. Evaluate: Z x
t2t et dt
−∞
SMT 2008 Calculus Solutions February 23, 2008

1
1. Answer: 2
Using the product to sum trigonometric identity, we get:
Z π/2 Z π/2
1
sin x cos x dx = (sin 2x − sin 0) dx
0 0Z 2
π/2
1
= 2 sin 2x dx
0
π/2
= 12 − 12 cos 2x 0
= 21 (1/2 − (−1/2))
= 12 (1)
= 1/2

2. Answer: 30
Two applications of L’Hôpital’s rule are necessary to obtain a fraction which does not divide by zero;
the result is
20 − 3x 20 − 0
2 x2 4 2 x2 = 2 = 30
3 e + 9 x e 3 +0

8
3. Answer: 3
The graph is a figure-8. By symmetry, the area it encloses is four times the area under the curve√that we
get from just considering 0 ≤ t ≤ π2 . For these values of t, we can write y = 2 sin(t) cos(t) = 2x 1 − x2
for 0 ≤ x ≤ 1. But this is integrable easily enough by letting u = 1 − x2 . The antiderivative that we
3
get is − 32 u 2 , which gives an area of 23 . Thus the total area enclosed by the parametric graph is 83 .
4. Answer: n!
You can think of taking the derivatives this way, in an extended version of the product rule: For each
j ≤ n, let fj (x) = sin(x). Then f (x) = f1 (x)f2 (x)...fn (x). Each time you take the derivative, you
pick one of the fj ’s and replace it with its derivative. Some fj ’s might be differentiated more than
once in this process. To find the derivative of f , take the results from each way of choosing the fj ’s to
differentiate and add them all up.
That said, since sin(0) = 0, if any fj doesn’t get differentiated, it’ll make the whole product zero
when x = 0, so these terms don’t matter. But we’re only taking the nth derivative, so if every fj
gets differentiated at least once, then they all have to be differentiated exactly once. There are n!
remaining terms (the number of ways to choose what order we differentiate the fj ’s in), each of which
is cosn (0) = 1, so the value we want is just n!.
125
5. Answer: 4
Let w(t) and s(t) denote the amounts of water and salt, respectively, in the tank at time t. We can
immediately see that w(t) = t + 100. Since the tank is constantly mixed, we know that

ds s(t)
=−
dt w(t)
ds dt
=−
s t + 100
ln(s) = − ln(C(t + 100))
C
s=
t + 100
Since s(0) = 50, C = 5000, so s(60) = 125/4.
1 3x
6. Answer: 10
e (3 sin(x) − cos(x))
SMT 2008 Calculus Solutions February 23, 2008

Let I be the answer. Integrating by parts twice:

Z
1 3x 1 3x
I = e sin(x) − e cos(x)dx
3 3
1 1 1
= e3x sin(x) − e3x cos(x) − I
3  9 9
9 1 3x 1 3x
I= e sin(x) − e cos(x)
10 3 9

1 2
7. Answer: 2
e
∞
X 2n
Multiply the sum by 2 and you get .
n=0
n!
Now notice that the nth derivative of e2x is 2n e2x , which gives 2n evaluated at x = 0. The Taylor
∞
X 2n n
series for e2x around x = 0 is thus x . Evaluated at x = 1, this gives the sum above, so the sum
n=0
n!
we want is equal to 21 e2 .

8. Answer: tan−1 x or tan−1 (x) + ax + b

1
The limit is, by definition, f 00 (x) . Therefore:

1 x4 + 2x2 + 1 (1 + x2 )2
= − = −
f 00 (x) 2x 2x
2x
f 00 (x) = −
(1 + x2 )2
1
f 0 (x) =
1 + x2
f (x) = tan−1 x

an−7 −(n−1)an−1
9. Answer: n(n−1)
Plugging the power series into the differential equation gives:
X X X
n(n − 1)an tn−2 + nan tn−1 = an tn+5
X X X
n(n − 1)an tn−2 + (n − 1)an−1 tn−2 = an−7 tn−2
n(n − 1)an tn−2 + (n − 1)an−1 tn−2 = an−7 tn−2
n(n − 1)an + (n − 1)an−1 = an−7

x(2e)x (2e)x )
10. Answer: 1+ln 2
− (1+ln 2)2

Z x  x Z 0
1 1
t2t et dt = t (2e)t − (2e)t dt
−∞ ln(2e) −∞ −∞ ln 2e
x
x(2e)x (2e)t

= −
1 + ln 2 (1 + ln 2) −∞
x(2e)x (2e)x
= −
1 + ln 2 (1 + ln 2)2
SMT 2008 Advanced Topics Test February 23, 2008

1. In ancient Roman games, gladiators were pitted against velociraptors in turn-based combat. In a stan-
dard match, one velociraptor fought n gladiators. The raptor attacked first, and hit one gladiator per
turn, killing him instantly. One gladiator attacked per turn, hitting the speedy raptor with probability
1
2 . If it took ten hits to kill the raptor, what is the smallest number of gladiators the emperor could
have sent into the arena in order to have at least 12 probability of killing the velociraptor?

2. In ancient Cartesia, city blocks were perfect squares, all with the same side length. East-west streets
were named First, Second, Third, and so on, while north-south streets were named Abel, Bernoulli,
Cauchy, and so on. Carrier pterodactyls were used to deliver large packages, since they could always fly
the shortest distance between two points. If a pterodactyl picked up two packages at First and Abel,
then dropped the first out of the sky onto Fourth and Erdős as it turned toward its next destination
at Eighth, by what angle did it turn above the first drop-off?

3. Two numbers are chosen randomly from the interval [0, 1]. What is the probability that they differ by
more than their average?
∞ ∞
1 π2 1
P P
4. Given that n2 = 6 , evaluate n2 +8n+16 .
n=1 n=1

5. Three X’s, three Y’s, and three Z’s are placed randomly in a 3 × 3 grid. What is the probability that
no row or column will contain two of the same letter?
6. A function f maps every sequence of integers to another sequence of integers as follows:
(
1 if n = 0
(f (a))n =
nan−1 6 0
if n =

If the sequence bn is a fixed point of f , what is b(2008)?

7. Lord Voldemort only does two things all day: curse Muggles, and kick puppies. Each Muggle he curses
has a 50% chance of dying while a puppy kick is always successful. Each dead Muggle gives him 3
units of satisfaction and each kicked puppy gives him 2 units. If an even number of Muggles die, he
doubles his satisfaction from each of them. If he can curse one Muggle or kick one puppy per hour,
how many Muggles should he curse in a day to maximize his expected satisfaction?

8. Bill has an infinite amount of time on his hands, which he spends by drawing fractals in full detail.
Beginning with a unit square, he inscribes a circle in every square he draws and inscribes a square in
every circle he draws. What is the total length he draws?
9. Two complex numbers z1 , z2 have purely imaginary product and purely real quotient. How many
ordered pairs (z1 , z2 ) are there such that |z1 | = |z2 | = 1?
∞ X
∞
X 1
10. Evaluate S = x+y+|x−y|
.
x=0 y=0
z
SMT 2008 Advanced Topics Solutions February 23, 2008

1. Answer: 21
The gladiators are able to attempt to attack n − 1 times. The expected number of hits is therefore
n−1 1
2 , and since hits are just as likely as misses, the mean is also the median. There is thus a 2 chance
n−1 n−1
that the gladiators hit at least 2 times, so 2 ≥ 10, giving n ≥ 21.

2. Answer: cos−1 24


25
The vectors from start to first drop-off and from there to the second are v~1 = (3, 4) and v~2 = (4, 3), both
with magnitude 5. The angle θ between these two vectors is the angle turned; v~1 · v~2 = |v1 | |v2 | cos θ,
so 12 + 12 = 5 · 5 · cos θ.
1
3. Answer: 3
Call the numbers x, y and suppose that x > y. Then the condition is that x − y > x+y 2 , or solving,
y < x3 . Therefore if y > x, we must have x < y3 . From the diagram below, we see that the probability
is therefore 61 + 16 .

13

13

π2 205
4. Answer: 6
− 144

∞ ∞
X 1 X 1
=
n=1
n + 8n + 16 n=1 (n + 4)2
2

∞
X 1 1 1 1
= 2
−1− − −
n=1
n 4 9 16

1
5. Answer: 280
9 9!


There are 3,3,3 = 3!3!3! = 1680 distinguishable ways of arranging the 9 letters on the grid. To satisfy
the conditions of the problem, we must have one of each letter per column and per row. We have 3
choices for X in row 1, then 2 for row 2, and only 1 choice for row 3, which makes 3 · 2 = 6 options in
all. Therefore, desired probability is 6/1680 = 1/280.
6. Answer: 2008!
Since bn is a fixed point, bn = (f (b))n , which is, as seen in the problem statement, a recursive definition
for factorial.
7. Answer: 24
Since each Muggle cursed has a 50 percent chance of dying, there is always a 50 percent chance that
the number of Muggles killed will be even (as long as the Dark Lord curses at least one Muggle). To
see this, note that the parity of the number of dead Muggles depends entirely on whether or not the
last Muggle cursed dies. No matter how many of the previous Muggles die, this last one will determine
whether or not an even number total die.
Now for a little trickery: Suppose Voldemort curses n Muggles. Let En be the expected value of the
number of Muggles killed given that this number is even, and F the expected value given that the
SMT 2008 Advanced Topics Solutions February 23, 2008

number killed is odd. Note that the overall expected value of the number of Muggles killed is n2 ,
which must be equal to En +F2
n
, so En + Fn = n. If n = 2k, then for each j, the probability of killing
k + j Muggles is the same as that of killing k − j Muggles, so En = Fn = k. Now if n = 2k + 1,
En = 21 En−1 + 12 (Fn−1 + 1), since if the number of Muggles killed is even, this is equally likely to result
from an even number of the first n − 1 dying and the last surviving, and an odd number of the first
n − 1 dying and the last also dying. But this gives En = n2 , so En = Fn .
Since En = Fn = n2 for all n, Voldemort’s expected satisfaction from cursing n Muggles is 3Fn + 6En =
9
2 n. This is greater than the expected satisfaction of 2n from kicking n puppies, so Voldy should ignore
the puppies and curse Muggles all day long.
√ √ √
8. Answer: (4 + π)(2 + 2) or 8 + 4 2 + 2π + π 2
The second square has diagonal the same length as the diameter of the first circle and therefore the
same as the edge of the first square. Therefore, each square is √12 the size of the previous, and the
circles scale in the same way. The first square has perimeter 4, and the first circle has circumference
∞  n √

√1
P
π. The total length is therefore (4 + π) 2
= (4 + π) 2 + 2 .
n=0

9. Answer: 8
Let zk = eiθk = cos θk + i sin θk for k = 1, 2. Recalling that multiplication corresponds to addition of
angles in the complex plane, while division corresponds to subtraction, we see that θ1 − θ2 must be
a multiple of π (real axis) while θ1 + θ2 must be an odd multiple of π/2 (imaginary axis). The first
condition need only be used for 0 · π and 1 · π (giving vectors in the complex plane either in the same
or in opposite directions). The second condition then reduces to either 2θ1 or 2θ1 + π being an odd
multiple of π/2, so θ1 must be an odd multiple of π/4, giving four possibilities from 0 to 2π. For each
of these, θ2 can either be the same or offset by π, giving a total of eight possibilities.
20
10. Answer: 9
Split the sum into three parts: One where y = x, one where y = x + d, and one where y = x − d for
∞ ∞ X ∞
X 1 1 4 X 1
some integer d > 0. The first is S1 = 2x
= 1 = . The second is S2 = 2x+2d
=
x=0
2 1− 4 3
d=1 x=0
2
∞ ∞ ∞
X 1 X 1 4X 1 4
= = . The third sum, S3 , is just obtained by switching x and y, so S3 = S2 .
22d x=0 22x 3 22d 9
d=1 d=1
20
Then S = S1 + 2S2 = 9 .
SMT 2008 General Test February 23, 2008

1. Calculate the least integer greater than 5(−6)(−5)(−4)···(2)(3)(4) .

2. How many primes exist which are less than 50?
3. Give the positive root(s) of x3 + 2x2 − 2x − 4.
4. A right triangle has sides of integer length. One side has length 11. What is the area of the triangle?

5. One day, the temperature increases steadily from a low of 45◦ F in the early morning to a high of 70◦ F
in the late afternoon. At how many times from early morning to late afternoon was the temperature
an integer in both Fahrenheit and Celsius? Recall that C = 59 (F − 32).
6. A round pencil has length 8 when unsharpened, and diameter 14 . It is sharpened perfectly so that it
remains 8 inches long, with a 7 inch section still cylindrical and the remaining 1 inch giving a conical
tip. What is its volume?
7. At the Rice Mathematics Tournament, 80% of contestants wear blue jeans, 70% wear tennis shoes,
and 80% of those who wear blue jeans also wear tennis shoes. What fraction of people wearing tennis
shoes are wearing blue jeans?

8. Terence Tao is playing rock-paper-scissors. Because his mental energy is focused on solving the twin
primes conjecture, he uses the following very simple strategy:
• He plays rock first.
• On each subsequent turn, he plays a different move than the previous one, each with probability
1
2.

What is the probability that his 5th move will be rock?

9. What is the sum of the prime factors of 20!?
10. Six people play the following game: They have a cube, initially white. One by one, the players mark
an X on a white face of the cube, and roll it like a die. The winner is the first person to roll an X (for
example, player 1 wins with probability 61 , while if none of players 1–5 win, player 6 will place an X
on the last white square and win for sure). What is the probability that the sixth player wins?
q √
11. Simplify: 3 17 47+45

12. If in the following diagram, m∠AP B = 16◦ and AP and BP are tangent to the circle, what is m∠ACB?

C P

13. Let N be the number of distinct rearrangements of the 34 letters in SUPERCALIFRAGILISTICEX-

PIALIDOCIOUS. How many positive factors does N have?
14. Suppose families always have one, two, or three children, with probability 14 , 21 , and 14 respectively.
Assuming everyone eventually gets married and has children, what is the probability of a couple having
exactly four grandchildren?
SMT 2008 General Test February 23, 2008

15. While out for a stroll, you encounter a vicious velociraptor. You start running away to the northeast
√
at 10 m/s, and you manage a three-second head start over the raptor. If the raptor runs at 15 2 m/s,
but only runs either north or east at any given time, how many seconds do you have until it devours
you?
16. Suppose convex hexagon HEXAGN has 120◦ -rotational symmetry about a point P — that is, if you
rotate it 120◦ about P , it doesn’t change. If P X = 1, find the area of triangle 4GHX.
17. An “expression” is created by writing down five random characters, taken from the digits 1 to 9, the
four basic operations, and parentheses. What is number of possible mathematically valid expressions
that can be created this way, without using implicit multiplication?

18. Cody enjoys his breakfast tacos with his favorite hot sauce, which comes in 88 mL bottles. On any
given day he eats 3, 4, or 5 tacos with probabilities 16 , 13 , and 12 , respectively. For the first taco, he
always uses 2 mL of hot sauce, but for each additional taco he uses 1 mL more than the previous. If
he starts with a new bottle, what is the probability that it is empty after five days?
19. Four pirates are dividing up 2008 gold pieces. They take turns, in order of rank, proposing ways to
distribute the gold. If at least half the pirates agree to a proposal, it is enacted; otherwise, the proposer
walks the plank. If no pirate ever agrees to a proposal that gives him nothing, how many gold pieces
does the highest-ranking pirate end up with? (Assume all pirates are perfectly rational and act in
self-interest, i.e. a pirate will never agree to a proposal if he knows he can gain more coins by rejecting
it.)
20. What is the smallest number which can be written as the sum of three distinct primes, the product of
two distinct primes and the sum of three distinct squares?
r
√
q p
21. Find the value of 2 + 2 + 2 + 2 + · · ·, assuming all square roots refer to the positive values.

22. What is the smallest possible surface area of an object constructed by joining the faces of five cubes
of edge length one?
23. You are standing at point A, which is 1000 feet from point B. You choose a random direction and
walk 1000 feet in that direction. What is the probability that you end up within 1000 feet of point B?
24. An isosceles right triangle with legs of length 1 has a semicircle inscribed within it and a semicircle
inscribed around it. Both have their diameter lying along the hypotenuse of the triangle. Find the
ratio of their radii (larger to smaller).

25. Three unit circles are mutually externally tangent. All three are internally tangent to a larger circle.
What is the radius of the larger circle?
SMT 2008 General Solutions February 23, 2008

1. Answer: 2
The exponent evaluates to 0; 50 = 1 so the least integer greater is 2.
2. Answer: 15
The primes less than 50 are: 2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47.
√
3. Answer: 2
√
√
√
The expression factors into x − 2 x + 2 (x + 2). Thus, 2 is the only positive root.
4. Answer: 330
11·60
The only Pythagorean triplet containing 11 is (11, 60, 61), giving a triangle with area 2 .

5. Answer: 3
The conversion formula C = (F − 32) · 59 requires that F − 32 be a multiple of 9 for C to be an integer.
Such integers between 45 and 70 are 50, 59, and 68.
11π
6. Answer: 96
1 2 1 2
+ 13 π · 1 · 7 1




The volume is π · 7 · 8 8 =π 64 + 192 .
32
7. Answer: 35
From the information provided, we can see that 64% of contestants are both earing blue jeans and
64
tennis shoes. Therefore, 70 = 32
35 of the contestants wearing tennis shoes are wearing blue jeans.
3
8. Answer: 8
If his nth move is rock, his (n + 1)th move will never be rock. If his nth move is not rock, then his
(n + 1)th move will be rock with probability 21 . So his second move will always be non-rock. His third
move will be rock with probability 21 . His fourth move can only be rock if his third is not, so this
happens with probability 14 . His fifth move can be rock only if his fourth is not, so the probability of
this happening is 34 · 12 = 38 .
9. Answer: 77
The product 1·2·3 · · · 19·20 contains all primes from 2 to 19, so the sum is 2+3+5+7+11+13+17+19 =
77.
5
10. Answer: 324
The probability of player n winning if no one has already done so is n6 . Restated, the probability that
player n will NOT win if no one has already done so is 6−n 6 . So the probability that the first 5 players
will all lose is 65 · 46 · 36 · 26 · 16 = 324
5
, and this is the only way the sixth player can win.
√
3+ 7
11. Answer: 2

s √ s √
q
3
√ √
3 17 + 7 + 45 3 34 7 + 90 (3 + 7)3 3+ 7
= = =
4 8 2 2
12. Answer: 98◦
Let O be the center of the circle. Since tangents and radii are at right angles, m∠AOB = 360◦ −
90◦ − 90◦ − 16◦ = 164◦ . The measure of the arc intercepted by ∠ACB is then 360◦ − 164◦ = 196, so
m∠ACB = 196/2
13. Answer: 3225600
This very long word contains 34 letters. One of them (I) is repeated seven times, four (S, C, A, L)
are repeated three times, five (U, P, E, R, O) are repeated twice, and the other five (F, G, T, X, D)
occur once. So N = (2!)534! 19 9 6 3 3 2 2
(3!)4 7! = 2 3 5 7 11 13 17 · 19 · 23 · 29 · 31, and the number of factors of N
is (19 + 1)(9 + 1)(6 + 1)(3 + 1)(3 + 1)(2 + 1)(2 + 1)(1 + 1)(1 + 1)(1 + 1) = 3225600.
SMT 2008 General Solutions February 23, 2008

27
14. Answer: 128
The ways to have four grandchildren are: two children each with two children, two children with one
and three children, and three children with one, one, and two children. There are two orders for the
second case and three for the third, so the total probability is:
 3  3  3
1 1 1 1 1
+2· · +3· ·
2 2 4 4 2
9
15. Answer: 4.5 or 2
√
Since the raptor is only running north and east, it √has to run 2 times as far as you. At time t,
you have√run 10t meters,
√ and the raptor has run 15 2(t − 3) meters. The raptor catches you when
10t = 15 2(t − 3) · 2, or 10t = 30(t − 3). This gives 20t = 90, or t = 29 = 4.5.
√
3 3
16. Answer: 4
Since the hexagon is convex, a 120-degree rotation around P must send H to X, X to G, and G to
H. But that means that 4GHX is equilateral with center P ! If√P X = 1, the triangle must have a
√
median length of 23 , and thus a side length of 3. So its area is 3 4 3 .
17. Answer: 150093
Let ◦ denote the four operations. For parentheses to be used, the expression must be of the form
(n) ◦ m or n ◦ (m), giving 2 · 4 · 92 possibilities. For two operations to be used, the expression must be of
the form l ◦ m ◦ n, giving 93 · 42 possibilities. For one operation to be used, the form must be k ◦ lmn,
kl ◦ mn, or klm ◦ n, giving 3 · 94 · 4 possibilities. Finally, there are 95 five-digit numbers. Adding up
gives the answer.
79
18. Answer: 288
Note that eating 3, 4, or 5 tacos requires 9, 14, or 20 mL of hot sauce, respectively. If x, y, z represent
the numbers of 3-taco, 4-taco, and 5-taco days, then emptying the bottle requires 9x + 14y + 20z ≥ 88,
which has three solutions in the integers: (0, 2, 3), (0, 1, 4), (0, 0, 5). The total probability of these three
cases is    2  3      4  5
5 1 1 5 1 1 1 1 1 1 1 1
+ + = 10 · · + 5 · · +
2 3 2 1 3 2 2 9 8 3 16 32

19. Answer: 2007

Notice that if only the third and fourth pirate remain, the third pirate can enact his proposal simply
by voting for it himself. Thus, he would be able to get all of the gold pieces, giving none to the fourth
pirate.
Therefore, if only the second, third, and fourth pirates remain, the second pirate can save 2007 gold
pieces for himself and offer 0 to the third pirate and just 1 to the fourth pirate. Notice that the fourth
pirate will support this proposal, since otherwise the second pirate would walk the plank, leaving the
above case in which the fourth pirate gets no gold at all.
So, when the first pirate proposes his offer, he need only offer the third pirate a single gold piece, and
save the rest for himself. Analogously, the third pirate will support this proposal, because otherwise
the second pirate’s proposal would leave him with nothing. Therefore, the highest-ranking pirate ends
up with 2007 gold pieces.

20. Answer: 14
The smallest sum of three distinct squares is 1 + 4 + 9 = 14, which conveniently can be expressed as
14 = 2 · 7 and 14 = 2 + 5 + 7.
SMT 2008 General Solutions February 23, 2008

21. Answer: 2
r
√ √
q p
Pick x = 2 + 2 + 2 + 2 + · · ·. Then by substitution x = 2 + x; squaring both sides gives
x2 = 2 + x. Since we assumed that all square roots in the definition for x refer to the positive values,
x is the positive root of the equation x2 − x − 2 = 0. Hence,
√
1+ 1+8
x= = 2.
2

22. Answer: 20
The most possible shared faces is 5, so the surface area is that of the five cubes minus two for each
shared face, or 6 · 5 − 2 · 5 = 20.
1
23. Answer: 3
Obviously, the closer your direction of travel is to the direction of B, the better. Let P be a point
such that P is 1000 feet from A and exactly 1000 feet from B. There are two such points, and they
are mirror images of each other across AB, so let the other one be P 0 . Since 4P AB is equilateral,
m∠P AB = 60◦ . Likewise, m∠P 0 AB = 60◦ , so m∠P AP 0 = 120◦ . If we walk in a direction that takes
us between P and P 0 , we will end up less than 1000 feet from B. The probability of this happening is
120 1
360 = 3 .
√
24. Answer: 2
Let the triangle be 4ABC with hypotenuse BC, and let P be the midpoint of √the hypotenuse, which
is also the center of each semicircle. Clearly, the outer semicircle has radius 22 . Now let X be the
point of tangency of the inner circle to AB. Then P X ⊥ AB. But then 4XBP ∼ √4ABC, so
P X = 21 BC = 12 . Since P X is a radius of the smaller circle, the ratio of the radii is thus 2.
√
2 3
25. Answer: 1 + 3
As seen in the figure, the centers of the unit circles form an equilateral triangle of side length 2.
√ √ √
Therefore an altitude of the triangle is 3 so OB = 32 3 and the radius OA = 1 + 2 3 3 .

A
SMT 2008 Power Test February 23, 2008

by Jeffrey Middleton
with thanks to Stephanie DeWet and Knots Knotes by Justin Roberts

Background and Definitions

What we usually think of as a knot is a 3-dimensional object. However, since we are unable to provide
3-dimensional paper, we will use a 2-dimensional definition. A knot is a single loop, possibly crossing over
itself, with gaps left whenever it crosses to indicate one arc “above” and one arc “below”. Only two arcs
may cross at a given point. (You can visualize this as a projection of a knot onto a sheet of paper.) Figure
1 shows two examples of knots.

(a) The unknot (b) The left trefoil knot (c) The right trefoil knot

Figure 1: Knots

Since our knots are two-dimensional, we must have rules for manipulating them. A convenient set of
such rules is called the Reidemeister moves, or the R-moves, which act on some small portion of the knot,
leaving the rest unchanged. (These should make sense if you again think of the knot as a projection. In
fact, Reidemeister’s Theorem states that the R-moves represent all possible manipulations of a 3D knot that
don’t involve cutting and rejoining it.) Figure 2 shows the four R-moves.

↔ ↔

(a) R0 (b) RI

↔ ↔

(c) RII (d) RIII

Figure 2: Reidemeister moves

Note that the R0 move is one we would take for granted; it allows deforming knots as desired as long
as it doesn’t affect crossings. You also need not explicitly mention when you use R0, and you may combine
more than one move in a single step, but be careful not to create confusing transformations by manipulating
a knot with multiple moves without drawing intermediate steps. Note also that although these moves still
apply to rotated versions of the depicted disks, they only apply with the crossings configured as shown (see
problem 1).
We can without much extra trouble also consider links, which are like knots except composed of multiple
loops. We call each loop a component; a knot is thus also a link (with one component). Link equivalence is
defined just as knot equivalence. Figure 3 shows a few examples of links.

(a) An unlink (b) The Hopf link (c) Borromean rings

Figure 3: Links
SMT 2008 Power Test February 23, 2008

Instructions
Write/draw all solutions neatly, with at most one question per page, clearly numbered. Turn in the solutions
in numerical order, with your team name at the upper right of every page. Be sure not to turn in two copies
of a question, and note that if we can’t decipher your drawings or handwriting, we can’t grade it. Partial
credit will be awarded where appropriate, and questions are weighted (though they are arranged in logical
order, not order of difficulty).
Finally, remember that for i < j you may use the results of problem i on problem j, even if you have not
solved problem i. For example, you may use the reverse R-moves of problem 1 in the rest of the problems.

Questions
1. Show that the R-moves imply the first two (RI2 and RIII2 ) reverse R-moves shown in Figure 4. You
may use the remainder of the reverse R-moves without proof for later problems.

↔ ↔ ↔

(a) RI2 (b) RIII2 (c) RIII3

↔ ↔ ↔

(d) RIII4 (e) RIII5 (f) RIII6

Figure 4: The reverse R-moves

2. Show that all knots with one or two crossings are equivalent to the unknot, a loop with no crossings,
and is denoted U .
3. Show that the following two knots are equivalent.

A knot invariant is a function which depends only on equivalence class of knots, not on which particular
representation is used. That is, if two knots are equivalent, the invariants of those knots are the same.
(Note that the invariant might also be the same for two non-equivalent knots!) Link invariants are defined
similarly.

4. (a) Let the crossing number c(K) be the minimal number of crossings of a knot K over all possible
manipulations. Show that crossing number is a knot invariant.
(b) Show that the number of components µ(L) of a link L is a link invariant.

An oriented knot is one with a chosen direction along the string, followed through the entire knot. An
oriented link is defined similarly (note that a direction is needed for each component). We define oriented knot
invariants and oriented link invariants as with unoriented knots and links. Note that it is sometimes helpful
to draw multiple arrows on a single component to keep track of the orientation, and also that properties of
unoriented knots generally apply to oriented versions as well. The R-moves and reverse R-moves work on
oriented knots and links as well, preserving the directions of arrows. See figure 5 for examples of oriented
knots and links. The R-moves apply to oriented knots and links as well, as long as you are careful to preserve
the orientation of the segments as you move them.
Let D be an oriented link. The linking number Lk(D) is the sum over all crossings of contributions given
by:
SMT 2008 Power Test February 23, 2008

(a) An oriented trefoil (b) An oriented knot (c) An oriented Hopf link (d) Oriented Borromean rings

Figure 5: Oriented knots and links

• 0 if the arcs of the crossing belong to the same component

• ±1 if the arcs of the crossing belong to different components (see figure 6). We call the ±1 the sign of
the crossings.

+1 −1

Figure 6: The signs of crossings

5. (a) Compute the linking number of the oriented Hopf link and Borromean rings shown in figures 5c
and 5d.
(b) Show that linking number is a knot invariant.
(c) Show that linking number is a link invariant.

Suppose we color a link D, using a single color for each connected arc, such that at every crossing the
three incident arcs (considering the arc on the bottom as two arcs instead of one) are either all the same
color or all different colors. For example, the three connected arcs of the left trefoil knot (figure 1b) cannot
be colored with two colors, since this would lead to two of one color and one of another at a crossing, but it
can be colored with a different color for each arc. This is called a 3-coloring, and τ (D) denotes the number
of possible 3-colorings of the link D.

6. Calculate τ for the trefoil knot (figure 1b).

7. Show that τ (D) is a link invariant.

The connect sum K1 #K2 of two oriented knots K1 and K2 is formed by cutting each in a single place,
and joining them together so that the orientations of each knot match up, forming a single new oriented
knot. Figure 7 shows an example of a connect sum. If the knots are unoriented, we form the connect sum
without worrying about matching up orientations.

8. Find, with proof, a formula for τ (K1 #K2 ) in terms of τ (K1 ) and τ (K2 ).

# =

Figure 7: A connect sum

9. Show that there are infinitely many distinct knots.

SMT 2008 Power Test February 23, 2008

−→ −→

Figure 8: Construction of a torus knot

A torus link Tp,q is created by placing p string segments from end to end of a cylinder, evenly spaced
around the circumference, twisting the cylinder through q/p full twists, and gluing the ends of the cylinder
together, forming a torus out of the cylinder which is then removed, leaving only the string. The example
in Figure 8 shows T3,3 , with all three segments moved to the front of the torus to make them visible.
10. Determine with proof the number of components of the link Tp,q . In particular, when is it a knot?
Another useful concept is the Kauffman bracket of an unoriented link D, denoted hDi. It is a polynomial
in integer powers of A defined by the following:
• It is invariant under R0 moves.
• It satisfies hDi = A hD+ i + A−1 hD− i, where D+ and D− are created by replacing a crossing as shown
below, while leaving the rest of the knot unchanged.

* + * + * +
−1
=A +A

• It satisfies hD U i = (−A2 − A−2 ) hDi, where

` `
represents a disjoint sum. (A disjoint sum of two
links is made by placing them together without entangling them and considering them one link. For
example, the unlink of figure 3a is the disjoint sum of two unknots.) That is, removing a disjoint
(unattached) unknot from the link results in a new link with Kauffman bracket (−A2 − A−2 ) times
that of the original link - see below for an example.
* + * +
2 −2


= −A − A

• hU i = 1; the bracket of a crossingless unknot is 1.

11. Calculate the Kauffman bracket of the Hopf link and left trefoil knot, shown in figures 3b and 1b.
12. Determine which, if any, of the R-moves the Kauffman bracket is invariant under.
If D is an oriented link, then the writhe w(D) is the sum of the signs of all crossings of D. That is, it is
the same as the linking number with self-crossings included as ±1, instead of being counted as 0.
13. Determine which, if any, of the R-moves the writhe is invariant under.
14. Show that the Jones polynomial fD (A) = (−A3 )−w(D) hDi is an invariant of oriented links.
15. Find a formula for fD1 ‘ D2 (A) in terms of fD1 (A) and fD2 (A), that is, for the Jones polynomial of a
disjoint union in terms of the Jones polynomials of the two parts of the union.
16. Find a formula for fD1 #D2 (A) in terms of fD1 (A) and fD2 (A), that is, for the Jones polynomial of a
connect sum in terms of the Jones polynomials of the two summands.
Reflecting a 3D knot across a plane in three dimensions causes all of the crossings in a corresponding
(2D) knot to flip (the arc that was on top goes to the bottom, and vice versa). A knot is called amphichiral
if it is equivalent to its mirror image, and chiral if it is not.
17. Find a method for determining whether a (2D) knot is amphichiral or chiral. Is the trefoil knot chiral?
SMT 2008 Power Solutions February 23, 2008

1.
RI RII
←→ ←→

RII RIII RII

←→ ←→ ←→

2. The basic one- and two-crossing knots are shown below; all others are the same as these with crossing
of the kinks reversed. They can all be transformed into the unknot using only RI.

3.

RII RIII RI RI RIII R0

←→ ←→ ←→ ←→ ←→ ←→

4. (a) The crossing number is a minimum over all manipulations, and if two knots are equivalent, they
can be transformed into each other, and therefore into all the same knots. They must therefore
have the same crossing number.
(b) The R-moves can never change the number of components, since they do not cut and rejoin arcs.
Thus the number of components for two equivalent links must be the same.
5. (a) The Hopf link has two positive crossings, giving it linking number 2. The Borromean rings have
three each positive and negative crossings, so they have linking number 0.
(b) The linking number of a knot is by definition 0, so it is a knot invariant.
(c) Since two knots are equivalent if and only if they are connected by a sequence of R-moves, we
must simply show that R-moves do not change the linking number. The part of a knot outside
the effects of the R-move does not change, so we just consider how the crossings changed by the
R-move affect the linking number. Clearly R0 moves do not affect it. RI moves remove/add a
crossing, but it is always a self-crossing, so it never contributes to the linking number. For RII
moves, we can similarly ignore the self-crossing case; if the two crossings belong two different
components, they always have opposite sign, so contribution to linking number is always zero.
Finally, RIII moves have three crossings. One is unchanged by the move, and the other two keep
the same sign, so the linking number is again unchanged. Since none of the R-moves change
linking number, it is a link invariant.
6. The three arcs of the trefoil may be colored all the same (3 possibilities) or all different (6 possibilities),
so τ = 9
7. We again consider the R-moves, with R0 obviously not affecting τ . We would like to show that every
coloring of the knot before the move corresponds to exactly one coloring of the knot after the move.
For an RI move, any coloring of the initial or final picture must be all the same color c, so the colorings
are essentially equivalent.
SMT 2008 Power Solutions February 23, 2008

For RII, the initial picture must have the top two ends the same color a, and from the constraint, the
bottom two ends must be the same color b (if a = b the middle arc must also be that color, and if
a 6= b the middle arc is the third color c). The final picture must also have the top two ends the same
color and the bottom two ends the same color, so a coloring of either defines a coloring of the other.
For RIII, there are five cases, based on the colors of the three left-hand ends. They could be all the
same, all different, or two the same and one different (giving three cases for the odd one out). It can
quickly be seen that each case determines the colors of the three right-hand ends, and matching end
colors to the final picture creates a unique coloring.
Therefore, all R-moves preserve τ , so it is a link invariant.
8. The splitting of each knot to form the connect sum breaks apart an arc which was all the same color.
A valid coloring of the sum can have any valid coloring of each summand, so long as they agree on
the joining segments. Suppose that the colors are a, b, c, and without loss of generality that the first
summand has color a on the split arc. Since a coloring of the second summand is still a valid coloring
with two colors switched, 31 of the colorings will have a on the split segment, 13 will have b, and 13 will
have c. Multiplying the number of choices for each knot gives the formula:
1
τ (K1 #K2 ) = τ (K1 )τ (K2 )
3

n−1 n
9. We calculated that τ of the trefoil knot is 9. Therefore, τ of a connect sum of n trefoils is 31 9 =
9 · 3n−1 . This takes a different value for each n, so the connect sums of trefoils are never equivalent to
each other, and there are infinitely many distinct knots.

10. Before turning the cylinder into a torus, number the points at the left end of the segments x1 , x2 , . . . , xp .
To find a component, start at a point xi1 and follow it around the torus once, so that we are at a
point xi2 which is rotated by 2π pq from xi1 . Repeat this until the component closes up, that is, until
xin = xi1 . We must have come an integer number of full rotations from xi1 , so we must have 2π pq n
p
be a multiple of 2π. The first time this happens is when n = gcd(p,q) . Each component therefore uses
p
up gcd(p,q) of the points, so the number of components is gcd pq. The special case of a knot has one
component, so torus knots must have gcd pq = 1.
11.
* + * + * +
=A + A−1
* + * +!
−1
=A A +A
* + * +!
+ A−1 A + A−1

= A2 (−A2 − A−2 ) + 1 + 1 + A−2 (−A2 − A−2 )

= −A4 − A−4

In finding the Kauffman bracket of the trefoil, we take advantage of the previous calculation for the
Hopf link, and the calculation for the one-crossing knot included therein.
SMT 2008 Power Solutions February 23, 2008

* + * + * +
=A + A−1

 
* + * +
= A(−A4 − A−4 ) + A−1 A + A−1
 


* + * +
4 −4 −2 2 −2
= A(−A − A )+ +A (−A − A )
* +
5 −3 −4
= −A − A + (−A )

= −A5 − A−3 + (−A−4 )(−A3 )

= −A5 − A−3 + A−1

12. The bracket is not RI invariant:

     
=A + A−1
   
=A + A−1 (−A2 − A−2 )
 
= (A − A − A−3 )

That is, an RI move changes the bracket by a factor of −A−3 . Similarly, a reverse RI move changes it
by a factor of −A3 .
The bracket is RII invariant:
     
=A + A−1
         
=A A + A−1 + A−1 A + A−1
       
= A2 + (−A2 − A−2 ) +A + A−2
 
=

The bracket is RIII invariant (using RII invariance in the calculation):

     
−1
=A +A
   
−1
=A +A
 
=

13. Clearly the writhe is invariant under R0. Our proofs for RII and RIII invariance of linking number
hold here as well. For RI, the crossing in the kink has sign +1 (and −1 for RI2 ) and is removed by RI,
so the writhe changes by −1 under RI (and +1 under RI2 ).
SMT 2008 Power Solutions February 23, 2008

14. Since both writhe and Kauffman brackets are invariant under RII and RIII, the Jones polynomial is
as well. Under an application of RI, the bracket changes by a factor of (−A3 ), which is made up for
by a loss of a factor of −A3 when the writhe in the exponent decreases by 1. We have shown that the
Jones polynomial is constant under all R-moves, and is therefore a link invariant.
15. As we’ve seen when calculating examples, the Kauffman bracket ultimately expands to powers of A
multiplied by n−1 factors of (−A2 −A−2 ), where n is the number of disjoint unknots left after applying
the recursive relation repeatedly. Suppose we first apply the relation to D1 in the disjoint union. If D1
creates n unknots, we must remove them all to begin working on D2 by itself - this is one more unknot
than we would have had to remove when working just with D1 , so all terms have an extra factor of
(−A2 − A−2 ) compared ` to hD1 i. All terms also still have hD2 i multiplied onto them. Factoring these
two parts out gives hD1 D2 i = (−A2 − A−2 ) hD1 i hD2 i. The writhe is a sum over all crossings, so it
is the sum of the writhes of the individual links D1 and D2 . The Jones polynomial is therefore

fD1 ‘ D2 (A) = (−A3 )−w(D1 +D2 ) (−A2 − A−2 ) hD1 i hD2 i = (−A2 + A−2 )fD1 (A)fD2 (A)

16. We can make a similar argument to that seen in the disjoint union, being careful to consider what
happens to the two arcs connecting D1 to D2 as we reduce it using the recursion relation by considering
the two crossings in D1 neighboring those two arcs. If we apply all the transformations for a particular
term in the bracket that would reduce D1 to n disjoint unknots, those two crossings must have been
reduced. Since the link must remain connected, the two arcs must eventually be joined, meaning that
effectively those two crossings became connected. Without the connect sum, this would have led to an
unknot, but now it simply creates an unknot joined to D2 by a connect sum. We therefore get only
n − 1 disjoint unknots from D1 before being fully reduced to D2 , which is exactly what would have
gone into the calculation of hD1 i. As with the previous problem, the writhe is unaffected, so the Jones
polynomial is
fD1 #D1 (A) = (−A3 )−w(D1 +D2 ) hD1 i hD2 i = fD1 (A)fD2 (A)

17. Mirroring a knot inverts the writhe (positive and negative crossings are switched). The Kauffman
bracket switches A and A−1 in the recursive part of the definition; everything else is unaffected. In
the Jones polynomial, we therefore flip the sign of the exponent of (−A3 ), and switch A and A−1 in
the bracket, therefore switching A and A−1 in the entire polynomial. Thus an amphichiral knot has
palindromic Jones polynomial (that is, the coefficient of Ak is the same as that of A−k , while a chiral
knot does not. As seen in the earlier calculation, the Kauffman bracket of one of the trefoil knots is
−A5 − A−3 + A−1 . This cannot be made palindromic by multiplying by a power of A, so the trefoil
knot must be chiral.
SMT 2008 Team Test February 23, 2008

1. Find the maximum value of esin x cos x tan x .

2. A fighter pilot finds that the average number of enemy ZIG planes she shoots down is 56z − 4z 2 , where
z is the number of missiles she fires. Intending to maximize the number of planes she shoots down, she
orders her gunner to “Have a nap . . . then fire z missiles!” where z is an integer. What should z be?

3. A sequence is generated as follows: if the nth term is even, then the (n + 1)th term is half the nth term;
otherwise it is two more than twice the nth term. If the first term is 10, what is the 2008th term?
√
4. Find the volume of the solid formed by rotating the area under the graph of y = x around the x-axis,
with 0 ≤ x ≤ 2.
5. Find the volume of a regular octahedron whose vertices are at the centers of the faces of a unit cube.

6. What is the area of the triangle with vertices (x, 0, 0), (0, y, 0), and (0, 0, z)?
7. Daphne is in a maze of tunnels shown below. She enters at A, and at each intersection, chooses a
direction randomly (including possibly turning around). Once Daphne reaches an exit, she will not
return into the tunnels. What is the probability that she will exit at A?

A C

8. In triangle AXE, T is the midpoint of EX, and P is the midpoint of ET . If triangle AP E is equilateral,
find cos(m∠XAE).
9. In rectangle XKCD, J lies on KC and Z lies on XK. If XJ and KD intersect at Q, QZ ⊥ XK, and
KC XD
KJ = n, find QZ .

10. Bill the magician has cards A, B, and C as shown. For his act, he asks a volunteer to pick any number
from 1 through 8 and tell him which cards among A, B, and C contain it. He then uses this information
to guess the volunteer’s number (for example, if the volunteer told Bill “A and C”, he would guess
“3”).
One day, Bill loses card C and cannot remember which numbers were on it. He is in a hurry and
randomly chooses four different numbers from 1 to 8 to write on a card. What is the probability Bill
will still be able to do his trick?

A: 2 3 5 7
B: 2 4 6 7
C: 2 3 6 1

11. Given that f (x, y) = x7 y 8 + x4 y 14 + A has root (16, 7), find another root.

12. How many nonrectangular trapezoids can be formed from the vertices of a regular octagon?
13. If reiθ is a root of x8 − x7 + x6 − x5 + x4 − x3 + x2 − x + 1 = 0, r > 0, and 0 ≤ θ < 360 with θ in
degrees, find all possible values of θ.
SMT 2008 Team Test February 23, 2008

Z π
2
14. For what real values of n is (tan(x))n dx defined?
−π
2

15. A parametric graph is given by: 

y = sin t
x = cos t + 12 t
How many times does the graph intersect itself between x = 1 and x = 40?
SMT 2008 Team Solutions February 23, 2008

1. Answer: e
1
The exponent simplifies to 2 sin(2x), which has maximum value 12 .
2. Answer: z = 7
d
Let dz (56z − 4z 2 ) = 56 − 8z = 0 ⇒ 8z = 56 ⇒ z = 7

3. Answer: 4
The first few terms of the sequence are 10, 5, 12, 6, 3, 8, 4, 2, 1, 4, 2, 1, . . .. Since the sequence repeats
every three terms and 2008 is one more than a multiple of 3, the 2008th term has the same value as
the 7th term.
4. Answer: 2π
We integrate using the disk method:
Z b Z 2 √
2
Z 2
V =π R(x)2 dx = π x dx = π xdx = 2π
a 0 0

1
5. Answer: 6
Remember that a regular octahedron is composed of two square pyramids. If the vertices are at the
centers of the faces of a unit cube, then the vertices of the bases of these pyramids are at the centers of
the sides of a unit square. Such a base has area 21 . The height of each pyramid is also 12 , which gives
a volume for each pyramid of 13 · 21 · 12 = 12
1
. The total volume is thus 16 .
p
6. Answer: 12 x2 y 2 + x2 z 2 + y 2 z 2
Let ~u = (−x, y, 0) and ~v = (−x, 0, z) be the vectors from the first point to the other two. The area of
the triangle is then half the magnitude of the cross product:

1 1
|~u × ~v | = |(yz, xz, xy)|
2 2
7
7. Answer: 15
Let the interior intersections from A, B, C, D be a, b, c, d, respectively. Let P (xY ) be the probability of
exiting at Y when currently at x. Then, because we enter at A, we are trying to find P (aA ) = 13 (1) +
1 1 1 1 1 1 1 1
3 P (bA ) + 3 P (dA ). Similarly, P (bA ) = 3 (0) + 3 P (aA ) + 3 P (cA ) and P (dA ) = 3 (0) + 3 P (aA ) + 3 P (cA )
1 1 1
and P (cA ) = 3 (0) + 3 P (bA ) + 3 P (dA ). Noting the symmetry between b and d, we have

 P (aA ) = 31 (1) + 23 P (bA )



P (bA ) = 13 (0) + 13 P (aA ) + 31 P (cA )

P (cA ) = 13 (0) + 23 P (bA )


This is a system of three equations with three unknowns. Solving for P (aA ) yields the required
probability of 7/15.
√
13
8. Answer: − 13
Let AE = EP = AP = 1. Then AP = P T = 1, so 4AP T is isosceles. Since m∠AP T√= 120◦ ,
m∠P AT = 30◦ , so ∠T AE is a right angle! This makes 4ET A a 30-60-90 triangle, so AT = 3. Since
T X = 2 and m∠P T A = 30◦ , m∠XT A = 150◦ , and we can use the Law of Cosines to find AX.
√ √ √
The Law of Cosines gives AX 2 = 3 + 4 − 4 3 cos(150◦ ) = 7 + 4 3 · 23 = 13.
√
AX 2 +AE 2 −EX 2 −2 13
Now use the Law of Cosines again to get cos(m∠XAE) = 2·AX·AE = √
2 13
=− 13 .
SMT 2008 Team Solutions February 23, 2008

1
9. Answer: n+1
Put XKCD in the coordinate plane with X at the origin, K on the positive y-axis, and D on the
positive x-axis. Let XK = t, and scale the rectangle so that KJ = 1 (and thus KC = n). Then the
length of QZ is just the x-coordinate of point Q, since it’s perpendicular to the y-axis. Line DK is
given by the equation y = − nt x + t, and XJ is given by y = tx. Q is their intersection point, so the
x-coordinate of Q is the solution to − nt x + t = tx. Multiply through by nt to get −x + n = nx, which
n
gives x = n+1 . Since XD = KC = n, XD 1
QZ = n+1 .

16 8
10. Answer: or 35
(84)
Notice that if we just consider cards A and B, there are four classes of numbers: 2 and 7 each appear
on both A and B; 3 and 5 appear only on A; 4 and 6 appear only on B; and 1 and 8 appear on neither
A nor B. So, using information from cards A and B alone, Bill can identify which of these four classes
the volunteer’s number belongs to, but nothing more. Thus, in order for Bill to be able to perform his
trick, he needs to be able to use card C to split each class in half; hence, C must contain one number
from each of the four classes. Thus, there are
24 choices of numbers that Bill can write that will allow
him to perform his trick. Since there are 84 choices available to Bill, the probability that he will be
4
able to perform his trick is 28 = 358
.
(4)
11. Answer: (49, 4)
√
Let g(x, y) = f (x, y) = x7 y 4 + x4 y 7 + A. g is then symmetric in x, y, so 0 = f (16, 7) = g(16, 49) =
g(49, 16) = f (49, 4), and therefore (49, 4) is a root.
12. Answer: 24
Label the vertices of the octagon ABCDEF GH. Then ABCD is a trapezoid, and there are 7 others
congruent to it. ACDF is also a trapezoid, and there are 7 others congruent to it. Finally, ABDE is
a trapezoid, and there are 7 congruent to it, making a total of 24 trapezoids.

13. Answer: 20, 60, 100, 140, 220, 260, 300, 340
9
The left side is equal to xx+1
+1
, so we must have x9 = −1 and x 6= −1. This gives 9θ = 360n + 180 for
some integer n, which gives θ = 20, 60, 100, 140, 220, 260, 300, or 340.
14. Answer: n < 1
∞
un
Z
2 2
Let u = tan(x), so du = sec (x)dx = (1 + u )du. Then the integral becomes 2
dx. For large
−∞ 1 + u
u, this function behaves almost exactly like un−2 , so if n < 1 it’s integrable, and if n > 1, it’s not. If
n = 1, the function’s antiderivative is 12 ln(1 + u2 ), but the limits of this function are ∞ as u goes to
±∞, so the integral is undefined. Thus we must have n < 1.

15. Answer: 12
At t = 0, (x, y) = (1, 0). This parametric equation makes a spiral where the loops do not touch. Hence,
every 2π steps, there will be an intersection, which corresponds to π steps in the x-axis. Since the
spiral intersects itself at a small distance before each 1 + 2πn, there are 39/π ≈ 12 intersections.
SMT 2009 Algebra Test and Solutions February 28, 2009

1. No math tournament exam is complete without a self referencing question. What is the product of
the smallest prime factor of the number of words in this problem times the largest prime factor of the
number of words in this problem?
Answer: 1681
There are 41 words in the problem statement. Since 41 is itself a prime, the answer is 412 = 1681.

2. King Midas spent 100x % of his gold deposit yesterday. He is set to earn gold today. What percentage
of the amount of gold King Midas currently has would he need to earn today to end up with as much
gold as he started?
100
Answer: x−1
%
After yesterday, the fraction of the initial gold remaining is 1− x1 = x−1
x . Therefore, in order to to reach
x 1
the original amount of gold, we must multiply by x−1 = 1 + x−1 . Thus, the gold must be increased
100
by x−1 percent.
3. Find all integer pairs (a, b) such that ab + a − 3b = 5.
Answer: (5, 0), (4, 1), (1, −2), and (2, −3)
We factor the expression as follows:

ab + a − 3b − 3 = 5 − 3
(a − 3)(b + 1) = 2

We can use a table to find appropriate values for a and b.

Thus, (5, 0), (4, 1), (1, −2), and (2, −3) are the desired solutions.

a−3 b+1 a b
2 1 5 0
-2 -1 1 -2
1 2 4 1
-1 -2 2 3

1


4. Find all the solutions for which f (x) + xf x = x.
Answer: x = 1

 
1
f (x) + xf = x
x
 
1 1 1
f + f (x) =
x x x
 
1 1 1
f = − f (x)
x x x
 
1 1
f (x) + x − f (x) = x
x x
x = 1

5. Find the minimum possible value of 2x2 + 2xy + 4y + 5y 2 − x for real numbers x and y.
Answer: − 54
SMT 2009 Algebra Test and Solutions February 28, 2009

We complete the square:

1 1
2x2 + 2xy + 4y + 5y 2 − x = (x2 + 2xy + y 2 ) + (x2 − x + ) + (4y 2 + 4y + 1) − ( + 1)
4 4
1 5
= (x + y)2 + (x − )2 + (2y + 1)2 −
2 4
1
Notice that x = 2 and y = − 12 would yield the minimum, which is − 54 .
1
6. The dollar is now worth 980 ounce of gold. After the nth $7001 billion “No Bank Left Behind” bailout
1
package passed by congress, the dollar gains 22n−1 of its (n − 1)th value in gold. After four bank
bailouts, the dollar is worth 1b 1 − 21c in gold, where b,c are positive integers. Find b + c.

Answer: 506
Let Pn be the value of the dollar in gold after the nth bailout. Let s = 21 . Then after the nth bailout,
n−1
the dollar is a factor of (1 + s2 ) of its (n − 1)th value. Thus,
1
P4 = (1 + s)(1 + s2 )(1 + s4 )(1 + s8 )
980
1
= (1 + s + s2 + s3 )(1 + s4 )(1 + s8 )
980
1
= (1 + s + s2 + s3 + s4 + s5 + s6 + s7 )(1 + s8 )
980
1
= (1 + s + s2 + s3 + s4 + s5 + s6 + s7 + s8 + s9 + s10 + s11 + s12 + s13 + s14 + s15 )
980 
1 − s16

1
= .
980 1−s
1 1 1


Plug in s = 2 , and we find that P4 = 490 1 − 216 . So b + c = 490 + 16 = 506.

2009
k
P
7. Evaluate b 60 c
k=1
Answer: 32670
Largest multiple of 60 below 2009 is 1980, so find the sum for k = 1 to 1979, so that we have each
32
value of bk/60c exactly 60 times. This sum is therefore 60(1 + 2 + . . . + 32) = 60(1 + 32) 22 = 31680.
The remaining terms are all 33, and there are 2009 − 1980 + 1 = 30 of them, giving an answer of
31680 + 30 × 33 = 32670.

8. “Balanced tertiary” is a positional notation system in which numbers are written in terms of the digits
1̄ (negative one), 0, and 1 with the base 3. For instance, 101̄1 = (1)30 + (−1)31 + (0)32 + 1(3)3 = 2510 .
Calculate (11̄00)(1̄1) + (11̄1) and express your answer in balanced tertiary.
Answer: 1̄01̄1
(11̄00)(1̄1) = 1̄1000 + 11̄00 = 1̄1̄00
1̄1̄00 + 11̄1 = 1̄01̄1
9. All the roots of x3 + ax2 + bx + c are positive integers greater than 2, and the coefficients satisfy
a + b + c + 1 = −2009. Find a.
Answer: −58
Let the roots be r, s, and t. Then they satisfy r + s + t = −a, rs + st + rt = b, and rst = −c. So we
have −(a + b + c + 1) = r + s + t − rs − rt − st + rst − 1 = (r − 1)(s − 1)(t − 1) = 2009 = 7 ∗ 7 ∗ 41.
Thus the roots are 8, 8, and 42, and a = −(r + s + t) = −58.
SMT 2009 Algebra Test and Solutions February 28, 2009

10. Let δ(n) be the number of 1s in the binary expansion of n (e.g. δ(1) = 1, δ(2) = 1, δ(3) = 2, δ(4) = 1).
Evaluate: P∞ δ(n) !
2
10 P∞ n=1 n
(−1)n−1 δ(n)
.
n=0 n2

Answer: 20
P∞ δ(n) P∞ (−1)n−1 δ(n)
For convenience, set x = n=1 n2 and y = n=0 n2 .

The crucial observation is that 12 (x + y) and 12 (x − y) give the same summation as x, restricted to the
terms with odd n and even n respectively. The latter summation is easily related to x using the fact
that δ(2n) = δ(n) (since multiplying by 2 is simply appending a 0 in the binary expansion), as follows.

1 X δ(n)
(x − y) =
2 n2
even n≥2
∞
X δ(2n)
=
n=1
(2n)2
∞
X δ(n)
=
n=1
4n2
1
= x.
4

Thus we have 12 (x − y) = 14 x. It follows that x = 2y, so x/y = 2. Thus, the desired answer is 20.
SMT 2009 Geometry Test and Solutions February 28, 2009

1. The sum of all of the interior angles of seven polygons is 180 · 17. Find the total number of sides of the
polygons.
Answer: 31
The angle sum of a polygon with n sides is 180(n − 2). The total sum is then 180(n1 − 2) + 180(n2 −
2) + · · · + 180(n7 − 2) = 180(n1 + n2 + · · · n7 ) − 7 · 2 · 180 = 180 · 17. Dividing through by 180 gives
n1 + n2 + · · · + n7 − 14 = 17, so the total number of sides 14 + 17 = 31.
2. The pattern in the figure below continues inward infinitely. The base of the biggest triangle is 1. All
triangles are equilateral. Find the shaded area.

√
3
Answer: 5
Rank
√
the shaded triangles by their area, largest to smallest. The largest shaded triangle has an area
of 163 . There are three of them, call them the first set. The ith set of triangles has base 14 that of the
1
(i − 1)th set, so the area of the ith set is 16 that of the of the (i − 1)th set. So the total shaded area
becomes an infinite geometric series:
√   √
3 1 3
·3· 1 = .
16 1 − 16 5

3. Given a regular pentagon, find the ratio of its diagonal, d, to its side, a.
Answer: 2 cos(36)
Consider the triangle formed by the diagonal and two sides of the pentagon. The interior angle of the
pentagon is 108◦ , so the other two angles are both 36◦ . By the law of sines, sin(36)
a = sin(108)
d . Thus,
d sin(108) sin(72) sin((2)(36))
a = sin(36) = sin(36) = sin(36) = 2 cos(36).

4. ABCD form a rhobus. E is the intersection of AC and BD. F lie on AD such that EF ⊥ F D. Given
EF = 2 and F D = 1. Find the area of the rhobus ABCD.
Answer: 20

B C

A F D

√
By the Pythogorean theorem, ED = 5. Since ABCD is a rhombus, AE ⊥ ED. So triangle
4F DE ∼ 4EDA. Thus we obtain the following ratio:
DF EF
=
ED AE
1 2
√ = .
5 AE
√ √ √
So AE = 2 5. Thus, the area is 12 (2 × AE)(2 × DE) = 12 (4 5)(2 5) = 20.
SMT 2009 Geometry Test and Solutions February 28, 2009

5. In the 2009 Rice Olympics, Willy and Sammy are two bikers. The circular race track has two lanes,
the inner lane with radius 11, and the outer with radius 12. Willy will start on the inner lane, and
Sammy on the outer. They will race for one complete lap, measured by the inner track. What is the
square of the distance between Willy and Sammy’s starting positions so that they will both race the
same distance? Assume that they are of point size and ride perfectly along their respective lanes.
√
Answer: 265 − 132 3
Denote r1 the inner radius and r2 the outer radius. Then the inner lane has distance 2πr1 and outer
lane 2πr2 . But since Sammy will only be racing for 2πr1 , there is 2π(r2 − r1 ) distance along the outer
lane which he will skip. Let W denote Willy’s starting position, S Sammy’s starting position, and O the
θ
origin of the circular race track. Let θ be the angle between W O and SO. Then 2π(r2 −r1 ) = 360 (2πr2 ).
◦
Plugging in r1 = 11 and r2 = 12 and solving √ for θ, we get θ = 30 . Using
√ coordinate geometry,
√
W = (11, 0) and S = (12 cos 30, 12 sin 30) = (6 3, 6). Thus, (W S)2 = (11 − 6 3)2 + 36 = 265 − 132 3.
6. Equilateral triangle ABC has side length of 24. Points D, E, F lie on sides BC, CA, AB such that
AD ⊥ BC, DE ⊥ AC, and EF ⊥ AB. G is the intersection of AD and EF . Find the area of the
quadrilateral BF GD.
√
117 3
Answer: 2

√ BC since 4ABC is equilateral, so CD = 12. 4ACD is a 30-60-90

AD bisects √ degree triangle, so
AD = 12 3. Likewise, 4DCE is also 30-60-90, so EC = 6 and ED√= 6 3. So AE = AC − EC =
24 − 6 = 18. 4EAF is also 30-60-90, so AF = 9 and EF = 9 3. Since ∠AEG = ∠AEF =
30◦ , ∠GED = 60◦ . Likewise, ∠CDE = 30◦ implies√EDG = 60◦ . So ∠DGE√must also √ be √ 60◦
and 4GED is equilateral, so EG = GD = ED = 6 3. F G = EF − EG = 9 3 − 6 3 = 3 3.
F B = AB − AF = 24 − 9 = 15 and BD = BC − CD = √12. So area of quadrilateral BF GD is
√ √
area4BF G + area4BDG = 12 (3 3)(15) + 12 (6 3)(12) = 1172 3 .
7. Four disks with disjoint interiors are mutually tangent. Three of them are equal in size and the fourth
one is smaller. Find the ratio of the radius of the smaller disk to one of the larger disks.
√
Answer: 3 − 32
Let r be the radius of the three largest circles and s be the radius of the smallest circles. Consider
the equilateral triangle 4ABC √ formed by the centers of the three largest circles. This triangle has
side length 2r and altitude r 3. Let O be the center of the smallest circle, and consider altitude AM ,
passing through O. AO = r + s = 32 AM , so the altitude is also 23 (r + s). Equating these and solving
√
for the ratio gives rs = 2 3 3 − 1.
8. Three points are randomly placed on a circle. What is the probability that they lie on the same
semicircle?
3
Answer: 4
Suppose the first two points are separated by an angle α. Note that α is therefore randomly chosen
between 0 and π. If we are to place the third point so that the three lie on the same semicircle, we
α
have an arc of measure 2π − α to choose from. The probability of this placement is therefore 1 − 2π .
1 3
This varies evenly from 1 at α = 0 to 2 at α = π. The average is therefore 4 .
SMT 2009 Geometry Test and Solutions February 28, 2009

9. Two circles with centers A and B intersect at points X and Y . The minor arc ∠XY = 120◦ with
respect to circle A, and ∠XY = 60◦ with respect to circle B. If XY = 2, find the area shared by the
two circles.
√ √
10π−12 3 10π 4 3
Answer: 9
, or 9
− 3
√
∠XAY = 120◦ , so the radius of circle A is 2 3 3 . ∠XBY in 60◦ , so the radius of circle B is 2. The area
of the sector AXY is 13 the area of circle A, so the area formed between segment XY and arc XY in
circle A is the area of sector AXY inus the area of 4AXY .
√ !2 √ √
1 2 3 1 3 4π − 3 3
π − ·2· = .
3 3 2 3 9

1
Similarly, sector BXY is of the area of circle B, so the area formed between segment XY and arc
6
XY in circle B is √ √
1 2 2 3 2π − 3 3
π(2) − (2) = .
6 4 3
The total area shared by the two circles is then:
√ √ √
4π − 3 3 2π − 3 3 10π − 12 3
+ =
9 3 9

10. Right triangle ABC is inscribed in circle W . ∠CAB = 65◦ , and ∠CBA = 25◦ . The median from C
to AB intersects W at D. Line l1 is drawn tangent to W at A. Line l2 is drawn tangent to W at D.
The lines l1 and l2 intersect at P . Compute ∠AP D.
Answer: 50◦

Note that ∠ACB = 90◦ , so AB must be the diameter of W . Then CO is the median from C to AB,
where O is the origin of W , and CD passes through O. Then CO = BO and ∠BCD = ∠CBA = 25◦ .
We calculate ∠COB = 180◦ − 2 × 25◦ = 130◦ . Then ∠AOD = 130◦ . Consider the quadrilateral
P DOA. ∠P = 360◦ − ∠BAD − ∠CDP − ∠AOD = 360◦ − 90◦ − 90◦ − 130◦ = 50◦ .
SMT 2009 Calculus Test and Solutions February 28, 2009

1 1
1. Find the exact value of 1 − + − . . ..
3! 5!
Answer: sin 1
x3 x5 x7
By Taylor Expansion, sin x = x − 3! + 5! − 7! + . . . . Let x=1, and the desired value equals sin 1.
2. At RMT 2009 is a man named Bill who has an infinite amount of time. This year, he is walking
1
continuously at a speed of 1+t 2 , starting at time t = 0. If he continues to walk for an infinite amount

of time, how far will he walk?

Answer: π2
R∞ 1 −1 −1
d = 0 1+t 2 dt = tan (t)|∞
0 = limt→∞ tan (t) − tan−1 (0) = π
2.

10x2
3. Evaluate lim .
x→0 sin2 (3x)
10
Answer: 9
By l’Hôpital’s rule,
10x2 20x
lim = lim
x→0 sin2 (3x) 6 sin(3x) cos(3x)
x→0

20x
= lim
x→0 3 sin(6x)
20
= lim
x→0 18 cos(6x)
10
= .
9
Z 1
4. Compute tan−1 (x)dx
0
π−2 ln(2)
Answer: 4
or equivalent expression
We integrate by parts:
Z 1 Z 1
−1
1 x
= x tan−1 (x) 0 −

1 · tan (x)dx 2+1
dx
0 0 x
 1
π 1
= −0− ln(x2 + 1)
4 2 0
π ln(2)
= − .
4 2
5. Let a(t) = cos2 (2t) be the acceleration at time t of a point particle traveling on a straight line. Suppose
at time t = 0, the particle is at position x = 1 with velocity v = −2. Find its position at time t = 2.
Answer: − cos(8)
32
+ 33
32

Z Z Z
1 + cos(4t) sin(4t) t
v(t) = a(t)dt = cos2 (2t)dt = dt = + + c1 ,
2 8 2
where c1 is a constant. Plug in t = 0 to find v(0) = c1 = −2. So v(t) = sin(4t)
8 + 2t − 2.
cos(4t) t2
Z Z
sin(4t) t
x(t) = v(t)dt = + − 2dt = − + − 2t + c2 .
8 2 32 4
1 33
Plug in t = 0 to get x(0) = − 16 + c2 = 1, so c2 = 32 . Thus,
cos(8) 33
x(2) = − + .
16 32
SMT 2009 Calculus Test and Solutions February 28, 2009

6. Find
∞
X dn −ax
(e )
n=2
dxn

for |a| < 1.

a2
Answer: 1+a
e−ax
dn −ax
Since dxn (e ) = (−a)n e−ax ,
∞ ∞
X dn −ax −ax
X
n
(e ) = e (−a)n .
n=2
dx n=2

This forms a geometric series with common ratio −a and first element a2 , which converges since |a| < 1.
a2 −ax
Thus the answer is 1+a e .
7. Compute
n  
X n−k 4k
lim cos .
n→∞ n2 n
k=1

1−cos(4)
Answer: 16
Define a partition on [0, 1] with n elements by setting xi = ni for 0 ≤ i ≤ n. Then xi − xi−1 = 1
n for
all i. If we let f (y) = (1 − y) cos(4y) and put yk = nk for 1 ≤ k ≤ n, then we have
n   n
X n−k 4k X
cos = f (yk )(xi − xi−1 ).
n2 n
k=1 k=1

Thus, we may conclude that

n
X Z 1
lim f (yk )(xi − xi−1 ) = f (y)dy
n→∞ 0
k=1
Z 1
= (1 − y) cos(4y)
0
  1
1−y cos(4y)
= sin(4y) −
4 16 0
− cos(4) 1
= + .
16 16
R∞
8. Evaluate 0
4bx + 7ce−2x dx. Remember to express your answer as a single fraction.
14e2 −12 14−12e−2
Answer: e2 −1
, or 1−e−2
To evaluate the floor function, split the integral into unit intervals:
Z ∞ ∞ Z
X k+1
4bx + 7ce−2x dx = 4(k + 7)e−2x dx
0 k=0 k
−0
= (14e − 14e−2 ) + (16e−2 − 16e−4 ) + (18e−4 − 18e−6 ) + . . .
= 12 + 2(e−0 + e−2 + e−4 + . . .)
2 14 − 12e−2 14e2 − 12
= 12 + = = .
1 − e−2 1 − e−2 e2 − 1
SMT 2009 Calculus Test and Solutions February 28, 2009

∞  n
X 1
9. Compute n .
n=0
5
5
Answer: 16
P∞ n
Let S = n=0 5n . Then
∞ ∞ ∞ ∞ ∞ 1
1 X n X n+1−1 X n+1 X 1 X n 5 1
S= = = − = − 1 =S−
5 n=0
5n+1 n=0
5n+1 n=0
5n+1 n=0
5n+1 n=1
5n 1− 5
4

S 1 4S 1 5
⇒ =S− ⇒ = ⇒S= .
5 4 5 4 16
∞
X 1
10. Evaluate 2 /80000
, as a decimal to the nearest tenth.
n=1
50 + n

Answer: 62.8
This sum is difficult to evaluate exactly. However, it can be closely approximated by the improper
integral of the same function, which is easily evaluated using u-substitution.

Z ∞ Z ∞
dx 1 dx
=
0 50 + x2 /80000 50 0 1 + (x/20000)2
Z ∞
1 2000du
=
50 0 1 + u2
2000  −1 ∞
= tan u 0
50
= 40 lim (tan−1 (b) − tan−1 (0))
b→∞
= 40 lim tan−1 (b)
b→∞
π
= 40
2
= 20π
≈ 62.83.

To see that this integral is correct to the nearest tenth, we observe that since the integrand is a mono-
tonic function, we can bound it above and below by Riemann sums. More precisely:

∞ ∞
X 1 X 1
2
≤ 20π ≤ 2
.
n=1
50 + n /80, 000 n=0
50 + n /80, 000

By rearranging terms, this implies that:

∞
1 X 1
20π − ≤ 2
≤ 20π.
50 n=1 50 + n /80, 000

From this it follows that 62.8 is indeed correct to the nearest tenth.
ERRATA

2009 Calculus #5:

The answer should be -cos(8)/32 - 63/32. In the final step, the t^2/4 and -2t terms both disappeared in
the official solution, but they actually add up to -3. Also, 1/16 and cos(8)/16 should be replaced by 1/32
and cos(8)/32 respectively.

This bug was reported by Colin Aitken on 21 December 2010.

SMT 2009 General Test and Solutions February 28, 2009

1. MacGyver is on a train, being pursued by a pack of velociraptors, starting with a 400m head start. He
frantically begins to build an anti-velociraptor weapon out of his paper clip, fork, spoon, energy bar,
and spring-loaded click pen; he needs 35 seconds to finish. The antiquated train is running at 10m/s,
and dramatically crashes to a stop after 20 seconds, but MacGyver finishes as the raptors close to only
5m away. How fast were the raptors running, in m/s?
Answer: 17
Let the answer be v. The raptors gain 20(v − 10) while the train is moving, then 13v after it stops, for
a total of 33v − 200. In this time they have gained 595, so we have 395 = 35v − 200, so v = 17.
2. I attach my pet snake, Earl, to one corner of my barn with a leash. The barn is square, with sides of
length 10, and the leash has a length of twenty, which wraps around the barn. I would like to make
sure that I am being humane to Earl, and would therefore like to know that area of my lawn he can
traverse while on the leash. What is this area?
Answer: 350π
We see that since this is at the corner of the barn, the snake is free to travel in a three quarter arc
around this point such that it does not intersect the barn. On the corners, we notice that the leash
will bend, and basically act as a shorter leash fixed at the other two corners. Thus, there is a leash of
length 10 for each of two quarter circles, which sum to half of an arc:
3 1
(20)2 π + (10)2 π = 350π.
4 2

3. A parallelogram is given with a base of length 2x + 15, and a height of 10 − x. Find x such that the
area is maximized.
5
Answer: 4
This can be expanded, and we see that it is quadratic:
−b 5
(2x + 15)(10 − x) = −2x2 + 5x + 150 =⇒ = =x
2a 4

4.
5. Find the sum of the distinct real roots of x4 − 3x3 + 4x2 − 3x + 1 = 0.
Answer: 1
It is apparent that 1 is a root. You can then factor the left side as (x − 1)(x3 − 2x2 + 2x − 1). The
second factor also has 1 as a root, so you can further factor it as (x − 1)(x − 1)(x2 − x + 1). The last
factor in this has no real roots, so 1 is the only real root.
6.
7. In a paralleogram, the measure of one angle is four times that of another angle. Find the measure of
the larger angle.
Answer: 144◦
The two different angles in a parallelogram sum to 180◦ . If x is the smaller angle, we have x + 4x =
5x = 180◦ , so x = 36◦ and the larger angle is 4x = 144◦ .
8. A fireworks factory can currently produce Fizzbangs for two dollars each. Upgrades are available for
the machinery, each of which will decrease cost by ten cents. The first upgrade costs $100, the second
costs $200, and so on. What is the cheapest cost (in dollars) to produce 10001 fizzbangs?
Answer: 15501
Each upgrade will reduce the total costs by $1000.1. Therefore the first ten upgrades, up to the one
that costs $1000, will save money. The cost of the fizzbangs will then be one dollar each, costing
$10001, and the upgrades will cost a total of $5500.
SMT 2009 General Test and Solutions February 28, 2009

√
9. The line y = mx intersects the line y = 10 − x at a distance of 82 from the origin. Find the product
of all possible values of m.
Answer: 1
Suppose the point (x, y) is the intersection point. The distance from (y, x) to the origin is the same,
and it will also be on the line y = 10 − x. The slope of the line from the origin to (y, x) is 1/m,
providing a second solution. The product of these two solutions is 1.
10. At the grocery store, Jeffrey notices berries on sale and decides to make an extremely large number
of berry tarts. Each tart uses 14 of a container of blueberries, 17 a container of blackberries, and 10
1

a container of strawberries. If he bakes an integer number of tarts, using all the berries, and resists
eating berries before baking, what is the fewest number of containers he could have used?
Answer: 69
140 140 140
The number of tarts made is 140, the least common multiple of 4, 7, and 10. This uses 4 + 7 + 10 =
35 + 20 + 14
11. An ice cream cone has a radius of 5 and a height of 8. After five hours in the sun, the ice cream melts,
filling the cone up to a height of 3. What is the volume of the melted ice cream?
225π
Answer: 64
3 15
The cone of melted ice cream is similar to the original cone, so it has a radius of 8 ·5 = 8 . The volume

2
of the cone is therefore 31 · 3 · 15
8 .
12. A brick wall containing several windows is built. In sections without windows, the wall uses 120 bricks
per meter, while in sections with windows, it uses only 80 bricks per meter. If the wall is 20 meters
long and contains 2120 bricks, how many meters of wall without windows are there?
Answer: 13
Let x be the length of wall without windows. The length of wall with windows is 20 − x, so the total
number of bricks used is 2120 = 120x + 80(20 − x). Solving for x gives x = 2120−1600
40 = 13.
13. A mouse factory makes 3- and 5-button mice. The factory normally uses 207 buttons a day, but one
day accidentally switches the orders and makes 5-button mice instead of 3-button mice, and vice versa,
and ends up using 281 buttons. How many 3-button mice does the factory normally make?
Answer: 49
Let x be the normal number of 3-button mice, and y the normal number of 5-button mice. We have
3x + 5y = 207 and 5x + 3y = 281. Subtracting 3 times the first equation from 5 times the second gives
25x − 9x = 5 · 281 − 3 · 207, so 16x = 784 and x = 49.

14. Find the largest prime factor of 3599.

Answer: 61
Notice that 3599 = 602 − 12 . Factoring the difference of two squares gives 602 − 12 = (60 − 1)(60 + 1).
59 and 61 are both prime, so the answer is 61.
15. How many factors does 12345 have?
Answer: 8
The number of factors of pn1 1 pn2 2 . . . pnmm , where each of pi is a prime, 1 ≤ i ≤ m, is

(n1 + 1)(n2 + 1) . . . (nm + 1).

Since 12345 = 3 × 5 × 823, we have (1 + 1)(1 + 1)(1 + 1) = 8 factors.

SMT 2009 General Test and Solutions February 28, 2009

16. Given a drawer with 8 white gloves, 12 black gloves, and 6 gray gloves, find the number of gloves you
need to pull out to ensure you have a pair of matching gloves. Assume that each glove has a matching
pair.
Answer: 14
We have three different colors, but each glove must have a matching pair of gloves. Thus, there are
8 12 6
2 = 4 white lefthanded gloves, 2 = 6 black lefthanded gloves, and 2 = 3 gray lefthanded gloves. We
have 4 + 6 + 3 = 13. But pulling out one more glove ensures a pair since all the remaining gloves are
righthanded. Thus, we need 14 gloves to make sure we have a matching pair.
17. Bill has made a bet with Tom. Bill will flip a fair coin 20 times; if all 20 come up heads, Bill wins a
million dollars. The first 19 coins come up heads. What is the probability that Bill will win?
1
Answer: 2
The first 19 flips are irrelevant. The probability that the last flip will be heads is still 12 .
18. As we know, a mathematician is a device for turning coffee into theorems. Nathan, a mathematician,
can prove a theorem in six hours (given enough coffee, of which he has an infinite supply). Silas’s
invention, the Lemm-o-Matic 1729, can prove a theorem in five hours. Working together, how long will
it take them to prove 100 theorems? Express your answer to the nearest hour.
Answer: 273
Nathan can proves theorems at a rate of 61 theorems per hour, while the Lemm-o-Matic 1729 can prove
theorems at a rate of 15 theorems per hour. Working together, Nathan and the Lemm-o-Matic can
prove 15 + 16 = 11
30 theorems per hour. To prove 100 theorems, it will Nathan and the Lemm-o-Matic a
100
total of 11/30 ≈ 273 hours.
19. In the parliament of Pythonistan, the Silly Party controls N seats, and the Sensible Party controls 25
seats. The Silly members always vote yes on everything, but they need at least 23 of the total members
of parliament to vote yes in order to pass a bill. For some values of N , they will find that by kicking
out one of their own members (and reducing the total membership by one), they will need fewer votes
from the Sensible party to pass bills. How many such values of N are there?
Answer: 0
There are 0 such values of N . By kicking out one of their own members, the Silly party can at most
reduce the number people needed to pass a bill by one, but at the same time, they reduce the number
of voters in the Silly party by one. Thus, the number of votes needed from the Sensible party does not
decrease.
20. Find the sum of the distinct real roots of x4 − 3x3 + 4x2 − 3x + 1 = 0.
Answer: 1
You can see immediately that 1 is a root, so factor the equation as (x − 1)(x3 − 2x2 + 2x − 1). The
second factor still has 1 as a root, so factor it further as (x − 1)2 (x2 − x + 1). The second factor now
has no real roots. The answer is 1, since 1 is the only real root.
21. Compute 1 + 2 + 3 + . . . + 1000000.
Answer: 500000500000
Pair the terms up: 1 pairs with 1000000, 2 pairs with 999999, etc. There are 500000 pairs, each of
which sums to 1000001, so the sum is 500000500000.
22. I want to join several equilateral triangles along their edges (so that vertices coincide) to form a 7-sided
polygon. What is the minimum number of triangles that I will need?
Answer: 5
Suppose we have just three triangles. The only way to join them (up to rotations and reflections) is to
make an isosceles trapezoid. If we add a fourth triangle to this, we can get at most six sides. However,
with a fifth triangle, we can get 7 sides. The arrangement that works comes from making a regular
hexagon out of six triangles, then removing one of them.
SMT 2009 General Test and Solutions February 28, 2009

23. An ellipse has semimajor axis 2 and semiminor axis 1. Find the distance between its foci.
√
Answer: 2 3
The focus, at points (c, 0) and (−c, 0), satisfy the equation a2 − b2 = c2 where a and b are the lengths
2 2 2
√ semiminor axes. Thus, we have that c = 2 − 1 = 3. The distance between
of the semimajor and the
the two foci are then 2 3.

24. Find all positive integers n such that n4 + n2 + 1 is prime.

Answer: 1
The expression factors as (n2 − n + 1)(n2 + n + 1). If it is to be prime, one of these must equal 1. But
the only way that can happen, for positive integers n, is if n = 1.

25. I flip 4 fair coins and eat all of the coins that land on tails. I flip all the uneaten coins again and then
eat all the coins that land on tails. What is the probability that I have eaten at least 3 coins?
189
Answer: 256
Probability is independent, so the probability

that any given
coin is eaten is 1 − (1/2 ∗ 1/2) = 3/4.
The probability of getting 3 or 4 coins is 43 (3/4)3 (1/4)+ 44 (3/4)4 = 108/256 + 81/256 = 189/256.

26. Cody can eat a 2 meter diameter pizza in 1 minute. Frank can eat a 2 meter diameter pizza in 2
minutes. Jeffrey can eat a 2 meter diameter pizza in 3 minutes. If they combine their powers together,
how many minutes does it take them to eat one pizza of diameter 6?
54
Answer: 11
minutes
The area of pizza Cody, Frank, and Jeffrey can eat per minute are π, π2 , π3 , respectively. So together,
they can eat π + π2 + π3 = 11π 6 54
6 square meters of pizza per minute. Thus, it takes 11π × 9π = 11 minutes
to eat a pizza of area 9π.
27. How many consecutive zeros occur at the end of the decimal expansion of (8!)!?.
Answer: 10076
The answer is the number of times 8! = 40320 is divisible by 5. This is equal to b 40320 40320
5 c + b 25 c +
40320
b 125 c + ..., which is 8064 + 1612 + 322 + 64 + 12 + 2 = 10076.
SMT 2009 Advanced Topics Test and Solutions February 28, 2009

1. Let a, b, c, and d be the numbers that show when four fair dice, numbered 1 through 6 are rolled.
What is the probability that |(a − 1)(b − 2)(c − 3)(d − 6)| = 1?
1
Answer: 324
The conditions implies that |a − 1| = |b − 2| = |c − 3| = |d − 6| = 1. a can equal 2, b can equal 1 or 3,
c can equal 2 or 4, and d can equal 5. So the probability is 16 ∗ 26 ∗ 26 ∗ 61 = 324
1
.

2. Find all possibilities for the second-to-last digit of a number whose square is of the form 1 2 3 4 5 6 7 8 9 0
(each is a digit).
Answer: 3, 7
Zero is the only digit with square ending in 0. The square of a number ending in zero will therefore
end in two zeros. Next digit of the number therefore needs a square ending in 9, so it is 3 or 7.

3. Ten gears are lined up in a single file and meshed against each other such that the ith gear from the
left has 5i + 2 teeth. Gear i = 1 (counting from the left) is rotated 21 times. How many revolutions
does gear 10 make?
147
Answer: 52
The number of teeth meshed does not vary. Thus, if n is the number of revolutions that gear 10 make,
then (5(1) + 21)(21) = (5(10) + 2)n ⇒ n = 7×21 147
52 = 52 .

4. In the game Pokeymawn, players pick a team of 6 different Pokeymawn creatures. There are 25 distinct
Pokeymawn creatures, and each one belongs to exactly one of four categories: 7 Pokeymawn are plant-
type, 6 Pokeymawn are bug-type, 4 Pokeymawn are rock-type, and 8 Pokeymawn are bovine-type.
However, some Pokeymawn do not get along with each other when placed on the same team: bug-type
Pokeymawn will eat plant-type Pokeymawn, plant-type Pokeymawn will eat rock-type Pokeymawn,
and bovine-type Pokeymawn will eat anything except other Bovines. How many ways are there to
form a team of 6 different Pokeymawn such that none of the Pokeymawn on the team want to eat any
of the other Pokeymawn?
Answer: 245
If we make our team all the same type, then there are 76 + 66 + 46 + 86 = 7 + 1 + 0 + 28 = 36

ways to do this. If we make our team partially bug and partially rock type, there are 62 44 + 63 43 +





6 4 6 4





4 2 + 5 1 = 15 ∗ 1 + 20 ∗ 4 + 15 ∗ 6 + 6 ∗ 4 = 15 + 80 + 90 + 24 = 209 ways. Any other combination
of types will not work. This gives a total of 245 ways.

5. Four cards are drawn from a standard deck (52 cards) with suits indistinguishable (for example, A♠
is the same as A♣). How many distinct hands can one obtain?
Answer: 1820, or 13 + 3 13 + 3 13 + 13





1 2 3 4

We proceed by casework.
Case 1 All cards have the same face value. There are 13


1 ways to choose the
face values.
Case 2 Some cards have face value A; some have face value B. There are 13 2
ways to choose A and
B. One can have the combinations ABBB, AABB, AAAB, so there are 3 13 2 distinct ways for this
case.
Case 3 Some cards have value A, some B, and some C. There are 13


3 ways
to choose the A, B, C.
One can have the combinations ABCC, ABBC, and AABC. There are 3 13 2 distinct ways for this
case.
Case 4 The cards are distinct: ABCD. There are 13


4
ways to do this. Since these cases are
mutually exclusive, we have 13 13 13 13




1 + 3 2 + 3 3 + 4 = 1820 distinct hands.

6. Find all complex numbers z such that z 5 = 16z̄, where if z = a + bi, then z̄ = a − bi.
√ √
Answer: 0, ±2, 1 ± i 3, −1 ± i 3
SMT 2009 Advanced Topics Test and Solutions February 28, 2009

Clearly 0 is a solution. Now we assume z 6= 0. We have |z 5 | = |16z̄|. By DeMoivre’s Theorem,

|z 5 | = |z|5 . The left hand side becomes |z 5 | = 16|z̄| = 16|z|. Equating the two sides, 16|z| = |z|5 ⇒
|z|4 = 16 ⇒ |z| = 2.
Multiplying both sides of the given equation by z,

z 6 = 16|z|2 = 64.

Let z = r(cos θ + i sin θ). Then r6 (cos(6θ) + i sin(6θ)) = 64. Thus, r = 2 and 6θ = 360k, for k =
0, 1, 2, 3, 4, 5. So our
√ other solutions
√ are 2, 2cis(60◦ ), 2cis(120◦ ), −2, 2cis(240◦ ), 2cis(300◦ ), which are
equal to ±2, 1 ± i 3, −1 ± i 3.
q √
7. Evaluate 1+2 3i
√
π
Answer: e 6 i , or ± 3+i
2
q √ √ √ √
π 1 π
Let x = 1+2 3i . Then x2 = 1+ 3i
2 . Converting to polar form, 1+2 3i = (e 3 i ) 2 = e 6 i = 3+i
2

8. Frank alternates between flipping a weighted coin that has a 23 chance of landing heads and a 13 chance
of landing tails and another weighted coin that has a 41 chance of landing heads and a 34 chance of
landing tails. The first coin tossed is the “2/3 − 1/3” weighted coin. What is the probability that he
sees two heads in a row before he sees two tails in a row?
13
Answer: 33
If the first toss comes up heads (2/3 probability), Frank has a 1/4 chance of getting another heads, a
(3/4) ∗ (1/3) = 1/4 chance of getting two successive tails, and a (3/4) ∗ (2/3) = 1/2 chance of getting
tails-heads and winding up back at his current position of tossing the “1/4−3/4” coin with the previous
toss being a heads. Expressing the probabilities as geometric series (or just the weighted probability
of the two nonrepeating options), he has a 1/2 chance of getting HH first and a 1/2 chance of getting
TT first. If instead, the first toss comes up tails (1/3 probability), he has a 3/4 chance of getting
another tails, a (1/4) ∗ (2/3) = 2/12 chance of getting two successive heads, and a (1/4) ∗ (1/3) = 1/12
chance of getting heads-tails and winding up back at my current state. Expressing the probabilities
as a geometric series, he has a 2/11 chance of getting HH first and a 9/11 chance of getting TT first.
The probability of getting HH before TT is (2/3) ∗ (1/2) + (1/3) ∗ (2/11) = 13/33.

9. The triangular numbers Tn = 1, 3, 6, 10, . . . are defined by T1 = 1 and Tn+1 = Tn + (n + 1). The square
numbers Sn = 1, 4, 9, 16, . . . are defined by S1 = 1 and Sn+1 = Tn+1 + Tn . The pentagonal numbers
Pn = 1, 5, 12, 22, . . . are defined by P1 = 1 and Pn+1 = Sn+1 + Tn . What is the 20th pentagonal
number P20 ?
Answer: 590
Expanding out the recurrence relations, we confirm that the triangular numbers are Tn = 1 + 2 + 3 +
· · · + n = n(n+1)
2 and the square numbers are Sn = n2 . A general formula for the pentagonal numbers
is therefore Pn = n2 + n(n − 1)/2 = n(3n − 1)/2. Substituting n = 20 gives P20 = 20(60 − 1)/2 = 590.

10. Evaluate eiπ/3 + 2e2iπ/3 + 2e3iπ/3 + 2e4iπ/3 + e5iπ/3 + 9e6iπ/3 .

Answer: 6
eiπ/3 + e2iπ/3 + e3iπ/3 + e4iπ/3 + e5iπ/3 + e6iπ/3 sum to 0 because the terms are sixth roots of unity (i.e.
they satisfy z 6 − 1 = 0, which is a 6th degree polynomial whose 5th degree coefficient is 0). Likewise,
e2iπ/3 + e4iπ/3 + e6iπ/3 sum to zero because the terms are cubic roots of unity. e3iπ/3 + e6iπ/3 sum to
0 because they are square roots of unity. Subtracting these sums from the original expression, we are
left with only 6e6iπ/3 , which is 6(cos(2π) + i sin(2π)) = 6.
 SMT/RMT Power Question 2009

By Silas Johnson, Nathan Pflueger, and Seva Tchernov

In this round, we will explore various fun things you can do with dice.
Throughout, we will refer to n-sided dice for various values of n. Don’t worry
about how you could actually construct such dice physically; just think of them
as random number selectors with n possible outcomes.
Prove all of your answers. Little or no credit will be given for numerical
answers without proof, even if correct.

Part I: Stacking Dice

“Standard dice” in this section refers to six-sided dice, numbered 1-6, with 1
opposite 6, 2 opposite 5, and 3 opposite 4. For all of these problems, ignore the
fact that there’s a table or other surface under the dice; the numbers on the
faces on the bottom of a stack count as being “visible”. Dice all have the same
chirality, meaning that a die and its mirror image do not both exist, all dice are
similar by rotation.

1. Suppose we stack up 7 identical standard dice in a 1*1*7 tower, in such

a way that any two faces that touch each other have the same number.
Find the sum of the numbers on the visible faces (remember, this includes
the bottom of the tower).
2. Now we arrange 4 identical standard dice in a 1*2*2 block, and impose
the additional condition that on each of the block’s 2*2 faces, the four
numbers showing are the same. Find the sum of the numbers on all the
visible faces.
3. Now we arrange 8 dice in a 2*2*2 cube. Find all possible values of the
sum of the numbers on all the visible faces.

Part II: Non-Transitive Dice

Suppose we have two dice A and B, and if we roll them both, A shows a higher
number than B more than half the time. Then we say A is “better” than B.

4. The following six-sided dice are known as Efron’s Dice:

A: 0, 0, x, x, x, x

1
 B: 3, 3, 3, 3, 3, 3
C: 2, 2, 2, 2, 6, 6
D: y, y, y, 1, 1, 1
They have the property that A is better than B, B is better than C, C
is better than D, and D is better than A. For this reason, we call them
nontransitive dice.

a. If x and y are integers, what must they be?

b. Suppose we roll all four dice at once, and the highest number showing
wins. Calculate the probability of winning for each die.

5. a. Design a set of three 3-sided non-transitive dice (i.e.A is better than

B, B is better than C, and C is better than A). Use the numbers 1
through 9 each once.
b. Suppose we want to do this in such a way that when all three dice
are rolled at once, each has an equal chance of winning. Prove that
this is impossible.
c. Do part (a), but use only the numbers 1-5, and don’t use any number
on two different dice (i.e. make ties impossible).
6. a. Prove that it is always possible to create a set of 2n + 1 dice (with
any number of sides) such that each die is better than n of the others
and worse than n of the others, for any positive integer n.
b. Do this with n = 2 (i.e. five dice), and make the dice 3-sided. Use
the numbers 1 through 15 each once.
c. Prove that you can always do this with 3-sided dice, and show how
to do it using the numbers 1 through 6n + 3.

7. Suppose we want to construct four six-sided dice so that A is better than

B, B is better than C, C is better than D, and D is better than A (like
Efron’s Dice). We also want to set them up so that in each of these cases,
the winning die wins with at least probability P. Prove that P cannot be
more than 23 .

Part III: Multiple Rolls

In this section, we examine a game where two players, Alice and Bob, have
different dice. They each roll their own die n times and add up the results.
Whoever has the highest total wins.

8. Suppose Alice’s die rolls 0 half the time and 2 half the time. Bob’s die
always rolls x, where x is an irrational number between 0 and 1. For what
values of n and x does Bob have a better chance of winning than Alice?
For what values does Alice have a better chance? For what values do they
each have an equal chance?
 Power Question Solutions

1. Answer: 105. Since the tower is of odd height, it is now the case that
the top and bottom faces are opposite faces, so every opposite pair of faces
sums to 7. Each of the 7 dice has 2 pairs of faces showing horizontally,
plus the one vertical pair up and down the tower. This gives (15*7) =

105.
2. Answer: 56. Having all 4 panels of the 2x2 face match implies that
adjacent die are rotated by 180 degrees. The blocks along the outside (not
the 2x2 face) therefore have the property that any neighboring blocks on
a 1x2 face contain opposite die sides. In other words, the arrows in the
accompanying figure point to blocks that sum to 7. There are 4 such pairs
of adjacent blocks and 4 pairs of front/back opposing sides on the 2x2

face, leading to a total sum of (4+4)*7 = 56.

3. Answer: {64, 72, 80, 88, 96, 104} and {68, 76, 92, 100} . There
are two families of solutions. The first is to take two blocks from Problem
2, flip one over, and stack them on top of each other. This gives 8 pairs of
matching sides but leaves the two 2x2 faces are all equal and allows their
die value to be set arbitrarily. The sum of all faces can take on the values
of 56 + 8i, i = 1, . . . , 6. This gives the first family in the answer. The other
valid arrangement, shown on the right, attempts to form a 2x2 face with
every die different. The other faces have matching pairs (the arrows in
the diagram on the right show equal numbered faces rather than opposite

1
 valued faces). The end result will be 8 opposite pairs that sum to 7 plus
four pairs that are 90 degrees away from each other on the die rather than
opposites. This gives the second family of values in the solution, plus some

redundant values with the first family.

4a. For A to be better than B, we must have x > 3. For D to be better than
A, we must have y > x. For C to be better than D, we must have y < 6.
From these we get that the only option is x = 4 and y = 5.
4b. Since B is always going to roll 3, the winning number die must roll at least
3. So if C rolls 6, C wins, and otherwise must lose. The probability of C
winning is 31 . For D to win, C must not roll 6, and D must roll 5. The
probability of this is 23 ∗ 12 = 13 also. For A to win, neither of these must
happen, and A must roll 4. This happens with probability 31 ∗ 23 = 92 . The
remainder, 19 , is the probability that B will win.
5a. One way to do it (there are many) is:
A: 1, 6, 8
B: 3, 5, 7
C: 2, 4, 9
5b. Suppose WLOG that the 9 goes on die A. This already gives A a 31 chance
of winning, so we must make sure A can’t win unless it rolls a 9. However,
when A doesn’t roll a 9, B and C must have an equal chance of winning
overall. This means that when only B and C are rolled, each has an equal
chance of winning. Thus B is not better than C, a contradiction.
5c. Here’s one way to do it:
A: 3, 3, 3
B: 5, 2, 2
C: 4, 4, 1
6a. You can do it like this:
D1 : 1, 4n + 2, 6n + 2, 8n + 2, ..., 2n(2n + 1) + 2
D2 : 2, 2n + 2, 6n + 3, 8n + 3, ..., 2n(2n + 1) + 3
D3 : 3, 2n + 3, 4n + 3, 8n + 4, ..., 2n(2n + 1) + 4
D4 : 4, 2n + 4, 4n + 4, 6n + 4, ..., 2n(2n + 1) + 5
...
D2n : 2n, 4n, 6n, 8n, ..., 2n(2n + 1) + (2n + 1)
D2n+1 : 2n + 1, 4n + 1, 6n + 1, 8n + 1, ..., 2n(2n + 1) + 1

2
 Notice how in each column, we write down the next 2n + 1 integers, but
each time we start one row lower. Each die in this set is better than the
previous n in the list (wrapping around if necessary) and better than the
following n.
6b. Here is a way to do it:
A: 1, 9, 14
B: 2, 10, 12
C: 3, 6, 15
D: 4, 7, 13
E: 5, 8, 11
6c. Construct the dice like this:
D1 : 1, 3n + 3, 6n + 2
D2 : 2, 3n + 4, 6n
D2 : 3, 3n + 5, 6n − 2
...
Dn : n, 4n + 2, 4n + 4
Dn+1 : n + 1, 2n + 2, 6n + 3
Dn+2 : n + 2, 2n + 3, 6n + 1
Dn+3 : n + 3, 2n + 4, 6n − 1
...
D2n : 2n, 3n + 1, 4n + 5
D2n+1 : 2n + 1, 3n + 1, 4n + 3
The procedure we’ve used here is as follows: First, write down the numbers
from 1 to 2n + 1, going straight down the list of dice. Then do the same
with the numbers from 2n + 2 to 4n + 2, but start at die Dn+1 . Finally, fill
in the numbers from 4n + 3 to 6n + 3, but as you go down the list, write
the even numbers first in descending order, followed by the odd numbers
in descending order. In this way, each die is better than the n that precede
it, and is worse than the n that follow it.
7. Let the dice be:
A: a1 , a2 , ..., a6
B: b1 , b2 , ..., b6
C: c1 , c2 , ..., c6
D: d1 , d2 , ..., d6
Assume each of the dice has the numbers on its faces written in descending
order (so a1 > a2 > ... > a6 , etc). Consider just A and B. Suppose we
want to make A beat B with a probability of greater than 32 . There are 36

3
 possible outcomes of rolling both dice, so A must win at least 25 of these.
Suppose a4 < b4 . Then A can roll a4 and win in at most two of the 36
cases. Since a5 and a6 are even worse, they can also account for at most
two winning cases for A. And obviously, a1 , a2 , and a3 can only account
for six winning cases each. This is a total of 24, which is not enough. This
shows that we must have a4 > b4 .
But the same argument applies to B and C, C and D, and D and A. So
we find that we need a4 > b4 > c4 > d4 > a4 , which is impossible.
8. Suppose n is even. Bob’s score will always be nx, which is less than n. If
Alice scores n or better, she will win. But this requires Alice to roll 2 on
at least half of her rolls, which she happens with probability greater than
1
2 . So if n is even, Alice always has the better chance of winning.
Now suppose n is odd. To score n or better (which would guarantee her
a win), Alice must roll 2 on more than half of her rolls. Since n is odd,
the probability of this occurring is exactly 12 . If Alice also wins when she
rolls 2 on n−1
2 of her n rolls (i.e. just under half), then she has the better
chance of winning. Otherwise, they have an equal chance of winning. The
condition for them to have equal chances is that nx > 2∗ n−1 n−1
2 , or x > n .
To summarize, Bob never has a greater chance of winning. They have
equal chances if n is odd and x > n−1
n , and otherwise Alice has the better
chance.

4
SMT 2009 Team Test and Solutions February 28, 2009

1. In the future, each country in the world produces its Olympic athletes via cloning and strict training
programs. Therefore, in the finals of the 200 m free, there are two indistinguishable athletes from each
of the four countries. How many ways are there to arrange them into eight lanes?
Answer: 2520

8! 7!
4
= = 2520
(2!) 2

2. Factor completely the expression (a − b)3 + (b − c)3 + (c − a)3 .

Answer: 3(a − b)(b − c)(c − a)
The expression is zero when any two of a, b, and c are equal. So it must have (a − b)(b − c)(c − a) as a
factor. But the original polynomial is degree 3, and so is this one, so the remaining factor must be a
constant. The original polynomial contains a term 3ab2 , but (a − b)(b − c)(c − a) only contains a term
ab2 , so the constant must be 3.
3. If x and y are positive integers, and x4 + y 4 = 4721, find all possible values of x + y.
Answer: 13
Consider the equation modulo 5. All fourth powers are either 0 or 1 mod 5. So one of x and y must be
divisible by 5; suppose it’s x. Then we must in fact have x = 5, since x = 10 is too large. This gives
y = 8, and this is the only possible solution. So the answer is 13.
4. How many ways are there to write 657 as a sum of powers of two where each power of two is used at
most twice in the sum? For example, 256+256+128+16+1 is a valid sum.
Answer: 41
A recursion relationship describing this problem is
a1 = 1, a2 = 2, a2n = an + an−1 , a2n+1 = an
where an is the number of valid sums for n. Thus,
a657 = a328 = a164 + a163 = a82 + 2a81 = a41 + 3a40 = 3a19 + 4a20
= 4a10 + 7a9 = 4a5 + 11a4 = 4a2 + 11a4 = 4 · 2 + 11 · 3 = 41.

5. Compute Z ∞
t5 e−t dt
0

Answer: 120
R∞
Define Γ(n) = 0 tn−1 e−t dt. Using integration by parts,
Z ∞
Γ(n + 1) = tn e−t dt
0
Z ∞
= −tn e−t |∞
0 + ntn−1 e−t dt
0
Z ∞
= 0+n tn−1 e−t dt
0
= nΓ(n).
R∞
Next we evaluate Γ(1) = 0 e−t dt = −e−t |∞
0 = 0 − −1 = 1. Thus, Γ(n + 1) = nΓ(n) = . . . = n!Γ(1) =
n!. So for the problem, Γ(6) = 5! = 120.
6. Rhombus ABCD has side length 1. The size of ∠A (in degrees) is randomly selected from all real
numbers between 0 and 90. Find the expected value of the area of ABCD.
2
Answer: π
SMT 2009 Team Test and Solutions February 28, 2009

1 θ θ
Area of Rhombus ABCD = 4∗ ∗ cos sin
2 2 2
θ θ
= 2 ∗ cos sin = sin θ
2 2

Z π
1 2 sin θ dx
E[Rhombus ABCD] = π
2 −0 0
2
= ∗1
π
2
= .
π
7. An isosceles trapezoid has legs and shorter base of length 1. Find the maximum possible value of its
area.
√
3 3
Answer: 4
Let the angle between the longer base and the leg be θ.
1
The Area of the Trapezoid ∆(θ) = sin θ + sin θ ∗ cos θ = sin θ + 2 sin 2θ
The area reaches extrema when its derivative is zero:
∆0 = cos θ + cos 2θ = 0
We use the formula cos 2θ = 2 ∗ cos2 θ − 1
2 ∗ cos2 θ + cos
√θ − 1 = 0
−1 ± 9 1
cos θ = = or −1 (omitted)
√ 4 2
3
sin θ =
2
√
∆M ax = 3 4 3

8. Simplify
n
X k 2 (k − n)
.
n4
k=1

−n2 +1
Answer: 12n2

n n
X k 2 (k − n) X k3 − k2 n
=
n4 n4
k=1 k=1
n n
X k3 X k2
= −
n4 n3
k=1 k=1
n n
1 X 3 1 X 2
= k − 3 k
n4 n
k=1 k=1
  2    
1 n(n + 1) 1 n(n + 1)(2n + 1)
= −
n4 2 n3 6
4 3 2
n + 2n + n 2n + 3n2 + n
3
= −
4n4 6n3
4 2
−n + n
=
12n4
−n2 + 1
= .
12n2
SMT 2009 Team Test and Solutions February 28, 2009

y2
9. Find the shortest distance between the point (6,12) and the parabola given by the equation x = 2 .
√
Answer: 2 17
Find the point on the parabola closest to the point (6,12). Call it (x, y) This point is where the normal
line at x crosses the parabola. We find the derivative by:
x = y2
dx = ydy
dy 1
=
dx y
The normal line will have slope of −y. It will contain (6, 12). Its equation is:
y − 12 = −y(x − 6)
y = −xy + 6y + 12
y3
y = − + 6y + 12
2
2y = −y 3 + 12y + 24
0 = y 3 − 10y − 24

The roots are 4 and two other imaginary answers, so 4 is the only one that works.

y − 12 = −y(x − 6)
−8 = −4(x − 6)
x=8

√
Find the distance between (8, 4) and (6, 12). The answer is 2 17.
X∞ n


2009
10. Evaluate .
n=2009
2n
Answer: 2
More generally, define a function G by

∞
X n


m
G(m) = .
n=m
2n

Thus we wish to evaluate G(2009). Observe that for all m ≥ 1:

∞
X n


m
G(m) =
n=m
2n
∞ n−1 n−1



X m−1 + m
=
n=m
2n
∞ n ∞ n



1 X m−1 1 X m
= +
2 n=m−1 2n 2 n=m−1 2n
1
= (G(m − 1) + G(m))
2
SMT 2009 Team Test and Solutions February 28, 2009

And thus G(m) = G(m − 1). Thus it suffices to evaluate G(0). However, this is simply a geometric
series:

∞
X 1
G(0) =
n=0
2n
= 2.
n 1

2009
NOTE: By noticing that 2009 is 2009! n
asymptotically, one can see this summation as a discrete
Z ∞ x−1
t
analogue of the Euler Γ function, which is defined by Γ(x) = dt. The solution above is similar
0 et
to the proof that Γ(n + 1) = nΓ(n).
11. Let z1 and z2 be the zeros of the polynomial f (x) = x2 + 6x + 11. Compute (1 + z12 z2 )(1 + z1 z22 ).
Answer: 1266

(1 + z12 z2 )(1 + z1 z22 ) = 1 + z12 z2 + z1 z22 + z13 z23

= 1 + z1 z2 (z1 + z2 ) + (z1 z2 )3 .
Since z1 + z2 = −6 and z1 z2 = 11,
(1 + z12 z2 )(1 + z1 z22 ) = 1 + 11(−6) + 113
= 1266.

12. A number N has 2009 positive factors. What is the maximum number of positive factors that N 2 could
have?
Answer: 13689
2009 = 72 × 41
We know for a number n = aα α2 αn
1 × a2 × . . . × an , it has (α1 + 1) × (α2 + 1) × . . . (αn + 1) factors.
1

Hence, for number N, we have the following options:

α1 = 7 − 1 = 6, α2 = 7 × 41 − 1 = 289 − 1 = 288
α1 = 7 − 1 = 6, α2 = 7 − 1 = 6, α3 = 41 − 1 − 40
By the same fact mentioned above, N 2 has: (2 ∗ α1 + 1) × (2 ∗ α2 + 1) × . . . (2 ∗ αn + 1) factors.
Calculating this number for both, we get the 2nd option gets us a bigger number: 13 × 13 × 81 =
13689
13. Find the remainder obtained when 17289 is divided by 7?
Answer: 3

   
289 289
17289 ≡ (14 + 3)289 ≡ 14288 3 + . . . + 13289−n 3n + . . .
1 n
3289 ≡ 3289 ( mod 7)

Note that 33 ≡ 27 ≡ −1( mod 7). Then 3289 ≡ 339̇6 3̇1 ≡ (−1)96 3̇1 ≡ 3 mod 7. Thus, the remainder
is 3.
14. Let a and b be integer solutions to 17a + 6b = 13. What is the smallest possible positive value for a − b?
Answer: 17
equation modulo 23, we get −6(a − b) ≡ −10 ( mod 23). Since -4 is an inverse of -6 modulo 23, then
we multiply to get (a − b) ≡ 17 ( mod 23). Therefore, the smallest possible positive value for (a-b) is
17. This can be satisfied by a = 5, b = −12.
SMT 2009 Team Test and Solutions February 28, 2009

2008!
15. What is the largest integer n for which 31n is an integer?
Answer: 66
b 2008 2008 2008 2008
31 c + b 312 c + b 313 c + b 314 c + · · · = 64 + 2 + 0 + 0 + · · · = 66.
SMT 2010 Algebra Test February 28, 2010

v s
u r
√
u q p
1. Compute 1 + 1 + 1 + 1 + 1 + 1 + ....
t

2. Write 0.2010228 as a fraction.

3. Bob sends a secret message to Alice using her RSA public key n = 400000001. Eve wants to listen in
on their conversation. But to do this, she needs Alice’s private key, which is the factorization of n.
Eve knows that n = pq, a product of two prime factors. Find p and q.
4. If x2 + 1/x2 = 7, find all possible values of x5 + 1/x5 .

5. A series of lockers, numbered 1 through 100, are all initially closed. Student 1 goes through and opens
every locker. Student 3 goes through and “flips” every 3rd locker (“flipping”) a locker means changing
its state: if the locker is open he closes it, and if the locker is closed he opens it. Student 5 then
goes through and “flips” every 5th locker. This process continues with all students with odd numbers
n < 100 going through and “flipping” every nth locker. How many lockers are open after this process?

6. Consider the sequence 1, 2, 1, 2, 2, 1, 2, 2, 2, 1, 2, 2, 2, 2, 1, ... Find n such that the first n terms sum up
to 2010.
7. Find all the integers x in [20, 50] such that 6x + 5 ≡ −19 mod 10, that is, 10 divides (6x + 15) + 19.
8. Let P (x) be a polynomial of degree n such that P (k) = 3k for 0 ≤ k ≤ n. Find P (n + 1).

9. Suppose xy − 5x + 2y = 30, where x and y are positive integers. Find the sum of all possible values of
x.
10. Find the sum of all solutions of the equation
1 2 3 4
+ + + = 2010x − 4.
x2 − 1 x2 − 2 x2 − 3 x2 − 4
SMT 2010 Algebra Solutions February 28, 2010

√
1. Answer: 1+2 5
√ √
q p p
Let x = 1 + 1 + 1 + .... Then x2 = 1 + 1 + 1 + .... Thus x2 = x + 1. The positive root of
√
1+ 5
x2 − x − 1 = 0 is 2 .
334703
2. Answer: 1665000
This is simply
2010 1 228 334703
+ · = .
10000 10000 999 1665000
3. Answer: 19801 and 20201
Notice that 4x4 + 1 = 4x4 + 4x2 + 1 − (2x)2 = (2x2 + 2x + 1)(2x2 − 2x + 1). Setting x = 100, we have
that 400000001 = 19801 · 20201.
4. Answer: ±123
1 2 1 1


Note that x + x = x2 + x2 + 2 = 9. Thus, x + x = ±3. Therefore,
  5   
1 1 1 1
x5 + = − 5 x3 + 3 − 10 x +
x+
x5 x x x
 5  3  
1 1 1
= x+ −5 x+ +5 x+
x x x
= ±123

5. Answer: 17
The open lockers will be the ones with an odd number of odd divisors. These numbers are of the form
2k · n2 , where n is odd. We can simply check that the open lockers are numbered

1, 2, 4, 8, 9, 16, 18, 25, 32, 36, 49, 50, 64, 72, 81, 98, 100.

6. Answer: 1027
If S(n) is the nth partial sum, note that if m is the kth triangular number, S(m) = k 2 . Since 442 = 1936
and 452 = 2025, we want to begin our search at 44(44 + 1)/2 = 990. Because (2010 − 1936)/2 = 37,
37 more 2s are needed, so the needed term is n = 990 + 37 = 1027.
7. Answer: 21,26,31,36,41,46

6x + 5 ≡ −19 mod 10
6x ≡ −24 mod 10
10
x ≡ −4 mod
gcd(10, 6)
x ≡ −4 mod 5
x≡1 mod 5

That is, x is in the form 5k + 1 where k is an integer.

8. Answer: 3n+1 − 2n+1
We use the fact that if P (x) is a polynomial of degree n, then P (x + 1) − P (x) is a polynomial of degree
n − 1. Define ∆P (x) = P (x + 1) − P (x). By induction on m, it can be easily proved that ∆m P (x) is a
polynomial of degree n − m such that ∆m P (k) = 2m · 3k for 0 ≤ k ≤ n − m when 0 ≤ m ≤ n. Moreover,
∆n+1 P is identically zero, since ∆n P is degree zero and applying ∆ to constants leaves zero. Thus
SMT 2010 Algebra Solutions February 28, 2010

P (n + 1) = P (n) + (P (n + 1) − P (n))
= P (n) + ∆P (n)
= P (n) + ∆P (n − 1) + (∆P (n) − ∆P (n − 1))
= P (n) + ∆P (n − 1) + ∆2 P (n − 1)
= P (n) + ∆P (n − 1) + ∆2 P (n − 2) + (∆2 P (n − 1) − ∆2 P (n − 2))
= P (n) + ∆P (n − 1) + ∆2 P (n − 2) + ∆3 P (n − 2)
= ···
Xn
= ∆i P (n − i) + ∆n+1 P (0)
i=0
n
X
= 2i 3n−i
i=0
= 3n+1 − 2n+1 .

9. Answer: 31
Factor the equation as (x + 2)(y − 5) + 10 = 30, or (x + 2)(y − 5) = 20. x must be 2 less than a factor
of 20. The solutions for x are thus 2, 3, 8, and 18, which sum to 31.
2
10. Answer: 1005
We can rewrite this equation as

x2 x2 x2 x2
+ + + =
x2 − 1 x2 − 2 x2 − 3 x2 − 4
1 + (x2 − 1) 2 + (x2 − 2) 3 + (x2 − 3) 4 + (x2 − 4)
+ + +
x2 − 1 x2 − 2 x2 − 3 x2 − 4
= (2010x − 4) + 4 = 2010x.

We divide by x; this makes us lose the solution x = 0, but this does not affect the sum of solutions.
Therefore, we have
x x x x
+ + + = 2010
x2 − 1 x2 − 2 x2 − 3 x2 − 4
Clearing denominators yields the polynomial equation

x((x2 − 2)(x2 − 3)(x2 − 4) + (x2 − 1)(x2 − 3)(x2 − 4)+

(x2 − 1)(x2 − 2)(x2 − 4) + (x2 − 1)(x2 − 2)(x2 − 3))
= 2010(x2 − 1)(x2 − 2)(x2 − 3)(x2 − 4)

The solutions that we want are therefore the roots of the polynomial

2010x8 − 4x7 + (lower order terms) = 0

4
By Vieta’s formulas, the sum of the roots of this polynomial equation is therefore .
2010
SMT 2010 Geometry Test February 28, 2010

Note: Figures may not be drawn to scale.

1. Find the reflection of the point (11, 16, 22) across the plane 3x + 4y + 5z = 7.
2. Find the radius of a circle inscribed in a triangle with side lengths 4, 5, and 6.
3. Find the volume of a regular cubeoctahedron of side length 1. This is a solid whose faces comprise 6
squares and 8 equilateral triangles, arranged as in the diagram below.

4. Given triangle ABC. D lies on BC such that AD bisects ∠BAC. Given AB = 3, AC = 9, and
BC = 8. Find AD.
5. Find the sum of angles A, B, C, D, E, F, G, H, I in the following diagram:
A

F E

6. In the diagram below, let OT = 25 and AM = M B = 30. Find M D.

SMT 2010 Geometry Test February 28, 2010

A
B

T
O

7. 4ABC is a triangle with AB = 5, BC = 6, and CA = 7. Squares are drawn on each side, as in the
image below. Find the area of hexagon DEF GHI.

I
H
A D
G
C
B

F E

8. A sphere of radius 1 is internally tangent to all four faces of a regular tetrahedron. Find the tetrahe-
dron’s volume.
9. For an acute triangle 4ABC and a point X satisfying ∠ABX + ∠ACX = ∠CBX + ∠BCX, find the
minimum length of AX if AB = 13, BC = 14, and CA = 15.

10. A, B, C, D are points along a circle, in that order. AC intersects BD at X. If BC = 6, BX = 4,

XD = 5, and AC = 11, find AB.
SMT 2010 Geometry Solutions February 28, 2010

Note: Figures may not be drawn to scale.

1. Answer: (−13, −16, −18)
The normal to the plane is in the direction < 3, 4, 5 > and so the line going through the point
perpendicular to the plane is (11 − 3t, 16 − 4t, 22 − 5t) which intersects the plane at t = 4 and hence
the reflection of the point occurs at t = 8, since the original point is at t = 0.
√
7
2. Answer: 2
q
15 15 7 5 3
First, use Heron’s Formula to find the area. The semiperimeter is s = 2 , so the area is 2 · 2 · 2 · 2 =
√ √
15 7 A 7
4 . Now the area is equal to the inradius times the semiperimeter, so r = s = 2 .
√
5 2
3. Answer: 3
√
The lengths of the sides of the large cube containing the cubeoctahedron are 2, so √
the√volume
 √ of √
the
√ 1 1 1 2 2 2 2
containing cube is 2 2. The volumes of the removed pyramids are 3 BH = 3 2 2 2 2 = 24 .
√ √
Because there are 8 pyramids removed, the total volume removed is 8 242 = 2
3 . Thus, the total volume
√ √ √
of the cubeoctahedron is 2 2 − 32 = 5 3 2 .
√
4. Answer: 15
By angle bisector theorem, AB BD 3
AC = DC ⇒ 9 =
BD
8−BD ⇒ 24 − 3BD = 9BD. This implies BD = 2 and
DC = 6. Now apply the Stewart’s theorem

AB 2 · CD + AC 2 · BD = BC · (AD2 + BD · BC),

then we have AD2 = 15.

5. Answer: 540◦
The path goes around the center three times. At each turning point, the external angle at that point
is the amount that you turn around the center. Hence, the external angles add up to 3 · 360◦ = 1080◦ .
Hence, the answer we want is 9 · 180◦ − 1080◦ = 540◦ .
√
25 13
6. Answer: 3
√
Let N be the opposite point of M in the circle. Then M N = 50 and N B = 502 − 302 = 40 because
4M BN is right triangle. Let C be the midpoint of AB, then 4M CB and 4M BN are similar, so
MB MB
BC = N B · M N = 24, M C = M B · M N = 18. Let L be the intersection of AB and the tangent.
Since we have AB and OT parallel, CL = OT = 25,√so BL = 1. Since 4M CB ∼ 4BLD, we have
√
BD = M B · MBL 5
C = 3 , so M D = M B 2 + BD2 = 25 3 13 .
√
7. Answer: 110 + 24 6
We can calculate the area of the middle triangle using Heron’s p formula. Hence, we can calculate √the
semiperimeter of the triangle, 9, and then calculate the area as 9 × (9 − 5) × (9 − 6) × (9 − 7) = 6 6.
Notice that ∠BCA and ∠ECD are supplementary. Hence, we can rotate 4ECD about point C so
that segments AC and CD overlap and the resulting figure BAE will be a triangle. In this position,
we can see that 4BAC and 4DCE have the same altitude, and since they have the same base (length
7) they must have the same area. By the same reasoning, all the triangles must have the same area.
Hence, the total area of the figure is simply
√ the areas√of the squares plus four times the area of the
middle triangle 25 + 36 + 49 + 4 × 6 × 6 = 110 + 24 6.
√
8. Answer: 8 3
The center of the sphere is located at the centroid of the tetrahedron, which is located 14 of the way up
the altitude from a face to the opposite vertex. In√
other words, the tetrahedron has height 4. Let its
edge length be s. Then the altitude of a face is s 23 , and the distance from the centroid of a face to a
SMT 2010 Geometry Solutions February 28, 2010

√
vertex is 32 of that, which is 33 . This length and the height of the tetrahedron form a right triangle,
with an edge as the hypotenuse. That is, 31 s2 + 16 = s2 . Thus s2 = 24, and so the area of a face is
√ √ √
6 3. The volume is 13 · 6 3 · 4 = 8 3.
√
9. Answer: 65

X
I

B C

Since ∠ABX + ∠ACX and ∠CBX + ∠BCX add up to ∠B + ∠C, we can see that ∠CBX + ∠BCX =
∠B+∠C
2 , so ∠BXC = 90 + ∠A 2 . Thus X should lie on a fixed circle, which also goes through I, the
incenter of the triangle.
Let P be the center of that circle. We have ∠BP C = 2(180 − ∠BXC) = 180 − ∠A, so P lies on the
circumcircle of 4ABC. And from BP = CP we have ∠BAP = ∠CAP , so A, I, P lie on the same
line. Thus X = I minimizes AI.
2S
Then it is the problem of finding AI. First we find the radius r of incircle, which is a+b+c = 2·84 =
p 13+14+15
2 2
4.
√ Suppose AB √ and the incircle meets at D. Then AD = (AB +AC −BC)/2 = 7, AI = AD + r =
2 2
7 + 4 = 65.
10. Answer: 6
Notice that ABCD is a cyclic quadrilateral. Set AB = x and AE = y, so that CE = 11 − y. We apply
the Power of a Point Theorem at the point E to get y(11 − y) = 4 · 5, so that 11y − y 2 = 20.
We have two pairs of similar triangles: 4ABE ∼ 4DCE and 4ADE ∼ 4BCE; these yield AD = 23 y
and CD = 5xy . Applying Ptolemy’s Theorem to quadrilateral ABCD then yields

5x 3
x· + 6 · y = 9 · 11
y 2
Clearly the denominators and rearranging, we see that the y terms are precisely those given by the
Power of a Point Theorem:
5x2 = 99y − 9y 2 = 9(11y − y 2 ) = 9 · 20 = 180,
and therefore x = 6.
SMT 2010 Calculus Test February 28, 2010

1. Evaluate     
1 −1 1 −1 1
lim tan − tan
t→0 t x+t x
3
2. Find the minimum value of ex − x − x3 .
Z ∞  p  Z ∞
3. Given etx f (x) dx = sin−1 t − 1/2 , find xf (x) dx.
−∞ −∞
 
 3x + 1 
4. Find the values of x that maximize f (x) =  .
9x2 + 6x + 2 
5. A rectangular pyramid tower is being built on a circular island of radius two. The height of the tower
is equal to its width. What is the maximum volume of the tower?
6. Evaluate
∞
X 13k
ke−13 .
k!
k=0
"Z #
ln 1/t
d
7. Calculate cos(tex )dx .
dt − ln 1/t

8. Evaluate
1 1 2 1 1 2 1 1 2
+ − + + − + + − + ···
1 2 3 4 5 6 7 8 9
9. Find the value of k which minimizes
Z 4
F (k) = |x(4 − x) − k| dx.
0

10. Let f (x) = x6 − 6x2 + 6x − 7. It is known that this polynomial has three critical points. Find the
parabola passing through these critical points.
SMT 2010 Calculus Solutions February 28, 2010

−1
1. Answer: x2 +1
Notice that as t → 0, both the numerator and the denominator approach 0. Thus, applying L’Hopital’s
rule on t (keeping x constant):
  
d −1 1 1
tan =−
dt x + t t=0 1 + x2

2. Answer: 1
3
Let f (x) = ex − x − x3 . Then f 0 (x) = ex − 1 − x2 . When x < 0, ex < 1 and 1 + x2 > 1, so f 0 (x) =
ex −(1+x2 ) < 0. Thus, f is decreasing on (−∞, 0). When x = 0, f 0 (x) = f 0 (0) = e0 −1−02 = 1−1 = 0.
Finally, for x > 0, f 0 (x) = ex − 1 − x2 > 0 by a Maclaurin series expansion, so f is increasing on (0, ∞).
3
Thus, f must attain its minimum when x = 0, at which point f has the value e0 − 0 − 03 = 1.
√
3. Answer: 2
Consider:
d ∞ tx
 Z  Z ∞  Z ∞
d −1
p   tx

sin (t − 1/2)
 = e f (x) dx
 = xe f (x) dx = xf (x) dx
dt t=0 dt −∞ t=0 −∞ t=0 −∞


√
 
d −1
p  1  1
sin (t − 1/2)
 = r  =p = 2.
dt t=0
 p  2
 1 − (1/2)
1− 1/2 − t 
t=0

4. Answer: x = − 23 and x = 0
Notice that f (x) → 0 as x → ±∞. Since 9x2 + 6x + 2 has no real roots, the maximum value of f (x)
is attained at the maximum of the absolute values of the critical points of 9x23x+1
+6x+2 .

The extrema of 9x23x+1 2

+6x+2 occur at x = − 3 and x = 0. It is easily checked that maxima of f (x) occur
at both of these points.
√
128 3
5. Answer: 27
Let the circular island be a circle of radius 2 centered at the origin. Without loss of generality, let
the length of the rectangular base be from −x to x and the width from −y to y. Notice that by the
equation of a circle, x2 = 4 − y 2 . Then
1 8 8 8
V =(2x)2 (2y) = x2 y = (4 − y 2 )y = (4y − y 3 )
3 3 3 r 3
dV 8 4
= (4 − 3y 2 ) = 0 =⇒ y =
dy 3 3
 r √
8 8 4 128 128 3
V = = √ = .
3 3 3 9 3 27

6. Answer: 13
This is the evaluation of the mean of a Poisson distribution: for any λ,
∞ k ∞ k ∞ ∞
X
−λ λ
X
−λ λ
X
−λ λk−1 −λ
X λm
ke = ke =λ e = λe = λe−λ eλ = λ.
k! k! (k − 1)! m=0
m!
k=0 k=1 k=1

−2 cos(t2 )
7. Answer: t
SMT 2010 Calculus Solutions February 28, 2010

By the Leibniz integral rule, the above integral becomes

Z ln 1/t
−ex sin(tex ) dx + cos(teln(1/t) )(−1/t) − cos(te− ln(1/t) )(1/t)
− ln 1/t
ln 1/t
cos(tex )  cos(1) + cos(t2 )
= −
t 
− ln 1/t t
−2 cos(t2 )
= .
t
8. Answer: ln 3
The partial sums of this sum are equal to
   
1 1 1 1 1 1
+ + ··· + −3 + + ··· +
1 2 3n 3·1 3·2 3·n
 
1 1 1 1 1 1 1
= + + ··· + = + + ··· +
n+1 n+2 3n n 1 + n1 1 + n2 1 + 2n
n
This is a Riemann sum, so as n → ∞ the partial sums converge to
Z 2
1
dx = ln 3.
0 1 + x
9. Answer: 3
4

1 2 3 4

As you can see it from this graph, F (k) is the area of region that “lies between” y = f (x) = x(4 − x)
and y = k. Let A be the region below y = f (x) and above y = k, and B be the region below y = k
and above y = f (x). Then F (k) = A + B. Meanwhile, we can find the area of A by integrating with
respect to y-variable. Since x belongs to the interval of length l(t) when y = t, we can say
Z 4
A= l(t) dt.
k
Apply the same reasoning to B, then we have
Z k
B= (4 − l(t)) dt.
0
Thus, by the fundamental theorem of calculus,
d d d
F (k) = A+ B
dk dk dk !
Z 4  Z k
d d
= l(t) dt + (4 − l(t)) dt
dk k dk 0

= −l(k) + (4 − l(k))
= 4 − 2l(k).
SMT 2010 Calculus Solutions February 28, 2010

d
Since l(k) is decreasing by k, F (k) achieves minimum when dk F (k) = 0. One can easily find that
k = 3 if l(k) = 2, so the answer is 3.
10. Answer: y = −4x2 + 5x − 7
Such a parabola intersects f (x) precisely where f 0 (x) = 0. Hence, the value of the intersection points do
not change when we replace f (x) by f (x)+g(x)f 0 (x) for any g(x). Therefore, since f 0 (x) = 6x5 −12x+6,
we must have that f (x) − 1/6xf 0 (x) = −4x2 + 5x − 7 passes through the three critical points. Since
three points determines a parabola uniquely, this must be the unique parabola passing through the
three critical points.
SMT 2010 Advanced Topics Test February 28, 2010

1. Find sin 18◦ .

2. The quarter-imaginary positional number system has radix 2i and uses the digits 0, 1, 2, and 3. It
can be used to represent any complex number without using a negative sign. For example, 10322i =
1(2i)3 + 0(2i)2 + 3(2i) + 1(2i)0 = 1 − 2i. Compute the base-10 representation of 3.01232i .

3. What is the smallest number of people that can be invited to the 2010 Stanford Mathematics Tour-
nament such that either three of them met each other last year, or three of them did not meet each
other last year?
Pn
4. Express cos5 θ in the form i=1 ai cos(kθ) for some positive integers k and n and real numbers ai .
5. Compute
2010
X 2010
X 
i
.
j=0 i=j
j

6. In an n-by-m grid, 1 row and 1 column are colored blue, the rest of the cells are white. If precisely
1
2010 of the cells in the grid are blue, how many values are possible for the ordered pair (n, m)?

7. A bug either splits into two perfect copies of itself or dies. If the probability of splitting is p > 12 (and is
independent of the bug’s ancestry), what is the probability that a bug’s descendants die out? Express
your answer as a function in terms of p.
8. Find all pairs of positive integers (x, y) such that 2x + 1 = 3y , and y is not divisible by 4.
9. How many ordered pairs of complex numbers (x, y) satisfy

1 1
x2 + y 2 = + = 9?
x y
2
10. Compute the product of all positive integers such that b n5 c is prime.
SMT 2010 Advanced Topics Solutions February 28, 2010

√
5−1
1. Answer: 4
.
Consider:

sin 18◦ = cos 72◦ = 2 cos2 36◦ − 1 = 2(1 − 2 sin2 18◦ )2 − 1

= 1 − 8 sin2 18◦ + 8 sin4 18◦
0 = 8x4 − 8x2 − x + 1 = (x − 1)(2x − 1)(4x2 + 2x + 1).

Clearly, x 6= 1, −1 ◦ ◦
2 , because 0 < sin 18 < sin 90 = 1. We solve the remaining term:
p √
2 −2 ± 4 + 4(4) −1 ± 5
0 = 4x + 2x + 1 =⇒ x = = .
2(4) 4
√
5−1
The only root that is within our bounds is 4 .
44 4
2. Answer: 15
+ 15
i.
Consider:
∞
X ∞
X ∞
X ∞
X
3.0123 = 3 + 0(2i)−(4n+1) + 1(2i)−(4n+2) + 2(2i)−(4n+3) + 3(2i)−(4n+4)
n=0 n=0 n=0 n=0
∞   n n  n 
i 1n

X 1 1 3 1
=3+ − + +
n=0
4 16 4 16 16 16
∞     n
X 1 1 1
=3+ − + i
n=0
16 4 16
 
1 1 1
=3+ − + i 1
16 4 1 − 16
 
16 1 1
=3+ − + i
15 16 4
44 4
= + i.
15 15

3. Answer: 6.
This is a trivial application of Ramsey Theory. It is isomorphic to Problem 2 of the Power Round.
Consider one of the people, P , in the group, and that he or she may or may not have met last year.
Assume without loss of generality that P met at least three of them last year: A, B, and C. If any
two of these met each other last year, then those two and P all met each other last year. Alternatively,
none of A, B, and C met each other last year.
1 5 5
4. Answer: 16
cos 5θ + 16
cos 3θ + 8
cos θ
Notice that
n
cos(nθ) + i sin(nθ) = (cos θ + i sin θ) = z n
and
−n
cos(−nθ) + i sin(−nθ) = (cos θ + i sin θ) = z −n .
Adding these two equations, we get that cos(nθ) = (z n + z −n )/2. Then (cos(θ))5 = (z + z −1 )5 /32.
Expanding yields the binomial coefficients:

(z + z −1 )5 = z 5 + 5z 4 (z −1 ) + 10z 3 (z −2 ) + 10z 2 (z −3 ) + 5z(z −4 ) + z −5 .

Then
1 5 5 5 1 5 5
(z + z −1 )5 /32 = (z + z −5 )/2 + (z 3 + z −3 )/2 + (z + z −1 )/2 = cos 5θ + cos 3θ + cos θ.
16 16 8 16 16 8
SMT 2010 Advanced Topics Solutions February 28, 2010

5. Answer: 22011 − 1
We evaluate the inner sum by the Hockey Stick Identity. This identity is
n     2010
X   
X i n+1 i 2011
= =⇒ = ,
i=r
r r+1 i=j
j j+1

so that
2010
X 2010
X  2010
X 
i 2011
= .
j=0 i=j
j j=0
j+1

Now, using the fact that

n  
X n
= 2n ,
i=0
i
we obtain
2010
X  2011
X  2011
X 2011 2011
2011 2011
= = − = 22011 − 1.
j=0
j+1 j=1
j j=0
j 0

6. Answer: 96
The number of blue cells is n + m − 1; the number of total cells is nm. So 2010(m + n − 1) = nm,
or nm − 2010n − 2010m + 2010 = 0. This factors as (n − 2010)(m − 2010) − 20102 + 2010 = 0, or
(n−2010)(m−2010) = 2010·2009. Thus each of n−2010 and m−2010 must be one of the positive factors
of 2010 ∗ 2009; for each positive factor, there is one ordered pair. Since 2010 · 2009 = 2 · 3 · 5 · 72 · 41 · 67,
there are 2 · 2 · 2 · 3 · 2 · 2 = 25 · 3 = 96 solutions.
1
7. Answer: p
−1
Let the probability that a bug’s descendant’s die out be x. There are two ways for the bugs to die out:
either the initial bug dies (with probability 1 − p), or the bug successfully splits (probability p) and
both of its descendants die out (each with probability x). Therefore, x = (1 − p) + px2 . Solving this
quadratic equation yields the two solutions x = 1 and x = p1 − 1. Which is correct?
Define pn to be the probability that the bug dies out within n generations. Then, by the same reasoning
as before, pn+1 = (1−p)+pp2n . From the definition of pn , we see that the sequence is always increasing.
We will show that pn < p1 − 1 for every n, which would imply that x = p1 − 1 is the correct solution.
This can be done by induction. Notice that p0 = 0 < p1 − 1. Now, suppose that pk < p1 − 1 for some
k. Then,
 
1 1 1 1 1
pk+1 = (1 − p) + pp2k < (1 − p) + p( − 1)2 = 1 − p + p − 2 + 1 = 1 − p + − 2 + p = − 1.
p p2 p p p
1
This completes the induction, so we indeed have pn < p − 1, and hence the correct answer is indeed
1
p − 1.

8. Answer: (1, 1) and (3, 2)

From that 31 ≡ 3, 32 ≡ 9, 33 ≡ 11, 34 ≡ 1 mod 16, we can see that 3y 6≡ 1 mod 16. Thus 2x 6≡ 0
mod 16, so x is at most 3. The solutions are (1, 1) and (3, 2).
9. Answer: 4
If (x, y) is a solution, then x and y are roots of the quadratic equation t2 + at + b = (t − x)(t − y) = 0.
Then −a = x + y and b = xy. We find that x2 + y 2 = a2 − 2b = 9 and x1 + y1 = x+y a
xy = − b = 9.
So a = −9b. Substituting this in the first equation, we get 81b2 − 2b − 9 = 0. This has two roots for
b, both of them real. Therefore there are two corresponding values of a, both real. In each case, the
quadratic leads to two ordered pairs, which gives four total ordered pairs. It is easy to check that they
are, indeed, distinct.
SMT 2010 Advanced Topics Solutions February 28, 2010

10. Answer: 120

Note that n2 ≡ 0, 1, 4 mod 5. We consider three cases.
2 2
Case 1: n2 ≡ 0 mod 5, so that b n5 c = n5 . In this case, n ≡ 0 mod 5, so n = 5a for some integer a.
2
Then n5 = 5a2 , which is not prime unless a = 1. Therefore, for this case, n = 5 is the only value of n
2
for which b n5 c is prime.
2 2
Case 2: n2 ≡ 1 mod 5, so that b n5 c = n 5−1 = (n−1)(n+1)
5 . In this case, we have either n = 5a + 1 or
n2
n = 5a − 1 for some integer a. Then 5 = a(n ± 1), which cannot be prime if a 6= 1. Therefore, for
2
this case, n = 4, 6 are the only values of n for which b n5 c might be prime. We can check that these
values of n do indeed yield primes 3 and 7.
2 2
Case 3: n2 ≡ 4 mod 5, so that b n5 c = n 5−4 = (n−2)(n+2)
5 . In this case, we have either n = 5a + 2 or
2
n = 5a − 2 for some integer a. Then n5 = a(n ± 2), which cannot be prime if a 6= 1. Therefore, for this
2
case, n = 3, 7 are the only values of n for which b n5 c might be prime. None of these values actually
2
yield primes however, as they give b n5 c = 1, 9.
2
Therefore, the only values of n for which b n5 c is prime are n = 4, 5, 6, and the product of these values
of n is 120 .
Generalization: Note that this procedure can be carried out when the denominator 5 is replaced by any
other number whose quadratic residues are all perfect squares. Which numbers satisfy this property?
SMT 2010 General Solutions February 28, 2010

1. Answer: 2
Let the set of coins be {A, B, C, D, E, F, G, H}. First, weigh {A, B, C} vs. {D, E, F }. If one group is
lighter (say {A, B, C}), then use the second weighing to measure A vs B. If we find that one is lighter
than the other, than it must be the counterfeit. If A vs B weigh the same, then we know that C must
be the counterfeit. If both groups weighed the same, then measure G vs H. If G vs H weigh the same,
then no counterfeit exists.
2. Answer: 29
Just try a bunch.
3. Answer: 3
We note that to get a zero at the end of a number, we must multiply by 10. Since 5 × 2 = 10,
and there are more factors of 2 in 200! than 5, it suffices to count how many 5’s appear in the prime
factorization of 200. Each of 5, 10, 15, . . . , 200 has a factor of 5 in it, which gives 40 factors. In addition,
25, 50, . . . , 200 gives a second factor of 5, so that is 8 additional factors of 5. 125 has a third factor of
5. So 200! has 49 factors of 5; it ends with 49 zeros.
Similarly, 124! ends with 28 zeros and 76! ends with 18 zeros. Therefore,
 
200 200!
=
124 124! 76!

ends in 49 − 28 − 18 = 3 zeros.
√
4. Answer: 1+2 5
√ √
q p p
Let x = 1 + 1 + 1 + .... Then x2 = 1 + 1 + 1 + .... Thus x2 = x + 1. The positive root of
√
1+ 5
x2 − x − 1 = 0 is 2 .

5. Answer: 19801 and 20201

Notice that 4x4 + 1 = 4x4 + 4x2 + 1 − (2x)2 = (2x2 + 2x + 1)(2x2 − 2x + 1). Setting x = 100, we have
that 400000001 = 19801 · 20201.
√
6. Answer: 14 5
7+9+12
We can use Heron’s formula to calculate the area of the triangle. The semiperimeter equals 2 = 14.
Applying Heron’s formula
p √
Area = 14 · (14 − 7) · (14 − 9) · (14 − 12) = 14 5
7. Answer: 67

Observe that if vertices V1 , V2 , . . . , Vn is a complete list of vertices that contain an edge leading into
a vertex X and it is known that there are p1 , p2 , . . . pn different paths into the vertices V1 , V2 , . . . , Vn ,
respectively, then the number of paths into vertex X is p1 + p2 + . . . pn . Using this rule, we can start
from the vertex A which trivially has 1 [empty] path into it and proceed inductively along the grid to
get the following diagram, which ends with 67 paths on vertex B.
SMT 2010 General Solutions February 28, 2010

√
a±a 5
8. Answer: x = 2

a2 = x2 − ax
0 = x2 − ax − a2
p
a ± a2 − 4 (−a2 )
x=
√2 √
a ± |a| 5 a±a 5
x= =
2 2
The ± takes care of the fact that we do not know the sign of a.

9. Answer: 180 meters

The regular trains will intersect at (150, 0) meters at t = 3 seconds. Regardless of the location of the
fly at any time, we know that the fly will be moving at 60 meters/second. Therefore, at t = 3 seconds
the fly will have traveled a total of 3 · 60 = 180 meters.

10. Answer: 100000000

4

P4 4


Note that the digits of 14641n in base n are the binomial coefficients i so that 14641n = i=0 i ni =
(n + 1)4 . When n = 99, this is simply 100000000.
√
11. Answer: 150 3
This regular hexagon has side length 10. Notice that a regular hexagon can be split up into 6 equilateral√
2
triangles, each with a side length of 10. The area of an equilateral triangle with side length 10 is 10 4 3
2
√ √
and hence the area of the hexagon will be 10 4 3 · 6 = 150 3
12. Answer: 1027
If S(n) is the nth partial sum, note that if m is the kth triangular number, S(m) = k 2 . Since 442 = 1936
and 452 = 2025, we want to begin our search at 44(44 + 1)/2 = 990. Because (2010 − 1936)/2 = 37,
37 more 2s are needed, so the needed term is n = 990 + 37 = 1027.

13. Answer: 21,26,31,36,41,46

6x + 5 ≡ −19 mod 10
6x ≡ −24 mod 10
10
x ≡ −4 mod
gcd(10, 6)
x ≡ −4 mod 5
x≡1 mod 5

That is, x is in the form 5k + 1 where k is an integer.

14. Answer: 17
The open lockers will be the ones with an odd number of odd divisors. These numbers are of the form
2k · n2 , where n is odd. We can simply check that the open lockers are numbered

1, 2, 4, 8, 9, 16, 18, 25, 32, 36, 49, 50, 64, 72, 81, 98, 100.
SMT 2010 General Solutions February 28, 2010

26
15. Answer: 15
p
√ √
p− 3q p2 −3q 2
Since q − 3= q = q(p+ √
3q)
, we should look for (p, q) which minimizes |p2 − 3q 2 |. There are
no solutions for p − 3q = −1 (consider mod 3) and for p2 − 3q 2 = 1, we have (2, 1), (7, 4), (26, 15)
2 2

as solutions where q ≤ 15. Since q = 15 also maximizes the denominator, (p, q) = (26, 15) is the best
choice.
16. Answer: 21 revolutions
The wheel goes through three laps of 2π · 7 = 14π distance, for a total distance of 15 · 14π = 210π. In
one full turn, the wheel goes through a distance of 2π 5̇ = 10π and thus after all three laps the wheel
undergoes 210π
10π = 21 revolutions.
√
17. Answer: R 3
Label the center of the circle O and the vertices of the triangle A, B, and C. Then we can find the
length of segment AB by using the law of cosines on triangle AOB. By symmetry, we can see that the
angle AOB equals 3603 = 120 degrees. Applying the law of cosines
(Length of AB) = R2 + R2 − 2 · R · R · cos(120◦ ) = 2R2 − 2R2 · − 12 = 3R2
2

18. Answer: 96
The number of blue cells is n + m − 1; the number of total cells is nm. So 2010(m + n − 1) = nm,
or nm − 2010n − 2010m + 2010 = 0. This factors as (n − 2010)(m − 2010) − 20102 + 2010 = 0, or
(n−2010)(m−2010) = 2010·2009. Thus each of n−2010 and m−2010 must be one of the positive factors
of 2010 · 2009; for each positive factor, there is one ordered pair. Since 2010 · 2009 = 2 · 3 · 5 · 72 · 41 · 67,
there are 2 · 2 · 2 · 3 · 2 · 2 = 25 · 3 = 96 solutions.
19. Answer: x = −2, −1, − 12 , 2
3
Simple trial and error of the first few integers yields the roots x = −2, −1. From there, the remaining
polynomial can be solved using the quadratic formula. The final factored form is

(2 + x)(1 + x) 6x2 − x − 2 = (2 + x)(1 + x)(1 + 2x)(−2 + 3x)

20. Answer: 20
Each circle can intersect with any other circle in at most two distinct points. As long as the inter-
section points between every two circles are all unique, then we can calculate the greatest number of
intersections possible starting with the case of two circles and working up to five circles:
2 circles → 2 points max
3 circles → 2 + 2 · 2 = 6 points max
4 circles → 6 + 2 · 3 = 12 points max
5 circles → 12 + 2 · 4 = 20 points max

21. Answer: 1
Factor it as (x2 − 1)(x2 + 1) = 2y 2 . Since x2 − 1 and x2 + 1 differ by 2, their gcd is 1 or 2. In either
case both x2 − 1 and x2 + 1 has even powers of all odd primes, so x2 − 1 = m2 , x2 + 1 = 2n2 or
x2 − 1 = 2m2 , x2 + 1 = n2 . In the first case x2 − m2 = 1, so we only have (x, m) = (1, 0) possible,
giving the answer (x, y) = (1, 0). In the second case n2 − x2 = 1, so x = 0. But this is impossible since
then x4 − 2y 2 = −2y 2 would not be 1. Thus we only have one solution (x, y) = (1, 0).
√
22. Answer: 12 5
Note that 1010100φ − .010101φ = φ6 + φ√
4
+ φ2 − φ−2 − φ−4 − φ−6 =√φ6 + φ4 + φ2 − (−φ)−2 −4
√ − (−φ) −
−6
(−φ) . Thus, 1010100φ − .010101φ = 5(F (6) + F (4) + F (2)) = 5(8 + 3 + 1) = 12 5.

23. Answer: 3
Note that the function is antisymmetric, so all three terms in the numerator are equal.
SMT 2010 General Solutions February 28, 2010

24. Answer: 6029

If we consider a single 1 × 1 square, and find two regions within it on which the center of the coin of
radius 14 can land — the center 12 × 12 square, which has area 14 , and the outside edge, where an overlap
will occur, of area 34 .
The total area that the center of the coin can land on is thus
40192
  
1 1
2010 − 2010 − = .
2 2 4
20102
Thus, the probability is 40192 , so a + b = 6029.
25. Answer: 7
Let 1, 2, 3, 4, 5 be five balls.
Compare 1, 2, Without Loss of Generality (WLOG) 1 < 2
Compare 3, 4, WLOG 3 < 4
Compare 1, 3, WLOG 1 < 3
Compare 3, 5.
(a) If 3 < 5, there are eight remaining cases: compare 4, 5 (WLOG 4 < 5), compare 2, 4
i. If 2 < 4, compare 2, 3; done
ii. If 4 < 2, compare 2, 5; done
(b) If 3 > 5, there are seven remaining cases: compare 2, 3.
i. If 2 < 3, compare 1, 5. If 5 < 1 we are done, if 1 < 5, compare 2, 5; done.
ii. If 2 > 3, compare 1, 5, and then compare 2, 4; done.
SMT 2010 General Test February 28, 2010

1. Given 8 coins, at most one of them is counterfeit. A counterfeit coin is lighter than a real coin. You
have a free weight balance. What is the minimum number of weighings necessary to determine the
identity of the counterfeit coin if it exists?
2. Find the smallest prime p such that the digits of p (in base 10) add up to a prime number greater than
10.

3. How many zeros are there at the end of 200

124 ?
v s
u r
√
u q p
4. Compute 1 + 1 + 1 + 1 + 1 + 1 + ....
t

5. Alice sends a secret message to Bob using her RSA public key n = 400000001. Eve wants to listen in
on their conversation. But to do this, she needs Alice’s private key, which is the factorization of n.
Eve knows that n = pq, a product of two prime factors. Find p and q.
6. A triangle has side lengths 7, 9, and 12. What is the area of the triangle?
7. How many paths are there from A to B in the directed graph below?

a x−a
8. Find all solutions of = for x.
x a
9. A straight line connects City A at (0, 0) to City B, 300 meters away at (300, 0). At time t = 0, a bullet
train instantaneously sets out from City A to City B while another bullet train simultaneously leaves
from City B to City A going on the same train track. Both trains are traveling at a constant speed of
50 meters/second. Also at t = 0, a super fly stationed at (150, 0) and restricted to move only on the
train tracks travels towards City B. The fly always travels at 60 meters/second, and any time it hits
a train, it instantaneously reverses its direction and travels at the same speed. At the moment the
trains collide, what is the total distance that the fly will have traveled? Assume each train is a point
and that the trains travel at their same respective velocities before and after collisions with the fly.
10. Compute the base 10 value of 1464199 .

11. What is the area of the regular hexagon with perimeter 60?
12. Consider the sequence 1, 2, 1, 2, 2, 1, 2, 2, 2, 1, 2, 2, 2, 2, 1, ... Find n such that the first n terms sum up
to 2010.
13. Find all the integers x in [20, 50] such that 6x + 5 ≡ −19 mod 10, that is, 10 divides (6x + 15) + 19.

14. A series of lockers, numbered 1 through 100, are all initially closed. Student 1 goes through and opens
every locker. Student 3 goes through and “flips” every 3rd locker (“flipping”) a locker means changing
its state: if the locker is open he closes it, and if the locker is closed he opens it). Thus, Student 3
will close the third locker, open the sixth, close the ninth. . . . Student 5 then goes through and “flips”
every 5th locker. This process continues with all students with odd numbers n < 100 going through
and “flipping” every nth locker. How many lockers are open after this process?
√
15. Find the best approximation of 3 by a rational number with denominator less than or equal to 15.
SMT 2010 General Test February 28, 2010

16. A wheel is rolled without slipping through 15 laps on a circular racecourse with radius 7. The wheel
is perfectly circular and has radius 5. After the three laps, how many revolutions around its axis has
the wheel been turned through?
17. An equilateral triangle is inscribed inside of a circle of radius R. Find the side length of the triangle.

18. In an n-by-m grid, 1 row and 1 column are colored blue, the rest of the cells are white. If precisely
1
2010 of the cells in the grid are blue, how many values are possible for the ordered pair (n, m)?

19. Find the roots of 6x4 + 17x3 + 7x2 − 8x − 4

20. Given five circles of radii 1, 2, 3, 4, and 5, what is the maximum number of points of intersections
possible (every distinct point where two circles intersect counts).

21. How many nonnegative integer solutions are there for x4 − 2y 2 = 1?

22. We need not restrict our number system radix √
to be an integer. Consider the phinary numeral system
in which the radix is the golden ratio φ = 1+2 5 and the digits 0 and 1 are used. Compute 1010100φ −
.010101φ .
23. Let f (X, Y, Z) = X 5 Y − XY 5 + Y 5 Z − Y Z 5 + Z 5 X − ZX 5 . Find

f (2009, 2010, 2011) + f (2010, 2011, 2009) − f (2011, 2010, 2009)

.
f (2009, 2010, 2011)

24. We are given a coin of diameter 12 and a checkerboard of 1 × 1 squares of area 2010 × 2010. We toss
the coin such that it lands completely on the checkerboard. If the probability that the coin doesn’t
2
touch any of the lattice lines is ab2 where ab is a reduced fraction, find a + b.

25. There are five balls that look identical, but their weights all differ by a little. We have a balance that
can compare only two balls at a time. What is the minimum number of times, in the worst case, we
have to use to balance to rank all balls by weight?
 2010 SMT Power Round

Definitions
A graph is a collection of points (vertices) connected by line segments (edges). In this test,
all graphs will be simple – any two vertices will be connected by at most one edge – and
connected – you can get from any vertex to any other by following edges.

A simple connected graph with 7 vertices and 11 edges.

An edge n-coloring of a graph G is an assignment of one of n colors to each edge of G.

blue
red
red blue
red red red
blue
red
blue
red
A 2-coloring of the earlier graph.

A complete graph is one in which any two vertices are connected by an edge.

1. a. (5 points) Draw a simple connected graph with 8 vertices and 7 edges, and 3-color its
edges.
Solution: Many graphs work. The only thing to note is that the graph must be a
tree, that is, it should have no cycles.
b. (5 points) Draw a complete graph on 5 vertices, and 2-color its edges so that it does
not contain a red triangle or a blue triangle (3 vertices, the edges between which are
all red or all blue).
Solution: The simplest depiction is a pentagon, with blue sides and red diagonals.

We will use Kn to denote a complete graph on n vertices. A monochromatic Kn is one in

which every edge has the same color. Hence, problem 1(b) could have been phrased “Color
K5 so that it has no monochromatic K3 ”.

2. (10 points) Show that no matter how you 2-color K6 , it will contain a monochromatic K3 .
(Hint: Think about all the edges coming from one vertex).
Solution: Pick any vertex v of K6 . Then there must be three edges of one color coming
from v, WLOG they are all blue. Now consider the three vertices on the other side of
these edges from v, and the edges between them. If any of the edges are blue, they form
a blue K3 with the blue edges from its endpoints going to v. However, if none of them
are blue, then we have a red K3 .
1
The Ramsey number R(k) is the least number n such that no matter how you 2-color the
edges of Kn , there will be a monochromatic Kk . In problems 1(b) and 2, you have shown
that R(3) = 6.
Interestingly, R(4) is a difficult quantity to calculate, and R(5) is still unknown! Since we
cannot go much further in this vein, let us try looking at generalizations of Ramsey numbers.
Define R(k, j) as the least n such that every red, blue edge 2-coloring of Kn contains either
a red Kk or a blue Kj . Then R(n) is just R(n, n) under this new definition.
3. a. (5 points) Show that R(4, 3) > 8 by exhibiting a 2-coloring.
Solution: A counterexample is given in the diagram below. Here, the graph shows
only blue edges; the edges that do not appear in the diagram are presumed to be red.

b. (15 points) Show that R(4, 3) = 9 (Hint: Use problem 2.)

Solution: Suppose that we have a 2-coloring of the edges of K9 in which there are
no blue triangles, nor red K4 . Then any vertex v can have no more than 5 red edges
coming from it, because if it had six, the points on the other end would form a K6 ,
and we already know that a K6 must contain a blue triangle – which we supposed to
be impossible – or a red triangle, which would form a red K4 with the edges from v.
Likewise, no vertex can have more than 3 blue edges coming from it, for if four blue
edges came from one vertex, the K4 on the other end would have to be all red, or have
a blue edge and complete a blue triangle. Now, since every vertex has 8 edges coming
from it, we see that every vertex must have exactly 3 blue and 5 red edges. However,
this violates the handshake theorem – basically, there are 27 requests for a blue edge,
but these can only be granted in pairs because each edge connects two vertices.

4. a. (15 points) Show that

R(n, m) ≤ R(n, m − 1) + R(n − 1, m).
(Hint: see hint to problem 2.)
Solution: Proof by induction on n + m. If n = 2, then R(n, m) = m, and likewise
2
 if m = 2. R(n, m) is not defined for smaller arguments. If we know the result for
n + m < k, let us take a complete graph on R(n − 1, m) + R(n, m − 1) vertices. Pick
any one vertex v, and consider the R(n − 1, m) + R(n, m − 1) − 1 edges coming from
it. By the pigeonhole principle, there are either R(n − 1, m) blue edges or R(n, m − 1)
red edges among these. WLOG, there are R(n − 1, m) blue ones. Now, the vertices on
the other side of those edges form a KR (n − 1, m), and so must contain either a red
Km or a blue Kn−1 , which with v forms a blue Kn .
b. (5 points) Conclude that R(n, m) is well defined, that is, that it exists for every n and
every m.
Solution: In the proof above, we showed that given m and n, every sufficiently large
graph must contain either a red Km or a blue Kn , which is precisely the condition for
R(n, m) to exist.
From here on, we will explore some interesting properties and generalizations of Ramsey
numbers. Each section is independent.
Bounds on Ramsey Numbers
5. Color a graph of n2 points, laid out in a n × n grid, as follows: The edge (u, v) is blue if
u and v are in the same row, and red otherwise.
a. (5 points) Show that any Kn+1 in that graph contains at least one red edge and at
least one blue edge.
Solution: By pigeonhole, there must be two points of the Kn+1 which are in the same
row, and so have a blue edge between them. Similarly, there must be two not in the
same row, and so those two will have a red edge between them.
b. (5 points) Conclude that R(n + 1, n + 1) > n2 .
Solution: We just gave an example of a graph that has n2 vertices, but no monochro-
matic Kn . Hence R(n, n) is greater than or equal to n2 .
Problem 6 gives us a polynomial lower bound for R(n, n), and it does so constructively –
we know exactly which graph will give a counterexample. Erdős has shown that, if we are
willing to be nonconstructive, we can get a much better lower bound:
6. a. (5 points) Show that if the edges of Km are colored red or blue randomly with equal
probability (i.e., by flipping a coin for each edge), then the probability that it contains
a monochromatic Kn is at most
 
m n
· 21−( 2 ) .
n

Solution: The probability that n vertices chosen at random form a monochromatic

n
Kn is 21−( 2 ) . There are m


n
ways to choose n vertices, so by the union bound the
chance of at least one of those choices being monochromatic is that product.
n
b. (5 points) Show that if m < 2( 2 )−1 , that probability is less than 1.


n
n
Solution: Multiply both sides by 2( 2 )−1 .
n 1
− 12 1−(n
2),
c. (10 points) Using the fact that m −n m

n


n
< m , show that if m = 2 2 then n
< 2
and conclude that n 1 1
R(n, n) > 2 2 − n − 2 .
3
 n2 n
Solution: mn = (2 2 − n − 2 )n = 2 2 − 2 −1 < 2( 2 )−1 . Now, from the previous parts we
n 1 1 n

have seen that this implies that a randomly colored Km will have probability less than
1 of containing a monochromatic Kn . But this means that there is some coloring which
does not contain a monochromatic Kn , and so R(n, n) > m.

7. (20 points) Prove a complementary upper bound: R(n, n) ≤ 4n .

Solution: Take any coloring of K4n . Pick a vertex v, and consider all 22n −1 edges coming
from it. Then there are at least 22n−1 edges of one color, so we give v a flag of that color
and restrict our attention to the smaller graph, of those vertices on the other ends of the
majority-color edges. In this graph we repeat this argument, and we continue doing this
until we get to a single point. At the end of this process we have a set of 2n vertices,
all flagged with different colors, so some color must have at least n flags, by pigeonhole,
and we can take the Kn given by the vertices flagged with this color. Now, if vi and vj
are in this subgraph, with i < j, then the edge from vi to vj is colored the same color
as the flag on vi . But all of the vertices in this subgraph have the same color flags, so
it is monochromatic. Then, since we have shown that every coloring of K4n must have a
monochromatic Kn , R(n, n) ≤ 4n .

k-color Ramsey Numbers

Similar to our definition R(n, m), we can define R(n1 , n2 , n3 , . . . , nk ) to be the least m such
that if Km is colored with k colors, there is some monochromatic Kni of color ci .
8. (15 points) Prove that R(3, 3, 3) ≤ 17. (In fact, R(3, 3, 3) = 17, but this is difficult to
show.)
Solution: Take a 3-coloring of K17 , and pick any vertex v. Consider the edges coming
from v. There are sixteen of them, and 16/3 > 5, so there must be some color with 6
edges of that color coming from v. Call this color green, the others red and blue. Consider
the K6 on the other end. Since all the edges going from v to it are green, if it contains
a single green edge, we have a green triangle. But if not, that is a 2 colored K6 , so since
R(3, 3) = 6, we must have a red or blue triangle instead.

9. (20 points) Show that

R(n1 , . . . , nk ) ≤ R(n1 , n2 , . . . nk−2 , R(nk , nk−1 ))

Solution: Take a k-coloring of Km where

m = R(n1 , . . . , nk ) ≤ R(n1 , n2 , . . . nk−2 , R(nk , nk−1 ))
and repaint it so that colors m and m − 1 look the same. Then by the definition of
the Ramsey number, we must either have a Kni in color i < k − 1, or we must have a
KR(nk−1 ,nk ) in the last two colors. But even in the last case, by the definition of R(n, m),
we must have a Knk−1 in color k − 1 or a Knk in color k, which completes the proof.
This gives us the existence of R(n1 , . . . , nk ) for all {n1 , . . . , nk }.
10. Prove that
4
 a. (10 points)
R(3, . . . , 3) ≤ 3r!
| {z }
r 3’s

Solution: Induction on r. The base case is done in problems 2 and 8. To induct, let
us suppose that
R(3, . . . , 3) ≤ 3(r − 1)!
| {z }
r − 1 3’s
and that we have a coloring of K3(r−1)! . Then we can choose a vertex v, and it has
many edges of up to r colors coming from it, but in particular we can choose some
color c so that there are at least 3(r − 1)! edges of color c coming from v. Now, if the
K3(r−1)! on the other end has any edges colored c, then we have a triangle of color c.
But if not, then that graph is colored with only r − 1 colors, and so by the inductive
hypothesis it contains a monochromatic triangle.
b. (15 points)
R(3, . . . , 3) > 2r
| {z }
r 3’s

Solution: As before, we induct, and the base case is problem 1a. Now, take a r − 1-
coloring of K2r−1 , which does not have any monochromatic triangles, put two copies
side by side, and connect all of their vertices by edges of color r. Then there are no
triangles of color r, because two points of the triangle would have to lie in the same
copy of the K2r−1 , and there are no monochromatic triangles of any other color by the
induction hypothesis.

Infinite Ramsey Numbers

11. (30 points) Define KN = (V, E), where V = {1, 2, 3, . . .}, and E = {(i, j) : i, j ∈ V, i <
j}. This is in some sense an infinite complete graph. Show that if every edge is colored
red or blue, there is some infinite subset V 0 of V such that all of the edges between points
of V 0 are the same color.
Solution: The proof is similar to that of problem 7. Take any vertex v1 and consider
the edges coming from it. There must be an infinite number of either red or blue edges,
so we put a flag of that color on v1 , and then we restrict our attention to the smaller (but
still infinite!) subgraph on the other ends of these edges. Repeat this process (infinitely).
When we finish, we have a sequence v1 , v2 , v3 , . . . of vertices of the graph, each with a
red or blue flag, such that if i < j, then vi and vj are connected by an edge of the same
color as the flag on vi . Well, this sequence must contain an infinite number of vertices
all flagged the same color, so take this subset, and the edges between them are all the
same color. This is an infinite monochromatic complete subgraph, as desired.

5
SMT 2010 Team Test February 28, 2010

x3
tan x − x − 3
1. Compute lim .
x→0 x
n
2. For how many integers n is 20−n equal to the square of a positive integer.
3. Find all possible solutions (x1 , x2 , . . . , xn ) to the following equations, where

x2
 
1
x1 = xn + n−1
2 xn
2
 
1 x
x2 = x1 + n
2 x1
x2
 
1
x3 = x2 + 1
2 x2
x2
 
1
x4 = x3 + 2
2 x3
x2
 
1
x5 = x4 + 3
2 x4
..
.
x2
 
1
xn = xn−1 + n−2 = 2010.
2 xn−1

4. Let ABCDEF be a convex hexagon, whose opposite angles are parallel, satisfying AF = 3, BC = 4,
and DE = 5. Suppose that AD, BE, CF intersect in a point. Find CD.
5. Rank the following in decreasing order:
√ √ √ √ √ √
1 1+2 2+3 3 12 + 2 2 + 3 2 1+2+3 1+ 2+ 3
A= √ √ √ ,B = ,C = ,D = 1 .
1+ 2+ 3 1+2+3 3 √ + √1 + √1
1 2 3

6. What is the least m such that for any m integers we can choose 6 integers such that their sum is
divisible by 6?
7. Find all positive integers n such that φ(n) = 16, where φ(n) is defined to be the number of positive
integers less than or equal to n that are relatively prime to n.
8. Suppose that for an infinitely differentiable function f ,
f (4x) + af (3x) + bf (2x) + cf (x) + df (0)
lim
x→0 x4
exists. Find 1000a + 100b + 10c + d.
9. Minimize x3 + 4y 2 + 9z under the constraints that xyz = 1, x, y, z ≥ 0.
10. Positive real numbers x, y, and z satisfy the equations

x2 + y 2 = 9
√
y2 + 2yz + z 2 = 16
√
z2 + 2zx + x2 = 25.
√
Compute 2xy + yz + zx.
11. Find the volume of the region given by the inequality

|x + y + z| + |x + y − z| + |x − y + z| + | − x + y + z| ≤ 4.
SMT 2010 Team Test February 28, 2010

12. Suppose we have a polyhedron consisting of triangles and quadrilaterals, and each vertex is shared by
exactly 4 triangles and one quadrilateral. How many vertices are there?
13. Rank the following in increasing order:
√ √ √ √
2011 + 2009 2011 2011 − 2009 2009
A= , B= ,
√ 2 √ √ 3
2011 + 2 2010 + 2009 √ √ 1
C= , D = 2010, E = 2011 − √
4 2 2011

14. Suppose f and g are continuously differentiable functions satisfying

f (x + y) = f (x)f (y) − g(x)g(y)

g(x + y) = f (x)g(y) + g(x)f (y)

Also suppose that f 0 (0) = 1 and g 0 (0) = 2. Find f (2010)2 + g(2010)2 .

15. Find the number of n-tuples (a1 , . . . an ) that maximize

a1 a2 a3 + a2 a3 a4 + · · · + an−2 an−1 an

under the constraints that n ≥ 3 and

a1 + a2 + · · · + an = 3m

for a fixed integer m, where ai are positive integers.

SMT 2010 Team Solutions February 28, 2010

1. Answer: 0
By L’Hopital’s Rule,
x3
tan x − x − sec2 x − 1 − x2
lim 3
= lim = 12 − 1 − 0 = 0
x→0 x x→0 1

2. Answer: 3 values
2
n 20k
If 20−n = k 2 then n = k 2 (20 − n) so n = 1+k 2 . However, since 1 + k
2
is always coprime to k 2 , 1 + k 2
must divide 20. Therefore, the only possible values of k are 1, 2, and 3, and the only possible values
of n are the corresponding values 10, 16, and 18.
3. Answer: x1 = x2 = x3 = . . . = xn = 2010
Notice that
x2n−1
 
1
x1 = xn + ⇒ 2x1 xn = x2n + x2n−1
2 xn
x2n
 
1
x2 = x1 + ⇒ 2x1 x2 = x21 + x2n
2 x1
x21
 
1
x3 = x2 + ⇒ 2x2 x3 = x22 + x21
2 x2
x22
 
1
x4 = x3 + ⇒ 2x3 x4 = x23 + x22
2 x3
..
.
x2
 
1
xn = xn−1 + n−2 ⇒ 2xn−1 xn = x2n−1 + x2n−2 .
2 xn−1

Adding all these equations up, subtracting the lefthand side to the righthand side of the equation, and
factoring, we get

(x1 − xn )2 + (x1 − x2 )2 + (x2 − x3 )2 + . . . (xn−1 − xn )2 = 0.

Since the sum of the squares is zero, each of the terms being added must be zero. Then it follows that
x1 = x2 = x3 = . . . = xn = 2010.
4. Answer: 3
Let O be the intersection of three diagonals. Since AF and CD are parallel, we have 4AOF and
4DOC similar, so AO/DO = F O/CO. Applying this method to other pairs of parallel sides, we have
AO/DO = BO/EO and BO/EO = CO/F O, so CO/F O = F O/CO. Thus CO = F O, 4AOF and
4DOC are congruent. Therefore CD = AF = 3.
5. Answer: B > A > C > D
It is easy to verify that
1n + 2 n + 3 n
f (n) =
1n−1 + 2n−1 + 3n−1
is a monotonically increasing function.
6. Answer: 11
We know that m must be at least 11 because we have a counterexample: the list of 10 integers
0, 0, 0, 0, 0, 1, 1, 1, 1, 1 does not satisfy the given property. Now, to see that 11 is sufficient, observe that
among any three integers we can find two whose sum is even. So, take three integers from a set of 11,
find the two whose sum is even, and replace the other. Now, we have two integers with even sum, and
nine others. We can repeat this process to get 5 pairs of integers, each with even sum, and the eleventh
SMT 2010 Team Solutions February 28, 2010

element is ignored. Among the pair-sums, we either have 3 numbers that all have the same remainder
when divided by 3, or 3 that all have different remainders. Either way, the sum of these is a multiple
of three, and then also of six, since we said that the pair-sums are all even.
7. Answer: 60, 48, 40, 32, 17
Given a number n = pa1 1 · pa2 2 . . . pakk , it is clear that a number is relatively prime to n if and only if
it is not divisible by any of the pi . A little experimentation with this idea, or previous knowledge of
number theory, gives the well-known formula:

φ(n) = pa1 1 −1 · (p1 − 1) · . . . · pakk −1 · (pk − 1)

With this, it is clear that any n satisfying φ(n) = 16 = 24 must be a power of two, times at most one
3 and at most one 5, or 17 by itself. Then, we have as possible solutions 60, 48, 40, 32, and 17, for five
solutions in total.
8. Answer: −3439
The fact that the limit exists implies that

lim (f (4x) + af (3x) + bf (2x) + cf (x) + df (0)) = (1 + a + b + c + d)f (0) = 0

x→0

therefore
a + b + c + d = −1.
Apply L’Hospital’s Rule once, then we have

f (4x) + af (3x) + bf (2x) + cf (x) + df (0) 4f 0 (4x) + 3af 0 (3x) + 2bf 0 (2x) + cf 0 (x)
lim = lim
x→0 x4 x→0 4x3
and for the following limit to exist we also need

lim (4f 0 (4x) + 3af 0 (3x) + 2bf 0 (2x) + cf 0 (x)) = (4 + 3a + 2b + c)f 0 (0) = 0,
x→0

therefore
3a + 2b + c = −4.
Repeat this process twice and get another two equations:

9a + 4b + c = −16

27a + 8b + c = −64
Solving these four equations one can get (a, b, c, d) = (−4, 6, −4, 1), giving the answer 1000a + 100b +
10c + d = −3439.
q
9. Answer: 11 11 27 4

This is an application of the arithmetic-geometric mean inequality:

3 2
s  
2 3  6
2 · x2 + 3 · 4y3 + 6 · 9z
6 11 x3 4y 2 9z
≥ · ·
11 2 3 6

When we use the fact that x6 y 6 z 6 = 1 and multiply both sides by eleven, we get the answer. (Recall
that the bound given by the AM-GM inequality is always achievable).
√
10. Answer: 12 2
Note that these equations√ come from
√ applying the Law of Cosines to the triangle in the figure. The
desired value is simply 2 2A = 12 2, where A is the area of the large triangle.
SMT 2010 Team Solutions February 28, 2010

x 5
3

135°

y 135°
z

20
11. Answer: 3
Let
F (x, y, z) = |x + y + z| + |x + y − z| + |x − y + z| + | − x + y + z|.
Note that
F (x, y, z) = F (−x, y, z) = F (x, −y, z) = F (x, y, −z)
so the region is symmetric with respect to all xy, yz, zx-planes. Thus one can only consider the first
octant part (where x, y, z ≥ 0) and reflect it to get a full figure.
Assume that x is the largest. Then x − y + z, x + y − z ≥ 0. Since the equation of the form A + |B| ≤ C
implies both A + B ≤ C and A − B ≤ C, we have

(x + y + z) + (x + y − z) + (x − y + z) + (−x + y + z) = 2(x + y + z) ≤ 4

and
(x + y + z) + (x + y − z) + (x − y + z) − (−x + y + z) = 4x ≤ 4.
Thus x + y + z ≤ 2, and x, y, z ≤ 1. We can now see that the region is cube 0 ≤ x, y, z ≤ 1 cut by the
plane x + y + z ≤ 2. Since the plane goes through three vertices (1, 1, 0), (1, 0, 1), (0, 1, 1) of the cube,
it cuts out a prism of volume 61 out of the cube. So the volume is 1 − 61 = 56 . We multiply 8 to get the
volume of whole region, so the answer is 65 · 8 = 20
3 .

12. Answer: 24
Let v, e, t, q be the number of vertices, edges, triangular faces, and quadrilateral faces respectively.
Note that each vertex is shared by exactly one quadrilateral, and a quadrilateral provides four vertices.
By simple counting we get v = 4q. Apply the same thing to triangular face, then we have 4v = 3t.
Meanwhile from each vertex we have 5 edges coming out, so 5v = 2e. Thus we have

q = 1/4v, t = 4/3v, e = 5/2v.

And from the Euler’s formula v − e + (t + q) = 2, we have (1 − 5/2 + 1/4 + 4/3)v = 1/12v = 2, v = 24.
13. Answer: A < C < B < D < E
These are approximations to the integral
2011
1√
Z
x dx
2009 2

A is a trapezoid approximation, B is the integral itself, C is a trapezoid approximation with a finer

partition (including the middle of the segment), D approximates the function by its value at 2010,
and then integrates, and E is an integral of the function approximated by its tangent line at 2011.
Simply
√ drawing the areas represented by these approximations, and keeping in mind the concavity
of x, makes it clear how these compare. Note that one should not try to simply write out decimal
approximations to these values, as they first differ in the 6th place after the decimal point.
SMT 2010 Team Solutions February 28, 2010

14. Answer: e4020

Let H(x) = f (x)2 + g(x)2 , then H(x + y) = H(x)H(y). We can prove that H(x) = eax for some a,
since H is continuously differentiable. To find a, consider H 0 (0). This is 2f 0 (0)f (0) + 2g 0 (0)g(0).
We claim f (0) = 1, g(0) = 0. From H(0) = 1 we have f (0)2 + g(0)2 = 1, so we can let f (0) = cos θ,
g(0) = sin θ. Substituting x = y = 0 in both equations, it is easy to see that we get cos θ = cos(2θ),
sin θ = sin(2θ). So θ = 2nπ for some n, f (0) = 1, g(0) = 0. Then 2f 0 (0)f (0) + 2g 0 (0)g(0) =
2 · 1 · 1 + 2 · 2 · 0 = 2, so a = 2, f (2010)2 + g(2010)2 = e2010a = e4020 .
15. Answer: m
Notice that if n ≥ 5, we can strictly increase that sum by replacing an−3 with an−3 + an , and omitting
an . Hence, we have either n = 3, and all three ai are equal to m, or we have n = 4, and then
a1 a2 a3 + a2 a3 a4 = (a1 + a4 )a2 a3 , so a2 = a3 = a1 + a4 = m, and there are m − 1 ways to make this
happen, for a total of m possibilities.
SMT 2011 Algebra Test and Solutions February 19, 2011

√
2 3
1. Let a, b ∈ C such that a + b = a2 + b2 = 3 i. Compute | Re(a)|.
1
Answer: √
2
√ √ √ √
From a + b = 2 33 i we can let a = 33 i + x and b = 3
3
i − x. Then a2 + b2 = 2(( 33 i)2 + x2 ) =
√ √ √ √
2(x2 − 31 ) = 2 3 3 i. So x2 = 1+3 3i = 23 eiπ/3 , x = ± √23 · 3+i
2 . Since | Re(a)| = | Re(x)|, the answer is
√ √
√2 · 3
= √1 .
3 2 2

2. Consider the curves x2 + y 2 = 1 and 2x2 + 2xy + y 2 − 2x − 2y = 0. These curves intersect at two
points, one of which is (1, 0). Find the other one.
Answer: − 35 , 45

From the first equation, we get that y 2 = 1 − x2 . Plugging this into the second one, we are left with
p p p
2x2 ± 2x 1 − x2 + 1 − x2 − 2x ∓ 2 1 − x2 = 0 ⇒ (x − 1)2 = ∓2 1 − x2 (x − 1)
p
⇒ x − 1 = ∓2 1 − x2 assuming x 6= 1
⇒ x2 − 2x + 1 = 4 − 4x2 ⇒ 5x2 − 2x − 3 = 0.

The quadratic formula yields that x = 2±8 3

10 = 1, − 5 (we said that x 6= 1 above but we see that it is still
valid). If x = 1, the first equation forces y = 0 and we easily see that this solves the second equation.
If x = − 35 , then clearly y must be positive or else the second equation will sum five positive terms.
q q
9
= 16 4 3 4


Therefore y = 1 − 25 25 = 5 . Hence the other point is − 5 , 5 .

3. If r, s, t, and u denote the roots of the polynomial f (x) = x4 + 3x3 + 3x + 2, find

1 1 1 1
2
+ 2 + 2 + 2.
r s t u
9
Answer: 4
First notice that the polynomial
 
4 1 3 3
g(x) = x + 3 + +2 = 2x4 + 3x3 + 3x + 1
x4 x x
is a polynomial with roots 1r , 1s , 1t , u1 . Therefore, it is sufficient to find the sum of the squares of the
roots of g(x), which we will denote as r1 through r4 . Now, note that
a3 2 a2
r12 + r22 + r32 + r42 = (r1 + r2 + r3 + r4 )2 − (r1 r2 + r1 r3 + r1 r4 + r2 r3 + r2 r4 + r3 r4 ) = (− ) −
a4 a4
by Vieta’s Theorem, where an denotes the coefficient of xn in g(x). Plugging in values, we get that
our answer is (− 32 )2 − 0 = 94 .
4. Find the 2011th-smallest x, with x > 1, that satisfies the following relation:
sin(ln x) + 2 cos(3 ln x) sin(2 ln x) = 0.

Answer: x = e2011π/5
Set y = ln x, and observe that
2 cos(3y) sin(2y) = sin(3y + 2y) − sin(3y − 2y) = sin(5y) − sin(y),
so that the equation in question is simply
sin(5y) = 0.
The solutions are therefore
nπ
ln x = y = =⇒ x = enπ/5 for all n ∈ N.
5
SMT 2011 Algebra Test and Solutions February 19, 2011

5. Find the remainder when (x + 2)2011 − (x + 1)2011 is divided by x2 + x + 1.

Answer: (−31005 − 1)x + (−2 · 31005 − 1)
The standard method is to use the third root of unity ω, ω 2 + ω + 1 = 0. Let (x + 2)2011 − (x + 1)2011 =
(x2 + x +√
1)Q(x) + ax + b and substitute x = ω. Then aω + b = (ω + 2)2011 − (ω + 1)2011 . Note that ω + 2
has size 3 and argument π/6, so (ω + 2)6 = −33 . Also ω + 1 has magnitude 1 and argument π/3, so
(ω + 1)6 = 1. Using this and 2011 = 6 · 335 + 1, we get that aω + b = (−31005 − 1)ω + (−2 · 31005 − 1).
Another solution is to note that (x + 2)2 ≡ x2 + 4x + 4 ≡ −3x2 (mod x2 + x + 1) and (x + 1)2 ≡
x2 + 2x + 1 ≡ x (mod x2 + x + 1). Then we have x3 ≡ 1 (mod x2 + x + 1) and we can proceed by
using periodicity.
6. There are 2011 positive numbers with both their sum and the sum of their reciprocals equal to 2012.
Let x be one of these numbers. Find the maximum of x + x−1 .
8045
Answer: 2012
Let y1 , y2 , · · · , y2010 be the 2010 numbers distinct from x. Then y1 + y2 + · · · + y2010 = 2012 − x and
1 1 1 1
y1 + y2 + · · · + y2010 = 2012 − x . Applying the Cauchy-Schwarz inequality gives

2010
! 2010
!
X X 1 1
yi = (2012 − x)(2012 − ) ≥ 20102
i=1 i=1
yi x

−1
2
so 2012 − 2012(x + x ) + 1 − 20102 ≥ 0, x + x−1 ≤ 8045/2012.
7. Let P (x) be a polynomial of degree 2011 such that P (1) = 0, P (2) = 1, P (4) = 2, ... , and P (22011 ) =
2011. Compute the coefficient of the x1 term in P (x).
1
Answer: 2 − 22010
We analyze Q(x) = P (2x) − P (x). One can observe that Q(x) − 1 has the powers of 2 starting from
1, 2, 4, · · · , up to 22010 as roots. Since Q has degree 2011, Q(x) − 1 = A(x − 1)(x − 2) · · · (x − 22010 )
for some A. Meanwhile Q(0) = P (0) − P (0) = 0, so

Q(0) − 1 = −1 = A(−1)(−2) · · · (−22010 ) = −2(2010·2011)/2 A.

Therefore A = 2−(1005·2011) . Finally, note that the coefficient of x is same for P and Q − 1, so it equals
1005·2011
(22011 −1) 1
A(−20 )(−21 ) · · · (−22010 )((−20 ) + (−2−1 ) + · · · + (−2−2010 )) = A·2 22010 = 2 − 2010 .
2
8. Find the maximum of
ab + bc + cd
a2 + b2 + c2 + d2
for reals a, b, c, and d not all zero.
√
5+1
Answer: 4
One has ab ≤ 2t a2 + 1 2
2t b , bc ≤ 12 b2 + 12 c2 , and cd ≤ 1 2
2t c + 2t d2 by AM-GM. If we can set t such that
t 2 2 2 2
t 1 2 (a +b +c +d )
2 = 2t + 12 , it can be proved that a2ab+bc+cd
+b2 +c2 +d2 ≤ = 2t , and this is maximal because we
a2 +b2 +c2 +d2
can set√a, b, c, d so that the equality√ holds in every inequality we used. Solving this equation, we get
t = 1+2 5 , so the maximum is 2t = 5+1 4 .

9. It is a well-known fact that the sum of the first n k-th powers can be represented as a polynomial in
n. Let Pk (n) be such a polynomial for integers k and n. For example,
n
X n(n + 1)(2n + 1)
i2 = ,
i=1
6

so one has
x(x + 1)(2x + 1) 1 1 1
P2 (x) = = x3 + x2 + x.
6 3 2 6
SMT 2011 Algebra Test and Solutions February 19, 2011

Evaluate P7 (−3) + P6 (−4).

Answer: −665
Since the equation
Pk (x) = Pk (x − 1) + xk
has all integers ≥ 2 as roots, the equation is an identity, so it holds for all x. Now we can substitute
x = −1, −2, −3, −4, · · · to prove
n−1
X
Pk (−n) = − (−i)k
i=1

so P7 (−3) + P6 (−4) = −(−1) − (−2) − (−3) − (−1)7 − (−2)7 = −665.

6 6 6

10. How many polynomials P of degree 4 satisfy P (x2 ) = P (x)P (−x)?

Answer: 10
2 3
Note that if r is a root of P then r2 is also a root. Therefore r, r2 , r2 , r2 , · · · , are all roots of P . Since
P has a finite number of roots, two of these roots should be equal. Therefore, either r = 0 or rN = 1
for some N > 0.
If all roots are equal to 0 or 1, then P is of the form axb (x − 1)(4−b) for b = 0, ..., 4.
Now suppose this is not the case. For such a polynomial, let q denote the largest integer such that
r = e2πi·p/q is a root for some integer p coprime to q. We claim that the only suitable q > 1 are q = 3
and q = 5.
√ √
First note that if r is a root then one of r or − r is also a root. So if q is even, then one of
e2πi·p/2q or e2πi·p+q/2q should also be root of p, and both p/q and (p + q)/2q are irreducible fractions.
This contradicts the assumption that q is maximal. Therefore q must be odd. Now, if q > 6, then
r−2 , r−1 , r, r2 , r4 should be all distinct, so q ≤ 6. Therefore q = 5 or 3.
If q = 5, then the value of p is not important as P has the complex fifth roots of unity as its roots,
so P = a(x4 + x3 + x2 + x + 1). If q = 3, then P is divisible by x2 + x + 1. In this case we let
P (x) = a(x2 + x + 1)Q(x) and repeating the same reasoning we can show that Q(x) = x2 + x + 1 or
Q(x) is of form xb (x − 1)2−b .
Finally, we can show that exactly one member of all 10 resulting families of polynomials fits the desired
criteria. Let P (x) = a(x − r)(x − s)(x − t)(x − u). Then, P (x)P (−x) = a2 (x2 − r2 )(x2 − s2 )(x2 −
t2 )(x2 − u2 ). We now claim that r2 , s2 , t2 , and u2 equal r, s, t, and u in some order. We can prove
this noting that the mapping f (x) = x2 maps 0 and 1 to themselves and maps the third and fifth roots
of unity to another distinct third or fifth root of unity, respectively. Hence, for these polynomials,
P (x)P (−x) = a2 (x2 − r)(x2 − s)(x2 − t)(x2 − u) = aP (x2 ), so there exist exactly 10 polynomials that
fit the desired criteria, namely the ones from the above 10 families with a = 1.
SMT 2011 Geometry Test and Solutions February 19, 2011

1. Triangle ABC has side lengths BC = 3, AC = 4, AB = 5. Let P be a point inside or on triangle ABC
and let the lengths of the perpendiculars from P to BC, AC, AB be Da , Db , Dc respectively. Compute
the minimum of Da + Db + Dc .
12
Answer: 5
Let M = Da + Db + Dc . Notice that the total area of ABC is (1/2)(3Da + 4Db + 5Dc ) = (1/2)[5(Da +
Db + Dc ) − 2Da − Db ] = (1/2)(5M − 2Da − Db ). Of course, since ABC is a right triangle, we know
its area is 3 · 4/2 = 6. We thus have M = (1/5)(2 · 6 + 2Da + Db ). This is clearly minimized when
Da = Db = 0, that is, when we set P to coincide with C. We then have M = 12/5.
Note: it is also possible to conclude this from a few applications of the triangle inequality.
2. Pentagon ABCDE is inscribed in a circle of radius 1. If ∠DEA ∼
= ∠EAB ∼
= ∠ABC, m∠CAD = 60◦ ,
and BC = 2DE, compute the area of ABCDE.
√
33 3
Answer: 28
Looking at cyclic quadrilaterals ABCD and ACDF tells us that m∠ACD = m∠ADC, so 4ACD is
equilateral and m∠DEA = 120◦ . Now, if we let m∠EAD = θ, we see that m∠CAB = 60◦ − θ =⇒
m∠ACB = θ =⇒ 4AED ∼ = 4CBA. Now all we have to do is √
calculate side lengths. After creating
some 30◦ − 60◦ − 90◦ triangles, it becomes evident that AC = 3. Now let AB = x, so BC = 2x.
By applying the Law of Cosines to triangle ABC, we find that x2 = 73 . Hence, the desired area
√ √ √
( 3)2 3
(ABCDE) = (ACD) + 2(ABC) = 4 + 2 · 21 (x)(2x)(sin 120◦ ) = 33 3
28 .

3. Let circle O have radius 5 with diameter AE. Point F is outside circle O such that lines F A and F E
◦
intersect circle O at points B and D, respectively. If
√ F A =√10 and
√m∠F AE = 30 , then the perimeter
of quadrilateral ABDE can be expressed as a + b 2 + c 3 + d 6, where a, b, c, and d are rational.
Find a + b + c + d.
Answer: 15
After some angle chasing, we find that m∠DBF = m∠DF B = 75◦ , which implies that DF = DB.
Hence the desired perimeter is equal to AF − BF + AE + F E = 20 − BF + F E.
√ √
By the law of sines, sinF 30
E 10
◦ = sin 75◦ =⇒ F E =
√
5√
6+ 2
= 5 6 − 5 2.
4

Now, to find BF , draw the altitude√from O to AB intersecting AB at√P . This forms a 30◦ − 60◦ − 90◦
triangle, so we can see that AP = 5 3/2 = 10−BF =⇒ BF = 10 − 5 3. Hence, the desired perimeter
√ √ √ √2 √ √
is 20 + (5 6 − 5 2) − (10 − 5 3) = 10 − 5 2 + 5 3 + 5 6, so the answer is 10 − 5 + 5 + 5 = 15.
4. Let ABC be any triangle, and D, E, F be points on BC, CA, AB such that CD = 2BD, AE = 2CE
and BF = 2AF . AD and BE intersect at X, BE and CF intersect at Y , and CF and AD intersect
Area(4ABC)
at Z. Find Area(4XY Z) .
Answer: 7
Using Menelaus’s Theorem on 4ABD with collinear points F, X, C and the provided ratios gives
DX/XA = 4/3. Using Menelaus’s Theorem on 4ADC with collinear points B, Y, E gives AY /Y D = 6.
We conclude that AX, XY, Y D are in length ratio 3 : 3 : 1. By symmetry, this also applies to the
segments CZ, ZX, XF and BY, Y Z, ZE. Repeatedly using the fact that the area ratio of two triangles
of equal height is the ratio of their bases, we find [ABC] = (3/2)[ADC] = (3/2)(7/3)[XY C] =
(3/2)(7/3)(2)[XY Z] = 7[XY Z], or [ABC]/[XY Z] = 7.
Alternate Solution
Stretching the triangle will preserve ratios between lengths and ratios between areas, so we may assume
that 4ABC is equilateral with side length 3. We now use mass points to find the length of XY . Assign
a mass of 1 to A. In order to have X be the fulcrum of 4ABC, C have mass 2 and B must have
mass 4. Hence, BX : XE = 4 : 3 and AX : XD = 6 : 1, the latter of which also equals BY : Y E
by symmetry. Hence, XY = 73 BE. To find BE, we apply the Law of Cosines to 4CBE to get that
√
BE 2 = 12 + 32 − 2 · 1 · 3 · cos 60◦ = 7 =⇒ XY = 3 7 7 . Since 4XY Z must be equilateral by symmetry,
AB 2
the desired ratio equals ( XY ) = 7.
SMT 2011 Geometry Test and Solutions February 19, 2011

5. Let ABCD be a cyclic quadrilateral with AB = 6, BC = 12, CD = 3, and DA = 6. Let E, F be the

intersection of lines AB and CD, lines AD and BC respectively. Find EF .
√
Answer: 10 2

We have 4ADE ∼ 4CBE, and their length ratio is AD : CB = 1 : 2. Let AE = p and DE = q.

Then we have AB = BE − AE = 2DE − AE = 2q − p and CD = 2p − q. Solving for p and q, we have
p = 4 and q = 5. Similarly we have F C = 8 and F D = 10. Let ∠B = θ. Then ∠F DE = π − θ. Apply
the Law of Cosines to 4EBF to get

EF 2 = BE 2 + BF 2 − 2BE · BF · cos θ = 102 + 202 − 2 · 10 · 20 cos θ = 500 − 400 cos θ

and to 4EDF to get

EF 2 = DE 2 + DF 2 + 2 · DE · DF cos θ = 52 + 102 − 2 · 5 · 10 cos θ = 125 + 100 cos θ.

Solving for EF 2 , we get EF 2 = 200.

6. Two parallel lines l1 and l2 lie on a plane, distance d apart. On l1 there are an infinite number of
points A1 , A2 , A3 , · · · , in that order, with An An+1 = 2 for all n. On l2 there are an infinite number of
points B1 , B2 , B3 , · · · , in that order and in the same direction, satisfying
P∞ Bn Bn+1 = 1 for all n. Given
that A1 B1 is perpendicular to both l1 and l2 , express the sum i=1 ∠Ai Bi Ai+1 in terms of d.
Answer: π − tan−1 ( d1 ) (or π/2 + tan−1 d or other equivalent form)

Construct points C1 , C2 , C3 , · · · on l1 progressing in the same direction as the Ai such that C1 =

A1 and Cn Cn+1 = 1. Thus we have C1 = A1 , C3 = A2 , C5 = A3 , etc., with C2n−1 = An in
SMT 2011 Geometry Test and Solutions February 19, 2011

general. We can write ∠Ai Bi Ai+1 = ∠C2i−1 Bi C2i+1 = ∠Ci Bi C2i+1 − ∠Ci Bi C2i−1 . Observe that
4Ci Bi Ck (for any k) is a right triangle with legs of length d and k − i, and ∠Ci Bi Ck = tan−1 k−i
d . So
∠Ci Bi C2i+1 − ∠Ci Bi C2i−1 = tan−1 i+1d − tan−1 i−1
d . The whole sum is therefore
∞  
−1 i + 1 −1 i − 1
X
tan − tan
i=1
d d

which has nth partial sum

n+1 n 1
tan−1 + tan−1 − tan−1
d d d
1
so it converges to π − tan−1 .
d
√
7. In a unit square ABCD, find the minimum of 2AP + BP + CP where P is a point inside ABCD.
√
Answer: 5

0 0
Rotate triangle AP B around A by 90 degrees as in the given figure.
√ Let P and B be the rotated
0 0 0
images of P and B respectively. Then we have B P = BP , P P = 2AP so
√ √
2AP + BP + CP = CP = P P 0 + P 0 B 0 ≤ CB 0 = 5.

8. We have a unit cube ABCDEF GH where ABCD is the top side and EF GH is the bottom side with
E below A, F below B, and so on. Equilateral triangle BDG cuts out a circle from the cube’s inscribed
sphere. Find the area of the circle.
π
Answer: 6
Consider the cube to be of side length 2 and divide the answer by 4 later. Set the coordinates of the
vertices of the cube to be (±1, ±1, ±1). Then the plane going through an equilateral triangle can be
described by the equation x+y+z = 1. The distance to the plane from the origin is √13 , as (1/3, 1/3, 1/3)
q q
is the foot of the perpendicular from (0, 0, 0). Thus the radius of the circle is 1 − ( √13 )2 = 23 , so
the area is 23 π. In the case of the unit cube we should divide this by 4 to get the answer π
6.

9. We have a circle O with radius 10 and four smaller circles O1 , O2 , O3 , O4 of radius 1 which are internally
tangent to O, with their tangent points to O in counterclockwise order. The small circles do not intersect
each other. Among the two common external tangents of O1 and O2 , let l12 be the one which separates
O1 and O2 from the other two circles, and let the intersections of l12 and O be A1 and B2 , with A1
denoting the point closer to O1 . Define l23 , l34 , l41 and A2 , A3 , A4 , B3 , B4 , B1 similarly. Suppose that
the arcs A1 B1 , A2 B2 , and A3 B3 have length π, 3π/2, and 5π/2 respectively. Find the arc length of
A4 B 4 .
Answer: 2π
The key is to note that A1 B1 + A3 B3 = A2 B2 + A4 B4 (in terms of arc length). Proof: let Pi be
the point of tangency between O and Oi , for i = 1, 2, 3, 4. Observe that it is enough to show that
SMT 2011 Geometry Test and Solutions February 19, 2011

Ai Pi = Bj Pj for all i, j ∈ {1, 2, 3, 4} where j − i ≡ 1 (mod 4). But these two arcs are symmetric with
respect to the perpendicular bisector of lij , as all the small circles have the same radius, so our initial
claim is correct.
10. Given a triangle ABC with BC = 5, AC = 7, and AB = 8, find the side length of the largest equilateral
triangle P QR such that A, B, C lie on QR, RP, P Q respectively.
√
Answer: 2 43

q
2 2
√
Let a = BC, b = AC, c = AB. We claim that in general, the answer is 3 (a + b2 + c2 + 4 3S) ,
where S is the area of ABC.
Suppose that P QR is an equilateral triangle satisfying the conditions. Then ∠BP C = ∠CQA =
∠ARB = 60◦ . The locus of points satisfying ∠BXC = 60◦ is part of a circle Oa . Draw Ob and Oc
similarly. These three circles meet at a single point X inside the triangle, which is the unique point
satisfying ∠BXC = ∠CXA = ∠AXB = 120◦ . Then the choice of P on Oa determines Q and R:
those two points should also be on Ob and Oc respectively, and line segments P CQ and P BR should
form sides of the triangle. Now one should find the maximum of P Q under these conditions. Note that
∠BP X and ∠BRX do not depend on the choice of P , so triangle P XR has the same shape regardless
of our choice. In particular, the ratio of P X to P R is constant, so P R is maximized when P X is the
diameter of Oa . This requires P Q, QR, RP to be perpendicular to XC, XA, XB respectively.
From this point there may be several ways to calculate the answer. One way is to observe that
P Q = √23 (AX + BX + CX) by considering (P QR) = (P XQ) + (QXR) + (RXP ). AX + BX + CX
can be computed by the usual rotation trick for the Fermat point: rotate 4BXA 60◦ around B
to 4BX 0 A0 . Observe that 4BXX 0 is equilateral, and so A0 , X 0 , X, and C are collinear. Hence,
A0 C = AX + BX + CX, and we can apply the Law of Cosines to 4A0 BC to get√that A0 C 2 = c2 + a2 −
2ac cos (B + 60◦ ) = a2 + c2 + 2ac sin 60◦ sin B − 2ac cos 60◦ cos B = a2 + c2 + 2S 3 − 12 (a2 + c2 − b2 ) =
a2 +b2 +c2
√ q
2 2
√
+ 2S 3 =⇒ P Q = 2 2
2 3 (a + b + c + 4 3S) (where S is again the area of ABC). Plugging
√ √
in our values for √ a, b, and c, and using Heron’s formula to find S = 10 ∗ 5 ∗ 3 ∗ 2 = 10 3, we can
calculate P Q = 2 43.
SMT 2011 Calculus Test and Solutions February 19, 2011

1. If f (x) = (x − 1)4 (x − 2)3 (x − 3)2 , find f 000 (1) + f 00 (2) + f 0 (3).

Answer: 0
A polynomial p(x) has a multiple root at x = a if and only if x − a divides both p and p0 . Continuing
inductively, the nth derivative p(n) has a multiple root b if and only if x − b divides p(n) and p(n+1) .
Since f (x) has 1 as a root with multiplicity 4, x − 1 must divide each of f, f 0 , f 00 , f 000 . Hence f 000 (1) = 0.
Similarly, x − 2 divides each of f, f 0 , f 00 so f 00 (2) = 0 and x − 3 divides each of f, f 0 , meaning f 0 (3) = 0.
Hence the desired sum is 0.
2. A trapezoid is inscribed in a semicircle of radius 2 such that one base of the trapezoid lies along the
diameter of the semicircle. Find the largest possible area of the trapezoid.
√
Answer: 3 3
Clearly, a trapezoid with maximal area will have a base equal to the
√ diameter. If x is the height of the
trapezoid, then the area of a trapezoid is h(b12+b2 ) = A(x) = 2 + 4 − x2 · x so the maximum occurs

when √
p x2 2 4 − x2 + 4 − 2x2
0 = A0 (x) = 2 + 4 − x2 − √ = √ ,
4 − x2 4 − x2
which is equivalent to
4(4 − x2 ) = (2x2 − 4)2 = 4x4 − 16x2 + 16.
√
Collecting like terms gives 4x4 = 12x2 , and
√ since x 6= 0 (the degenerate case), we get that x = 3.
Thus the desired maximum occurs at x = 3 and so the maximum area is
√ √
√ √
A( 3) = 2 + 4 − 3 · 3 = 3 3.

3. A sector of a circle has angle θ. Find the value of θ, in radians, for which the ratio of the sector’s area
to the square of its perimeter (the arc along the circle and the two radial edges) is maximized. Express
your answer as a number between 0 and 2π.
Answer: 2
Suppose that the circle has radius r. Then the area of the circle is πr2 , so the area of the sector is
θ 2 1 2 θ
2π πr = 2 θr . The arc of the perimeter of the sector has length 2π 2πr = θr, and the two straight
edges of the sector each has length r, so the perimeter has length θr + 2r = (θ + 2)r, and hence the
square of the perimeter is (θ + 2)2 r2 . The ratio that we want to maximize is therefore
1 2
2 θr θ
= .
(θ + 2)2 r2 2(θ + 2)2

To do this, differentiate to find the critical points:

2(θ + 2)2 − 4θ(θ + 2)

 
d θ 2(θ + 2) − 4θ 2−θ
0= 2
= = = =⇒ θ = 2.
dθ 2(θ + 2) 4(θ + 2)4 4(θ + 2)3 2(θ + 2)3

Observe that the derivative is decreasing at θ = 2, which implies that this is a local maximum, as
desired.
Alternate Solution:
Equivalently, we can minimize the reciprocal:

d 2(θ + 2)2
 
d
4θ−1 + 4 + θ = 2 −4θ−2 + 1 =⇒ θ2 = 4 =⇒ θ = 2.



0= =2
dθ θ dθ

2
x 3 ex
4. Let f (x) = 1−x2 . Find f (7) (0), the 7th derivative of f evaluated at 0.
Answer: 12600
SMT 2011 Calculus Test and Solutions February 19, 2011

Since f (n) (0) = an n!, where an is the nth Taylor series coefficient, we just need to find the Taylor series
of f and read off the appropriate coefficient. The Taylor series is given by

x2 x4
 
3
1 + x2 + x4 + · · · .


f (x) = x 1 + + + ···
1! 2!
1 1
The coefficient of x7 is 2! + 1! + 1 = 52 , so f (7) (0) = 7! · 5
2 = 12600.
5. The real-valued infinitely differentiable function f (x) is such that f (0) = 1, f 0 (0) = 2, and f 00 (0) = 3.
Furthermore, f has the property that

f (n) (x) + f (n+1) (x) + f (n+2) (x) + f (n+3) (x) = 0

for all n ≥ 0, where f (n) (x) denotes the nth derivative of f . Find f (x).
Answer: 2e−x − cos x + 4 sin x
We solve the differential equation f + f 0 + f 00 + f 000 = 0. Let f + f 0 = g. Then we need to solve
g + g 00 = 0, which has solution g(x) = a cos x + b sin x. Then

ex (f + f 0 ) = (ex f )0 = aex cos x + bex sin x,

so that Z 
−x
f =e (ae cos x + be sin x) dx + c = ce−x + a0 cos x + b0 sin x.
x x

Finally, we find f (0) = c + a0 , f 0 (0) = −c + b0 , and f 00 (0) = c − a0 and solve for a0 , b0 , c.

Alternate Solution: Observe that since the given equation holds for all n, by moving the index
up one and then subtracting, we get f (n) (x) − f (n+4) (x) = 0, so that f (n) (x) = f (n+4) (x). That
is, any function that satisfies the given equation must also have the property that the derivatives
repeat in cycles of 4. However, as we will see, this is only a necessary property, not a sufficient one.
The characteristic equation of the given differential equation is λn+4 − λn = 0, or λn (λ4 − 1) = 0.
The roots of this equation are 0 and the fourth roots of unity, so a complete set of solutions is
given by f (x) = aex + be−x + ceix + de−ix (the terms eix and e−ix can be written in terms of sine
and cosine, as is boxed above). Note however, that aex does not satisfy the original differential
equation as all of its derivatives have the same sign. Relabelling the constants, the solution set is
f (x) = ae−x + beix + ce−ix = ae−x + b(cos x + i sin x) + c(cos x − i sin x).
Z π
x2
6. Compute p dx.
−π 1 + sin x + 1 + sin2 x
π3
Answer: 3
Use symmetry around the origin. Substitute x to −x, so the integral is now
Z π
x2 dx
p .
−π 1 − sin x + 1 + sin2 x
Add the two integrals, and note that
p
1 1 2 + 2 1 + sin2 x
p + p = p = 1,
1 + sin x + 1 + sin2 x 1 − sin x + 1 + sin2 x 2 + sin2 x + 2 1 + sin2 x − sin2 x
Rπ 3
so the integral is the same as 21 −π x2 dx = π3 .

7. For the curve sin(x) + sin(y) = 1 lying in the first quadrant, find the constant α such that

d2 y
lim xα
x→0 dx2
SMT 2011 Calculus Test and Solutions February 19, 2011

exists and is nonzero.

3
Answer: 2
Differentiate the equation to get
dy
cos(x) + cos(y) = 0
dx
and again to get
2
d2 y

dy
− sin(x) + 2 cos(y) − sin(y) = 0.
dx dx
By solving these we have
dy cos(x)
=−
dx cos(y)
and
d2 y sin(x) cos2 (y) + sin(y) cos2 (x)
2
= .
dx cos3 (y)
√ p p
Let sin(x) = t, then sin(y) = 1 − t. Also cos(x) = 1 − t2 and cos(y) = 1 − (1 − t)2 = t(2 − t).
Substituting gives
d2 y t2 (2 − t) + (1 − t)(1 − t2 ) −3/2 1 − t + t
2
= = t .
dx2 t3/2 (2 − t)3/2 (2 − t)3/2
2
t 3 d y 1
Since limx→0 x = 1, α = 2 should give the limit limx→0 xα dx 2 =
√
2 2
.
2
tan−1 x
Z
8. Compute dx.
1
2
x2 − x + 1
√
π2 3
Answer: 18
dx dy
Take y = 1/x, then x2 −x+1 = − y2 −y+1 . Note furthermore by the tangent addition formula that
−1 −1
tan (x) + tan (y) = π/2. The original integral is equal to the average of these two integrals:
!
Z 2 Z 2 π −1
tan−1 x 2 − tan y π 2
Z
1 dx
2
dx + 2
dy = 2
.
2 1 x − x + 1 1 y −y+1 4 1/2 x − x + 1
2 2

√
3
Substitute x = 2 θ + 1/2, then
2
√ Z √3 √
π2 3
Z
π dx π4 3 1
= dθ = .
4 1/2 x2 − x + 1 4 3 2 0 θ2 + 1 18

9. Solve the integral equation

Z x
f (x) = ex−y f 0 (y) dy − (x2 − x + 1)ex .
0

Answer: f (x) = (2x − 1)ex

Differentiate both sides to get
Z x
d x d 2
f 0 (x) = e e−y f 0 (y) dy − (x − x + 1)ex
dx 0 dx
Z x
f 0 (x) = f 0 (x) + ex−y f 0 (y) dy − (x2 + x)ex .
0
But Z x
ex−y f 0 (y) dy = f (x) + (x2 − x + 1)ex
0
SMT 2011 Calculus Test and Solutions February 19, 2011

so by substituting it we get
f (x) + (x2 − x + 1)ex − (x2 + x)ex = 0,
and f (x) = (2x − 1)ex .
10. Compute the integral Z π
ln(1 − 2a cos x + a2 ) dx
0
for a > 1.
Answer: 2π ln a
Solution 1:
This integral can be computed using a Riemann sum. Divide the interval of integration [0, π] into n
parts to get the Riemann sum
     
π 2 π 
2 2π 2 (n − 1)π
ln a − 2a cos + 1 + ln a − 2a cos + 1 + · · · + ln a − 2a cos +1 .
n n n n
Recall that
eiθ + e−iθ
cos θ = .
2
We can rewrite this sum of logs as a product and factor the inside to get
"n−1  # "n−1 #
π Y kπ π Y  
ln a2 − 2a cos +1 = ln a − ekπi/n a − e−kπi/n .
n n n
k=1 k=1

±kπi/n
The terms e are all of the 2n-th roots of unity except for ±1, so the inside product contains all
of the factors of a2n − 1 except for a − 1 and a + 1. The Riemann sum is therefore equal to
π a2n − 1
ln 2
n a −1
To compute the value of the desired integral, we compute the limit of the Riemann sum as n → ∞;
this is r
π a2n − 1 n a
2n − 1
lim ln 2 = lim π ln = lim π ln a2 = 2π ln a.
n→∞ n a −1 n→∞ a2 − 1 n→∞

(This is problem 471 of Răzvan Gelca and Titu Andreescu’s book Putnam and Beyond. The solution
is due to Siméon Poisson.)
Solution 2:
Let the desired integral be I(a), where we think of this integral as a function of the parameter a. In
this solution, we differentiate by a to convert the desired integral to an integral of a rational function
in cos x: Z π Z π
d d 2a − 2 cos x
I(a) = ln(1 − 2a cos x + a2 ) dx = dx.
da da 0 0 1 − 2a cos x + a2
All integrals of this form can be computed using the substitution t = tan x2 . Then x = 2 arctan t, so
2
dx = 1+t 2 dt and

1 − t2
 
1
cos x = cos(2 arctan t) = 2 cos(arctan t)2 − 1 = 2 2
−1= ,
1+t 1 + t2
so our integral becomes
2
2a − 2 1−t
Z ∞ Z ∞
d 1+t2 2 a(1 + t2 ) − (1 − t2 ) 1
I(a) = 1−t 2 2
dt = 4 2 2 ) + a2 (1 + t2 ) 1 + t2
dt
da 0 1 − 2a 1+t2 + a2 1+t
0 (1 + t ) − 2a(1 − t
Z ∞
(a + 1)t2 + (a − 1) 2 ∞ a2 − 1 2 ∞ 1
Z Z
=4 dt = dt + dt.
0 ((a + 1)2 t2 + (a − 1)2 )(1 + t2 ) a 0 (a + 1)2 t2 + (a − 1)2 a 0 1 + t2
SMT 2011 Calculus Test and Solutions February 19, 2011

a−1 a−1
In the first integral, we do the substitution t = a+1 u. Then dt = a+1 du and we have
Z ∞ Z ∞
2 1 2 1 2  π π  2π
= du + dt = + = .
a 0 1 + u2 a 0 1 + t2 a 2 2 a
Therefore, our desired integral is the integral of the previous quantity, or
Z π
I= ln(1 − 2a cos x + a2 ) dx = 2π ln a.
0

Solution 3:
We use Chebyshev polynomials1 . First, define the Chebyshev polynomial of the first kind to be
Tn (x) = cos(n arccos x). This is a polynomial in x, and note that Tn (cos x) = cos(nx). Note that

cos((n + 1)x) = cos nx cos x − sin nx sin x

cos((n − 1)x) = cos nx cos x + sin nx sin x,

so that cos((n + 1)x) = 2 cos nx cos x − cos((n − 1)x) and hence the Chebyshev polynomials satisfy the
recurrence Tn+1 (x) = 2xTn (x) − Tn−1 (x).
Therefore, the Chebyshev polynomials satisfy the generating function
∞
X 1 − tx
Tn (x)tn = .
n=0
1 − 2tx + t2

Now, substituting x 7→ cos x and t 7→ a−1 , we have

∞
X a − cos x
cos(nx)a−n = a .
n=0
a2 − 2a cos x + 1

So
∞
X 2a − 2 cos x
2 cos(nx)a−n−1 = .
n=0
1 − 2a cos x + a2
Then
Z π ∞
πX ∞  π 
2a − 2 cos x
Z X Z
−n−1
dx = 2 cos(nx)a dx = 2 a−n−1 cos(nx) dx = 2πa−1 .
0 1 − 2a cos x + a2 0 n=0 n=0 0

Now, since
2a − 2 cos x
Z
2
ln(1 − 2a cos x + a ) = da,
1 − 2a cos x + a2
we see that Z π Z
ln(1 − 2a cos x + a2 ) dx = 2πa−1 da = 2π ln a.
0

Solution 4:
We can also give a solution based on physics. By symmetry, we can evaluate the integral from 0 to 2π
and divide the answer by 2, so
Z π Z 2π p
2
ln(1 − 2a cos x + a ) dx = ln 1 − 2a cos x + a2 dx.
0 0

Now let’s calculate the 2D gravitational potential of a point mass falling along the x axis towards a
unit circle mass centered around the origin. We set the potential at infinity to 0. We also note that,
1 http://en.wikipedia.org/wiki/Chebyshev_polynomials
SMT 2011 Calculus Test and Solutions February 19, 2011

since the 2D gravitational force between two masses is proportional to 1r , the potential between two
masses is proportional to − ln r. So to calculate the gravitational potential, we integrate − ln r over
the unit circle. But if the point mass
√ is at (a, 0), then the distance between the point mass and the
section of the circle at angle x is 1 − 2a cos x + a2 . So we get the integral
Z 2π p
− ln 1 − 2a cos x + a2 dx
0

This is exactly the integral we want to calculate! We can also calculate this potential by concentrating
the mass of the circle at its center. The circle has mass 2π and its center is distance a from the point
mass. So the potential is simply −2π ln a. Thus, the final answer is 2π ln(a).
Solution 5:
This problem also has a solution which uses the Residue Theorem from complex analysis. It is easy to
show that Z π Z 2π
2 ln(1 − 2a cos(x) + a2 ) dx = ln(1 − 2a cos(x) + a2 ) dx.
0 0

Furthermore, observe that 1 − 2a cos x + a = (a − e )(a − e−ix ). Thus, our integral is

2 ix

Z 2π  Z 2π Z 2π 
1 1
I= ln[(a − eix )(a − e−ix )]dx = ln(a − eix )dx + ln(a − e−ix )dx ,
2 0 2 0 0

where the integrals are performed on the real parts of the logarithms in the second expression. In the
first integral, substitute z = eix , dz = ieix dx = iz dx; the resulting contour integral is

ln(a − z)
I
dz.
kzk=1 iz

ln(a−z)
By the Residue Theorem, this is equal to 2πi Resz=0 iz = 2π ln(a). The second integral is identical.
Thus, the final answer is 12 (4π ln(a)) = 2π ln(a).
SMT 2011 Advanced Topics Test and Solutions February 19, 2011

1. Five students at a meeting remove their name tags and put them in a hat; the five students then each
randomly choose one of the name tags from the bag. What is the probability that exactly one person
gets their own name tag?
3
Answer: 8
Assume without loss of generality that the first person gets a correct nametag. Let’s call the other
people B, C, D, and E. We can order the four people in nine ways such that none of the persons gets
his own nametag; CBED, CDEB, CEBD, DBEC, DEBC, DECB, EBCD, EDBC, EDCB. Therefore,
9
the desired probability is 4! = 38 .
Alternative Solution: The selection of random nametags amounts to a selection of a random per-
mutation of the five students from the symmetric group S5 . The condition will be met if and only if
the selected permutation σ has exactly one cycle of length one (i.e., exactly one fixed point). The only
5!
distinct cycle types with exactly one fixed point are (1, 4) and (1, 2, 2). There are = 30 permutations
4
5! 30 + 15 3
of the first type and 3 = 15 permutations of the second. Thus, the desired probability is = .
2 5! 8
2. Compute
∞
X (7n + 32) · 3n
.
n=1
n · (n + 2) · 4n

33
Answer: 2
7n+32 16 9
Note that n(n+2) = n − n+2 so that

∞ ∞ ∞
X (7n + 32) 3n X 16 3n X 9 3n
= −
n=1
n(n + 2) 4n n=1
n 4n n=1 n + 2 4n
∞ ∞
X 16 3n X 16 3n+2
= −
n=1
n 4n n=1 n + 2 4n+2
∞ ∞
X 16 3n X 16 3n
= −
n=1
n 4n n=3 n 4n
16 3 16 9 33
= + = .
1 4 2 16 2

3. Find the unique polynomial P (x) with coefficients taken from the set {−1, 0, 1} and with least possible
degree such that P (2010) ≡ 1 (mod 3), P (2011) ≡ 0 (mod 3), and P (2012) ≡ 0 (mod 3).
Answer: P (x) = 1 − x2
First suppose P (x) is constant or linear. Then we have P (2010) + P (2012) = 2P (2011), which is a
contradiction because the left side is congruent to 1 (mod 3) and the right is congruent to 0 (mod 3). So
P must be at least quadratic. The space of quadratic polynomials in x is spanned by the polynomials
f (x) = 1, g(x) = x, and h(x) = x2 . Applying each of these to 2010, 2011, and 2012, we have the mod
3 equivalences:
f (2010, 2011, 2012) ≡ (1, 1, 1)
g(2010, 2011, 2012) ≡ (0, 1, 2)
h(2010, 2011, 2012) ≡ (0, 1, 1)
Subtracting the third row from the first, we have P (x) = f (x)−h(x) = 1−x2 , giving P (2010, 2011, 2012) ≡
(1, 0, 0) (mod 3), as desired. Uniqueness follows from the observation that the three vectors above form
a basis for (Z/3Z)3 .
SMT 2011 Advanced Topics Test and Solutions February 19, 2011

4. Let Tn denote the number of terms in (x + y + z)n when simplified, i.e. expanded and like terms
collected, for non-negative integers n ≥ 0. Find
2010
X
(−1)k Tk = T0 − T1 + T2 − · · · − T2009 + T2010 .
k=0

Answer: 10062
First note that the expression (x + y + z)n is equal to
X n!
xa y b z c
a!b!c!
where the sum is taken over all non-negative integers
a, b, and c with
a + b + c = n. The number of
non-negative integer solutions to a + b + c = n is n+2 2 , so Tk =
k+2
2 for k ≥ 0. It is easy to see that
Tk = 1 + 2 + · · · + (k + 1), so Tk is the (k + 1)st triangular number. If k = 2n − 1 is odd, then for all
positive integers i, T2i − T2i−1 = 2i + 1 and therefore 1
k−1
X n−1
X
(−1)j Tj = T0 + (T2j − T2j−1 )
j=0 j=1
n
X
=1+ (2j − 1)
j=2

= n2 .

Therefore, since T2010 is the 2011th triangular number and 2011 = 2(1006) − 1, we can conclude that
the desired sum is 10062 .
5. Two ants begin on opposite corners of a cube. On each move, they can travel along an edge to an
adjacent vertex. Find the probability they both return to their starting position after 4 moves.
49
Answer: 729
Let the cube be oriented so that one ant starts at the origin and the other at (1, 1, 1). Let x, y, z be
moves away from the origin and x0 , y 0 , z 0 be moves toward the origin in each the respective directions.
Any move away from the origin has to at some point be followed by a move back to the origin, and
if the ant moves in all three directions, then
it can’t get back to its original corner in 4 4!
moves. The
number of ways to choose 2 directions is 32 = 3 and for each pair of directions there are 2!2! = 6 ways
to arrange four moves a, a0 , b, b0 such that a precedes a0 and b precedes b0 . Hence there are 3 · 6 = 18
ways to move in two directions. The ant can also move in a, a0 , a, a0 (in other words, make a move,
return, repeat the move, return again) in three directions so this gives 18 + 3 = 21 moves. There are
34 = 81 possible moves, 21 of which return the ant for a probability of 21 7
81 = 27 . Since this must happen
7 7 49
simultaneously to both ants, the probability is 27 · 27 = 729 .
6. An unfair coin has a 2/3 probability of landing on heads. If the coin is flipped 50 times, what is the
probability that the total number of heads is even?
1+(1/3)50
Answer: 2
The coin can turn up heads 0,2,4,..., or 50 times to satisfy the problem. Hence the probability is
   0  50    2  48    50  0
50 2 1 50 2 1 50 2 1
P = + + ··· + .
0 3 3 2 3 3 50 3 3

Note that this sum is the sum of the even-powered terms of the expansion (1/3 + 2/3)50 . To isolate
these terms, we note that the odd-powered terms of (1/3 − 2/3)50 are negative. So by adding (1/3 +
1 For a quick visual proof of this fact, we refer the reader to http://www.jstor.org/stable/2690575.
SMT 2011 Advanced Topics Test and Solutions February 19, 2011

2/3)50 + (1/3 − 2/3)50 , we get rid of the odd-powered terms and we are left with two times the sum of
the even terms. Hence the probability is

(1/3 + 2/3)50 + (1/3 − 2/3)50 1 + (1/3)50

P = = .
2 2

7. Compute the sum of all n for which the equation 2x + 3y = n has exactly 2011 nonnegative (x, y ≥ 0)
integer solutions.
Answer: 72381
Observe that if the equation ax + by = n has m solutions, the equation ax + by = n + ab has m + 1
solutions. Also note that ax + by = ax0 + by0 for 0 ≤ x0 < b, 0 ≤ y0 < a has no other solution than
(x, y) = (x0 , y0 ). (It is easy to prove both if you consider the fact that the general solution has form
(x0 + bk, y 0 − ak).) So there are ab such n and their sum is
X ab(2ab − a − b)
(ax + by + 2010ab) = 2010a2 b2 + .
2
0≤x<b
0≤y<a

8. Let {ai }i=1,2,3,4 , {bi }i=1,2,3,4 , {ci }i=1,2,3,4 be permutations of {1, 2, 3, 4}. Find the minimum of a1 b1 c1 +
a2 b2 c2 + a3 b3 c3 + a4 b4 c4 .
Answer: 44
The minimum can be obtained by

1 · 3 · 4 + 2 · 2 · 3 + 3 · 4 · 1 + 4 · 1 · 2 = 12 + 12 + 12 + 8 = 44.

We claim that 44 is optimum. Denote xi = ai bi ci . Since x1 x2 x3 x4 = (1 · 2 · 3 · 4)3 = 29 · 33 , xi should

only consist of prime factors of 2 and 3. So between 8 and 12 xi can only be 9.

Case 1. There are no 9 among xi Then xi are not in (8, 12). And x1 x2 x3 x4 = 12 · 12 · 12 · 8, so if x1
is minimum then x1 ≤ 8. Then by AM-GM inequality x2 + x3 + x4 ≥ 3(x2 x3 x4 )1/3 . If we let
(x2 x3 x4 )1/3 = 12y then x1 = 8y −3 , and for y ≥ 1 8y −3 + 36y attains minimum at y = 1. So
x1 + x2 + x3 + x4 ≥ 8y −3 + 36y ≥ 44.
Case 2. x1 is 9. Then x2 x3 x4 is divisible by 3 but not 9. So only x2 is divisible by 3 and others are
just powers of 2. x2 can be 3, 6, 12, 24 or larger than 44.
Case 2-1 x2 = 3 : x3 x4 = 29 , x3 + x4 ≥ 25 + 24 = 48 > 44.
Case 2-2 x2 = 6 : x3 x4 = 28 , x3 + x4 ≥ 24 + 24 = 32, x1 + x2 + x3 + x4 ≥ 9 + 6 + 32 = 47.
Case 2-3 x2 = 12 : x3 x4 = 27 , x3 + x4 ≥ 24 + 23 = 24, x1 + x2 + x3 + x4 ≥ 9 + 12 + 24 = 45.
Case 2-4 x2 = 24 : x3 x4 = 26 , x3 + x4 ≥ 23 + 23 = 16, x1 + x2 + x3 + x4 ≥ 9 + 24 + 16 = 49.

9. How many functions f that take {1, 2, 3, 4, 5} to {1, 2, 3, 4, 5}, not necessarily injective or surjective
(i.e. one-to-one or onto), satisfy f (f (f (x))) = f (f (x)) for all x in {1, 2, 3, 4, 5}?
Answer: 756
For any such function f , let A = {n | f (n) = n} be the set of elements fixed by f and let B = {n |
f (n) ∈ A and n ∈/ A} be the set of elements that are sent to an element in A, but are not themselves in
A. Finally, let C = {1, 2, 3, 4, 5} \ (A ∪ B) be everything else. Note that any possible value of f (f (x))
is in A so A is not empty. We will now proceed by considering all possible sizes of A.
(a) A has one element: Without loss of generality, let f (1) = 1, so we will multiply our result by
5 at the end to account for the other possible values. Suppose that B has n elements so C has
the remaining 4 − n elements. Since f (f (x)) = 1 for each x so any element c in C must satisfy
f (c) = b for some b in B, because f (c) 6= 1 and the only other numbers for which f (x) = 1 are
the elements of B. This also implies that B is not empty. Conversely, any function satisfying
SMT 2011 Advanced Topics Test and Solutions February 19, 2011

P4
f (c) = b works, so the total number of functions in this case is 5 n=1 n4 n4−n because there are


4


n ways to choose the elements in B, and each of the 4 − n elements in C can be sent to any
element of B (there are n of them). This sum is equal to 5(4 + 6 · 4 + 4 · 3 + 1) = 205, so there
are 205 functions in this case that A has one element.
(b) A has two elements: This is similar to the first case, except that each element in B can now
correspond
to one of
two possible elements in A, so this adds a factor of 2n . The sum now
5 P3 3 n 3−n
becomes 2 n=1 n 2 n = 10(3 · 2 + 3 · 4 · 2 + 8) = 380, so there are 380 functions in this
case.
(c) A has three elements: This is again similar to the

prior 2cases, except there are 3 possible targes
5 P2
n

n 2−n
in A, adding a factor of 3 . Then the sum is 3 n=1 n 3 n = 10(2 · 3 + 9) = 150, so there
are 150 functions in this case.
(d) A has four elements: The logic is the same as the prior cases and there are 5(4) = 20 functions in
this case.
(e) A has five elements: The identity function is the only possible function in this case.

Adding together the five cases, we see that there are 205 + 380 + 150 + 20 + 1 = 756 such functions.
10. Find the number of ways of filling a 2 × 2 × 8 box with 16 1 × 1 × 2 boxes (rotations and reflections of
the 2 × 2 × 8 box are considered distinct).
Answer: 23409
Let an be the number of ways of filling the 2 × 2 × n box, and let bn be the number of ways of filling it
with one 1 × 1 × 2 box fixed at the “bottom face” (2 × 2 face). It is easy to see that bn = an−1 + bn−1 .
It is then simple to verify that an = 2bn + 2bn−1 + an−2 . The base cases a1 = 2, b1 = 1, a2 = 9, and
b2 = 3 are trivial to calculate. Using these values to calculate a8 recursively gives a8 = 23409.
SMT 2011 General Test and Solutions February 19, 2011

1. Let F (x) be a real-valued function defined for all real x 6= 0, 1 such that
 
x−1
F (x) + F = 1 + x.
x
Find F (2).
3
Answer: 4
Setting x = 2, we find that F (2) + F 12 = 3. Now take x = 12 , to get that F 12 + F (−1) = 3



2.
Finally, setting x = −1, we get that F (−1) + F (2) = 0. Then we find that
   
1 3 3 3
F (2) = 3 − F =3− − F (−1) = + F (−1) = − F (2)
2 2 2 2
3
⇒ F (2) = .
4
Alternate Solution: We can explicitly solve for F (x) and then plug in x = 2. Notice that for x 6= 0, 1,
F (x) + F x−1


x = 1 + x so
     
x−1 1 x−1 1 1
F +F =1+ and F + F (x) = 1 + .
x 1−x x 1−x 1−x
Thus
       
x−1 x−1 1 1
2F (x) = F (x) + F −F −F +F + F (x)
x x 1−x 1−x
 
x−1 1
=1+x− 1+ +1+
x 1−x
1−x 1
=1+x+ + .
x 1−x
 
It follows that F (x) = 12 1 + x + 1−x
x + 1
1−x and the result follows by taking x = 2.

2. Given that a1 = 2, a2 = 3, an = an−1 + 2an−2 , what is a100 + a99 ?

Answer: 298 × 5

an = an−1 + 2an−2
an + an−1 = 2(an−1 + an−2 )
= 2n−2 (a1 + a2 ).

So a100 + a99 = 298 × 5.

j−1
3. Let sequence A be { 47 , 76 , 79 , . . .} where the j th term is given by aj = 47 23 . Let B be a sequence
where the j th term is defined by bj = a2j + aj . What is the sum of all the terms in B?
861
Answer: 80
49 4 j−1


Split B into two series C and D where the terms of C are cj = a2j = 16 9 and the terms of D are
7 2 j−1


dj = aj = 4 3 . Since both C and D are geometric series with ratios less than 1, the sum of their
49/16 7/4
terms yields 1−4/9 = 441 21 441 21
80 and 1−2/3 = 4 . Therefore, the sum of the terms in B equals 80 + 4 = 80 .
861

4. Find all rational roots of |x − 1| × |x2 − 2| − 2 = 0.

Answer: x = −1, 0, 2
There√are four intervals to consider, each with their own restrictions. Consider the case in which
x > 2. Then the equation becomes (x − 1)(x2 − 2) − 2 = x(x − 2)(x + 1) = 0. Thus, x = 2 is
SMT 2011 General Test and Solutions February 19, 2011

√ √
the only rational root for x > 2. Consider the case in which − 2 < x < 1. Then the equation
2
√ (x − 1)(x − 2) − 2 = x(x − 2)(x + 1) = 0. Thus,
becomes √ x = 0 and x = −1 are the rational
√ roots
for − 2 < x < 1. Consider the case in which x < − 2 or the case in which 1 < x < 2. In these
cases, the equation becomes (1 − x)(x2 − 2) − 2 = −x3 + x2 + 2x − 4. By the rational root theorem, the
rational roots of this polynomial can only be ±4, ±2, ±1 and a quick check shows that none of these
are roots, so this polynomial has no rational roots.

5. Let S = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10}. In how many ways can two (not necessarily distinct) elements a, b
be taken from S such that ab is in lowest terms, i.e. a and b share no common divisors other than 1?
Answer: 63
This amounts to determining, for a given numerator, how many elements in S are relatively prime to
the numerator. If we let f (n) be the number of positive integers relatively prime to n and less than or
equal to 10, it is obvious that f (1) = 10, f (2) = f (4) = f (8) = 5, f (3) = f (9) = 7, f (5) = 8, f (7) = 9,
f (6) = 3, and f (10) = 4. Therefore, the answer is 10 + 3 · 5 + 2 · 7 + 8 + 9 + 3 + 4 = 63.
6. Find all square numbers which can be represented in the form 2a + 3b , where a, b are nonnegative
integers. You can assume the fact that the equation 3x − 2y = 1 has no integer solutions if x ≥ 3.
Answer: 22 , 32 , 52
For b = 0 one has 2a + 1 = c2 , 2a = (c + 1)(c − 1). Thus both c + 1 and c − 1 should be powers of 2.
The only possibility is c = 3, which gives a solution 23 + 30 = 9 = 32 .
For b ≥ 1, 2a + 3b is not divisible by 3, so it should be ≡ 1(mod 3). This requires a to be even. Let
a = 2d, then 3b = c2 − 22d = (c + 2d )(c − 2d ). Let c + 2d = 3p and c − 2d = 3q . Eliminating c, one has
2d+1 = 3p − 3q . For q ≥ 1 the right-hand side is divisible by 3, so q = 0. From what we know, there are
only two solutions (d, p) = (0, 1), (2, 2). These solutions give 20 + 31 = 4 = 22 and 24 + 32 = 25 = 52
respectively.
7. A frog is jumping on the number line, starting at zero and jumping to seven. He can jump from x to
either x + 1 or x + 2. However, the frog is easily confused, and before arriving at the number seven,
he will turn around and jump in the wrong direction, jumping from x to x − 1. This happens exactly
once, and will happen in such a way that the frog will not land on a negative number. How many ways
can the frog get to the number seven?
Answer: 146
Let fn be the number of ways to jump from zero to n, ignoring for the time being jumping backwards..
We have f0 = 1, f1 = 1, and fn = fn−1 + fn−2 when n ≥ 2. Therefore, we have that f2 = 2, f3 = 3,
f4 = 5, f5 = 8, f6 = 13, and f7 = 21. Note that we can describe the frog’s jumping as jumping forward
n numbers, jumping backward 1 number, and jumping forward 8 − n numbers. Therefore, the desired
X6
answer is simply fi f8−i = 146 .
i=1

8. Call a nonnegative integer k sparse when all pairs of 1’s in the binary representation of k are separated
by at least two zeroes. For example, 9 = 10012 is sparse, but 10 = 10102 is not sparse. How many
sparse numbers are less than 217 ?
Answer: 872
Let an denote the number of sparse numbers with no more than n binary digits. In particular, for
numbers with less than n binary digits after removing leading zeroes, append leading zeroes so all
numbers have n binary digits when including sufficiently many leading zeroes. We have that a0 = 1,
a1 = 2, and a2 = 3 since for these lengths, either zero digits are 1 or one digit is 1. We claim that the
recurrence a = an−1 + an−3 holds for n ≥ 3. We split this analysis into two cases; numbers where the
nth binary digit is 0 or 1. When the nth binary digit is zero, we can remove that zero to get a valid
number with n − 1 binary digits. When the nth binary digit is one, it is known that the (n − 1)th and
(n − 2)th digits are both zero, so we can truncate those to get a valid number with n − 3 binary digits.
Therefore, the recurrence holds. With the given initial conditions, a17 = 872 .
SMT 2011 General Test and Solutions February 19, 2011

9. Two ants begin on opposite corners of a cube. On each move, they can travel along an edge to an
adjacent vertex. Find the probability they both return to their starting position after 4 moves.
49
Answer: 729
Let the cube be oriented so that one ant starts at the origin and the other at (1, 1, 1). Let x, y, z be
moves away from the origin and x0 , y 0 , z 0 be moves toward the origin in each the respective directions.
Any move away from the origin has to at some point be followed by a move back to the origin, and
if the ant moves in all three directions, then
it can’t get back to its original corner in 4 4!
moves. The
number of ways to choose 2 directions is 32 = 3 and for each pair of directions there are 2!2! = 6 ways
to arrange four moves a, a0 , b, b0 such that a precedes a0 and b precedes b0 . Hence there are 3 · 6 = 18
ways to move in two directions. The ant can also move in a, a0 , a, a0 (in other words, make a move,
return, repeat the move, return again) in three directions so this gives 18 + 3 = 21 moves. There are
34 = 81 possible moves, 21 of which return the ant for a probability of 21 7
81 = 27 . Since this must happen
7 7 49
simultaneously to both ants, the probability is 27 · 27 = 729 .
10. An unfair coin has a 2/3 probability of landing on heads. If the coin is flipped 50 times, what is the
probability that the total number of heads is even?
1+(1/3)50
Answer: 2
The coin can turn up heads 0,2,4,..., or 50 times to satisfy the problem. Hence the probability is
   0  50    2  48    50  0
50 2 1 50 2 1 50 2 1
P = + + ··· + .
0 3 3 2 3 3 50 3 3

Note that this sum is the sum of the even-powered terms of the expansion (1/3 + 2/3)50 . To isolate
these terms, we note that the odd-powered terms of (1/3 − 2/3)50 are negative. So by adding (1/3 +
2/3)50 + (1/3 − 2/3)50 , we get rid of the odd-powered terms and we are left with two times the sum of
the even terms. Hence the probability is

(1/3 + 2/3)50 + (1/3 − 2/3)50 1 + (1/3)50

P = = .
2 2

11. Find the unique polynomial P (x) with coefficients taken from the set {−1, 0, 1} and with least possible
degree such that P (2010) ≡ 1 (mod 3), P (2011) ≡ 0 (mod 3), and P (2012) ≡ 0 (mod 3).
Answer: P (x) = 1 − x2
First suppose P (x) is constant or linear. Then we have P (2010) + P (2012) = 2P (2011), which is a
contradiction because the left side is congruent to 1 (mod 3) and the right is congruent to 0 (mod 3). So
P must be at least quadratic. The space of quadratic polynomials in x is spanned by the polynomials
f (x) = 1, g(x) = x, and h(x) = x2 . Applying each of these to 2010, 2011, and 2012, we have the mod
3 equivalences:
f (2010, 2011, 2012) ≡ (1, 1, 1)
g(2010, 2011, 2012) ≡ (0, 1, 2)
h(2010, 2011, 2012) ≡ (0, 1, 1)
Subtracting the third row from the first, we have P (x) = f (x)−h(x) = 1−x2 , giving P (2010, 2011, 2012) ≡
(1, 0, 0) (mod 3), as desired. Uniqueness follows from the observation that the three vectors above form
a basis for (Z/3Z)3 .
√
2 3
12. Let a, b ∈ C such that a + b = a2 + b2 = 3 i. Compute | Re(a)|.
1
Answer: √
2
√ √ √ √
From a + b = 2 33 i we can let a = 33 i + x and b = 3
3
i − x. Then a2 + b2 = 2(( 33 i)2 + x2 ) =
√ √ √ √
2(x2 − 31 ) = 2 3 3 i. So x2 = 1+3 3i = 23 eiπ/3 , x = ± √23 · 3+i
2 . Since | Re(a)| = | Re(x)|, the answer is
√ √
√2 · 3
= √1 .
3 2 2
SMT 2011 General Test and Solutions February 19, 2011

13. Let Tn denote the number of terms in (x + y + z)n when simplified, i.e. expanded and like terms
collected, for non-negative integers n ≥ 0. Find
2010
X
(−1)k Tk
k=0
= T0 − T1 + T2 − · · · − T2009 + T2010 .

Answer: 10062
First note that the expression (x + y + z)n is equal to
X n!
xa y b z c
a!b!c!
where the sum is taken over all non-negative integers
a, b, and c with
a + b + c = n. The number of
non-negative integer solutions to a + b + c = n is n+2 2 , so Tk = k+2
2 for k ≥ 0. It is easy to see that
Tk = 1 + 2 + · · · + (k + 1), so Tk is the (k + 1)st triangular number. If k = 2n − 1 is odd, then for all
positive integers i, T2i − T2i−1 = 2i + 1 and therefore 1
k−1
X n−1
X
(−1)j Tj = T0 + (T2j − T2j−1 )
j=0 j=1
n
X
=1+ (2j − 1)
j=2

= n2 .

Therefore, since T2010 is the 2011th triangular number and 2011 = 2(1006) − 1, we can conclude that
the desired sum is 10062 .
14. Let M = (−1, 2) and N = (1, 4) be two points in the plane, and let P be a point moving along the
x-axis. When ∠M P N takes on its maximum value, what is the x-coordinate of P ?
Answer: 1
Let P = (a, 0). Note that ∠M P N is inscribed in the circle defined by points M , P , and N , and that it
intercepts M N . Since M N is fixed, it follows that maximizing the measure of ∠M P N is equivalent to
minimizing the size of the circle defined by M , P , and N . Since P must be on the x-axis, we therefore
want this circle to be tangent to the x-axis. Since the center of this circle must lie on the perpendicular
bisector of M N , which is the line y = 3 − x, the center of the circle has to be of the form (a, 3 − a), so
a has to satisfy (a + 1)2 + (1 − a)2 = (a − 3)2 . Solving this equation gives a = 1 or a = −7. Clearly
choosing a = 1 gives a smaller circle, so our answer is 1.
15. Consider the curves x2 + y 2 = 1 and 2x2 + 2xy + y 2 − 2x − 2y = 0. These curves intersect at two
points, one of which is (1, 0). Find the other one.
Answer: − 53 , 45

From the first equation, we get that y 2 = 1 − x2 . Plugging this into the second one, we are left with
p p p
2x2 ± 2x 1 − x2 + 1 − x2 − 2x ∓ 2 1 − x2 = 0 ⇒ (x − 1)2 = ∓2 1 − x2 (x − 1)
p
⇒ x − 1 = ∓2 1 − x2 assuming x 6= 1
⇒ x2 − 2x + 1 = 4 − 4x2 ⇒ 5x2 − 2x − 3 = 0.
The quadratic formula yields that x = 2±8 3
10 = 1, − 5 (we said that x 6= 1 above but we see that it is still
valid). If x = 1, the first equation forces y = 0 and we easily see that this solves the second equation.
If x = − 35 , then clearly y must be positive or else the second equation will sum five positive terms.
q q
9
= 16 4 3 4


Therefore y = 1 − 25 25 = 5 . Hence the other point is − 5 , 5 .

1 For a quick visual proof of this fact, we refer the reader to http://www.jstor.org/stable/2690575.
SMT 2011 General Test and Solutions February 19, 2011

16. If r, s, t, and u denote the roots of the polynomial f (x) = x4 + 3x3 + 3x + 2, find
1 1 1 1
+ 2 + 2 + 2.
r2 s t u
9
Answer: 4
First notice that the polynomial
 
4 1 3 3
g(x) = x + 3 + +2 = 2x4 + 3x3 + 3x + 1
x4 x x

is a polynomial with roots 1r , 1s , 1t , u1 . Therefore, it is sufficient to find the sum of the squares of the
roots of g(x), which we will denote as r1 through r4 . Now, note that
a3 2 a2
r12 + r22 + r32 + r42 = (r1 + r2 + r3 + r4 )2 − (r1 r2 + r1 r3 + r1 r4 + r2 r3 + r2 r4 + r3 r4 ) = (− ) −
a4 a4
by Vieta’s Theorem, where an denotes the coefficient of xn in g(x). Plugging in values, we get that
our answer is (− 32 )2 − 0 = 94 .
17. An icosahedron is a regular polyhedron with 12 vertices, 20 faces, and 30 edges. How many rigid
rotations G are there for an icosahedron in R3 ?
Answer: 60
There are 12 vertices, each with 5 neighbors. Any vertex and any of its neighbors can be rotated to
any other vertex-neighbor pair in exactly one way. There are 5 · 12 = 60 vertex-neighbor pairs.
18. Pentagon ABCDE is inscribed in a circle of radius 1. If ∠DEA ∼
= ∠EAB ∼
= ∠ABC, m∠CAD = 60◦ ,
and BC = 2DE, compute the area of ABCDE.
√
33 3
Answer: 28
Looking at cyclic quadrilaterals ABCD and ACDF tells us that m∠ACD = m∠ADC, so 4ACD is
equilateral and m∠DEA = 120◦ . Now, if we let m∠EAD = θ, we see that m∠CAB = 60◦ − θ =⇒
m∠ACB = θ =⇒ 4AED ∼ = 4CBA. Now all we have to do is √
calculate side lengths. After creating
some 30◦ − 60◦ − 90◦ triangles, it becomes evident that AC = 3. Now let AB = x, so BC = 2x.
By applying the Law of Cosines to triangle ABC, we find that x2 = 73 . Hence, the desired area
√ √ √
( 3)2 3
(ABCDE) = (ACD) + 2(ABC) = 4 + 2 · 21 (x)(2x)(sin 120◦ ) = 33 3
28 .

19. Five students at a meeting remove their name tags and put them in a hat; the five students then each
randomly choose one of the name tags from the bag. What is the probability that exactly one person
gets their own name tag?
3
Answer: 8
Assume without loss of generality that the first person gets a correct nametag. Let’s call the other
people B, C, D, and E. We can order the four people in nine ways such that none of the persons gets
his own nametag; CBED, CDEB, CEBD, DBEC, DEBC, DECB, EBCD, EDBC, EDCB. Therefore,
9
the desired probability is 4! = 38 .
Alternative Solution: The selection of random nametags amounts to a selection of a random per-
mutation of the five students from the symmetric group S5 . The condition will be met if and only if
the selected permutation σ has exactly one cycle of length one (i.e., exactly one fixed point). The only
5!
distinct cycle types with exactly one fixed point are (1, 4) and (1, 2, 2). There are = 30 permutations
4
5! 30 + 15 3
of the first type and 3 = 15 permutations of the second. Thus, the desired probability is = .
2 5! 8
20. Find the 2011th-smallest x, with x > 1, that satisfies the following relation:

sin(ln x) + 2 cos(3 ln x) sin(2 ln x) = 0.

SMT 2011 General Test and Solutions February 19, 2011

Answer: x = e2011π/5
Set y = ln x, and observe that

2 cos(3y) sin(2y) = sin(3y + 2y) − sin(3y − 2y) = sin(5y) − sin(y),

so that the equation in question is simply

sin(5y) = 0.

The solutions are therefore

nπ
ln x = y = =⇒ x = enπ/5 for all n ∈ N
5

21. An ant is leashed up to the corner of a solid square brick with side length 1 unit. The length of the
ant’s leash is 6 units, and it can only travel on the ground and not through or on the brick. In terms
of x = arctan 34 , what is the area of region accessible to the ant?

79π 3
Answer: 2
+ 2
− 5x

Label the top left corner of the square as the origin O. By keeping the leash straight, the ant can
travel through 34 of a circle of radius 6 (A1 = 34 × 36π = 27π). The ant can also bend the leash around
the two nearest corners of the square to where it is leashed (A2 = 2 × 41 × 25π = 25 2 π). However, this
double counts the area enclosed by BCDE, which is equal to two times the area of BCD. To calculate
the latter, notice that ACD is the sector of the circle centered at A with radius 5. We can calculate the
coordinates of D usign the two equations y = −x (from the symmetry) and
x2 + (y + 1)2 = 52 which
yields D = (4, −4). Since A = (0, −1), the angle of sector ACD is arctan 34 = x. The area of triangle
ABD equals 23 (base times height)
so BCD has area 5x − 32 and BCDE has area 10x − 3. Hence, the
total area is A1 + A2 − 5x − 2 = 2 + 23 − 5x.
3 79π

22. Compute the sum of all n for which the equation 2x + 3y = n has exactly 2011 nonnegative (x, y ≥ 0)
integer solutions.
Answer: 72381
Observe that if the equation ax + by = n has m solutions, the equation ax + by = n + ab has m + 1
solutions. Also note that ax + by = ax0 + by0 for 0 ≤ x0 < b, 0 ≤ y0 < a has no other solution than
(x, y) = (x0 , y0 ). (It is easy to prove both if you consider the fact that the general solution has form
(x0 + bk, y 0 − ak).) So there are ab such n and their sum is
X ab(2ab − a − b)
(ax + by + 2010ab) = 2010a2 b2 + .
2
0≤x<b
0≤y<a

23. Let ABC be any triangle, and D, E, F be points on BC, CA, AB such that CD = 2BD, AE = 2CE
and BF = 2AF . AD and BE intersect at X, BE and CF intersect at Y , and CF and AD intersect
Area(4ABC)
at Z. Find Area(4XY Z) .
SMT 2011 General Test and Solutions February 19, 2011

Answer: 7
Using Menelaus’s Theorem on 4ABD with collinear points F, X, C and the provided ratios gives
DX/XA = 4/3. Using Menelaus’s Theorem on 4ADC with collinear points B, Y, E gives AY /Y D = 6.
We conclude that AX, XY, Y D are in length ratio 3 : 3 : 1. By symmetry, this also applies to the
segments CZ, ZX, XF and BY, Y Z, ZE. Repeatedly using the fact that the area ratio of two triangles
of equal height is the ratio of their bases, we find [ABC] = (3/2)[ADC] = (3/2)(7/3)[XY C] =
(3/2)(7/3)(2)[XY Z] = 7[XY Z], or [ABC]/[XY Z] = 7.
Alternate Solution
Stretching the triangle will preserve ratios between lengths and ratios between areas, so we may assume
that 4ABC is equilateral with side length 3. We now use mass points to find the length of XY . Assign
a mass of 1 to A. In order to have X be the fulcrum of 4ABC, C have mass 2 and B must have
mass 4. Hence, BX : XE = 4 : 3 and AX : XD = 6 : 1, the latter of which also equals BY : Y E
by symmetry. Hence, XY = 73 BE. To find BE, we apply the Law of Cosines to 4CBE to get that
√
BE 2 = 12 + 32 − 2 · 1 · 3 · cos 60◦ = 7 =⇒ XY = 3 7 7 . Since 4XY Z must be equilateral by symmetry,
AB 2
the desired ratio equals ( XY ) = 7.

24. Let P (x) be a polynomial of degree 2011 such that P (1) = 0, P (2) = 1, P (4) = 2, ... , and P (22011 ) =
2011. Compute the coefficient of the x1 term in P (x).
1
Answer: 2 − 22010
We analyze Q(x) = P (2x) − P (x). One can observe that Q(x) − 1 has the powers of 2 starting from
1, 2, 4, · · · , up to 22010 as roots. Since Q has degree 2011, Q(x) − 1 = A(x − 1)(x − 2) · · · (x − 22010 )
for some A. Meanwhile Q(0) = P (0) − P (0) = 0, so

Q(0) − 1 = −1 = A(−1)(−2) · · · (−22010 ) = −2(2010·2011)/2 A.

Therefore A = 2−(1005·2011) . Finally, note that the coefficient of x is same for P and Q − 1, so it equals
1005·2011
(22011 −1) 1
A(−20 )(−21 ) · · · (−22010 )((−20 ) + (−2−1 ) + · · · + (−2−2010 )) = A·2 22010 = 2 − 2010 .
2
25. Find the maximum of
ab + bc + cd
a2 + b2 + c2 + d2
for reals a, b, c, and d not all zero.
√
5+1
Answer: 4
One has ab ≤ 2t a2 + 1 2
2t b , bc ≤ 12 b2 + 12 c2 , and cd ≤ 1 2
2t c + 2t d2 by AM-GM. If we can set t such that
t 2 2 2 2
t 1 2 (a +b +c +d )
2 = 2t + 12 , it can be proved that a2ab+bc+cd
+b2 +c2 +d2 ≤ = 2t , and this is maximal because we
a2 +b2 +c2 +d2
can set√a, b, c, d so that the equality√ holds in every inequality we used. Solving this equation, we get
t = 1+2 5 , so the maximum is 2t = 5+1 4 .
SMT 2011 Team Test and Solutions February 19, 2011

1. Let ABCD be a unit square. The point E lies on BC and F lies on AD. 4AEF is equilateral. GHIJ
is a square inscribed in 4AEF so that GH is on EF . Compute the area of GHIJ.

√
Answer: 312 − 180 3
√ √
First√let a be the length of AE. Then√CE = a/ 2, BE√= 1 − a/ 2 so√AE 2 √ = a2 = 1 + BE 2 =
2 2 2
2 − 2a + a /2. Solving it gives a + 2 2a − 4 = 0, (a + 2) = 6 so a = 6 − 2. √
Next let b be the length of IJ. Then AIJ is equilateral so AJ = b. Also JE = 2/ 3b, so AE = a =
2+
√ √ √ √ √ √ √ √
√ 3 b, b = (2 − 3)( 3)( 6 − 2) = 2(9 − 5 3). Squaring it gives 312 − 180 3.
3

2. Find all integers x for which |x3 + 6x2 + 2x − 6| is prime.

Answer: 1, −1
The whole equation is ≡ 0 (mod 3), so x3 + 6x2 + 2x − 6 should be 3 or −3. The equation (x3 + 6x2 +
2x − 6)2 = 32 can be rewritten using difference of squares as (x − 1)(x2 + 7x − 9)(x + 1)(x2 + 5x − 3) = 0,
so only 1 and −1 work for x.

3. Let A be the set of points (a, b) with 2 < a < 6, −2 < b < 2 such that the equation

ax4 + 2x3 − 2(2b − a)x2 + 2x + a = 0

has at least one real root. Determine the area of A.

Answer: 12
After dividing the equation by 4x2 , we can re-write it as
 2  
x 1 x 1
a + + + − a = b.
2 2x 2 2x

Set y = x2 + 2x
1
, which has range (−∞, −1] ∪ [1, ∞). Therefore, we need all b in (−2, 2) such that b is in
the range of f (y) = ay 2 + y − a for the domain y ∈ (−∞, −1] ∪ [1, ∞). The vertex of this parabola lies
1
at y = − 2a ∈ (−1/4, −1/12), so the desired range is just all values greater than f (−1) = −1. Hence,
A is the set of all points where −1 < b < 2 and 2 < a < 6, so the area is 12.
4. Three nonnegative reals x, y, z satisfy x + y + z = 12 and xy + yz + zx = 21. Find the maximum of
xyz.
Answer: 10
Consider the graphs of y = t3 − 12t2 + 21t and y = p(p ≤ 0). These two graphs intersect at three
points (counting multiplicity) if and only if there are three nonnegative x, y, z satisfying xyz = p. In
order for these two to intersect at three points, p should lie between the local maximum and the local
minimum of the cubic function y = t3 − 12t2 + 21t, so the maximal p will lie at the local maximum of
this cubic. Since y 0 = 3t2 − 24t + 21 = 3(t − 1)(t − 7), the local maximum occurs at t = 1, so the local
maximum is 13 − 12 · 12 + 21 · 1 = 10 (this can be achieved by letting (x, y, z) = (1, 1, 10)).
SMT 2011 Team Test and Solutions February 19, 2011

5. Let 4ABC be equilateral. Two points D and E are on side BC (with order B, D, E, C), and satisfy
∠DAE = 30◦ . If BD = 2 and CE = 3, what is BC?
A

30◦

B 2 D E 3 C
√
Answer: 5 + 19
Rotate the figure around A by 60◦ so that C coincides with B. Let B 0 , C 0 , D0 , E 0 be the points
corresponding to B, C, D, E in the rotated figure. Since ∠E 0 AD = ∠E 0 AC 0 + ∠C 0 AD = ∠EAC +
∠BAD = 30◦ = ∠EAD, E 0 A = EA and DA√= D0 A, one has E 0 D = ED. So BC = BD √ + DE + EC
can √ know E 0 D. But E 0 D = E 0 B 2 + BD2 − 2 · E 0 B · BD · cos 120◦ = 19, so BC =
be found if we √
2 + 19 + 3 = 5 + 19.
A
B0
D0

E0

B = C0 D E C

6. Three numbers are chosen at random between 0 and 2. What is the probability that the difference
between the greatest and least is less than 14 ?
11
Answer: 256
Call the three numbers x, y, and z. By symmetry, we need only consider the case 2 ≥ x ≥ y ≥ z ≥ 0.
Plotted in 3D, the values of (x, y, z) satisfying these inequalities form a triangular pyramid with a leg-2
right isosceles triangle as its base and a height of 2, with a volume of 2 · 2 · 12 · 2 · 13 = 43 . We now
need the volume of the portion of the pyramid satisfying x − z ≤ 41 . The equation z = x − 14 is a
plane which slices off a skew triangular prism along with a small triangular pyramid at one edge of
the large triangular pyramid. The prism has a leg- 14 right isosceles triangle as its base and a height
of 74 , so has volume 41 · 14 · 21 · 74 = 277 . The small triangular pyramid also has a leg- 14 right isosceles
1 1 1 1 1 1 1
7 1 4

and a height of 4 , so has volume 4 · 4 · 2 · 4 · 3 = 3·27 . Then our probability is
triangle as
its base
27 + 3·27 / 3 = 11/256.

7. Tony the mouse starts in the top left corner of a 3x3 grid. After each second, he randomly moves to an
adjacent square with equal probability. What is the probability he reaches the cheese in the bottom
right corner before he reaches the mousetrap in the center?
1
Answer: 7
Let x be the probability that Tony reaches the cheese before the mousetrap, starting from the top left.
Let y be the probability that Tony reaches the cheese before the mousetrap, starting from the top right
or the bottom left (which are symmetric).
After 2 moves from the top left there is 31 chance that Tony returns to the top left corner, there is
1 1
3 chance that Tony reaches the mousetrap, and there is 3 chance that Tony reaches the top right or
bottom left corners. This gives us the relation
SMT 2011 Team Test and Solutions February 19, 2011

1 1 1
x= x + 0 + y.
3 3 3
After 2 moves from the top right corner there is 31 chance that Tony returns to the top right corner, 13
chance that Tony reaches the mousetrap, 61 chance that Tony reaches the top left corner, and 16 chance
that Tony reaches the cheese. This gives the relation

1 1 1 1
y= y+ 0+ x+ .
3 3 6 6
Now we have a system of linear of equations and we solve, obtaining x = 71 .
8. Let A = (0, 0), B = (1, 0), and C = (0, 1). Divide AB into n equal segments, and call the endpoints
of these segments A = B0 , B1 , B2 , · · · , Bn = B. Similarly, divide AC into n equal segments with
endpoints A = C0 , C1 , C2 , · · · , Cn = C. By connecting Bi and Cn−i for all 0 ≤ i ≤ n, one gets
a piecewise curve consisting of the uppermost line segments. Find the equation of the limit of this
piecewise curve as n goes to infinity.

√ √
Answer: x + y = 1 or equivalent form
The limiting curve is the boundary of a region given by the union of all line segments connecting (q, 0)
y
and (0, 1 − q) for all numbers 0 ≤ q ≤ 1. Such a line segment has equation xq + 1−q = 1. Thus a
y
point (x0 , y0 ) is in that region if and only if the equation xq + 1−q = 1, (1 − q)x + qy = q(1 − q) has
a solution in 0 ≤ q ≤ 1. Let F (q) = (1 − q)x + qy − q(1 − q) = q 2 − (1 + x − y)q + x. Note that
F (0) = x ≥ 0 and F (1) = y ≥ 0, and the minimum of F at 1+x−y 2 is always between 0 and 1. So F has
1+x−y (1+x−y)2 √
a root in √
[0, 1] if and only√if F ( 2 ) =√ − + x ≤ 0. So 4x√ ≤ (1 + x − y)2 , 2 x ≤ 1 + x − y,
√ 4 √
y ≤ 1 − 2 x + x = (1 − x)2 , y ≤ 1 − x, and finally we have x + y ≤ 1.
9. Determine the maximum number of distinct regions into which 2011 circles of arbitrary size can par-
tition the plane.
Answer: 20112 − 2011 + 2 = 4042112
Let f (n) denote the maximum number of regions into which n circles can partition the plane. We will
show that f (n) is a quadratic polynomial in n. Indeed, let A be a planar arrangement of n circles.
Note that A is a graph: Each intersection point is a vertex, and the arcs connecting them are edges.
Having recognized this, we can apply Euler’s theorem, V − E + F = 2, to find the number of regions
(i.e., F ). It is easy to see that an arrangement
  with the maximum number of vertices is optimal. The
n
maximum number of vertices is V = 2 = n(n − 1), since each circle can intersect each other circle
2
in at most two vertices. In this optimal arrangement, each circle contains 2(n − 1) vertices and the
same number of edges; thus, the total number of edges is E = 2n(n − 1). Thus, the desired quantity
is f (n) = E − V + 2 = n2 − n + 2, so our answer is 20112 − 2011 + 2 = 4042112.
Alternative Solution: As before, we apply Euler’s theorem for planar graphs. Given that circles
are defined by quadratic polynomials, it is clear that V and E are each quadratic in n. In particular,
SMT 2011 Team Test and Solutions February 19, 2011

Euler’s theorem implies that F is quadratic in n. Moreover, it is easy to check that f (1) = 2, f (2) = 4,
and f (3) = 8. Interpolating gives f (n) = n2 − n + 1, as in the first solution.
xyz(x + y + z)
10. For positive reals x, y, and z, compute the maximum possible value of .
(x + y)2 (y + z)2
1
Answer: 4
If we consider the triangle ABC with side length AB = x + y, BC = y + z, CA = z + x, the equation
becomes
|ABC|2 sin2 B 1
= ≤ .
AB 2 · BC 2 4 4

11. Find the diameter of an icosahedron with side length 1 (an icosahedron is a regular polyhedron with
20 identical equilateral triangle faces; a picture is given below).

√ √
10+2 5
Answer: 4
Note that opposite vertices of the icosahedron can be seen as vertices of pyramids whose bases are
regular pentagons of side length 1. Now, note that if we select two parallel diagonals of these two
pentagons, these two diagonals are also two sides of a rectangle whose other sides are length 1 and
whose diagonals are diameters of the icosahedron. The diagonal of the pentagon can be found with
similar triangles: in regular pentagon ABCDE, let AD and BE intersect at F. Angle chasing shows that
√
4ACD ∼ 4DEF , both are isosceles, and F E = F A, so we get that AD 1
= AD−1 =⇒ AD = 1+2 5 .
q √ √ 1
√
Hence, the diameter of the icosahedron equals 12 + ( 1+2 5 )2 = 10+22
5
.

12. Find the boundary of the projection of the sphere x2 + y 2 + (z − 1)2 = 1 onto the plane z = 0 with
respect to the point P = (0, −1, 2). Express your answer in the form f (x, y) = 0, where f (x, y) is a
function of x and y.
Answer: x2 − 4y − 4 = 0
Let O = (0, 0, 1) be the center of the sphere. For a point X = (x, y, 0) on the boundary of the
projection, the angle ∠XP O is constant as X varies, since it is just the angle between OP and any
tangent from P to the sphere. Considering the case when X = (0, −1, 0), we can see that ∠XP O = 45◦ .
−−→ −−→ −−→ −−→
Writing this in terms of the dot product, one has (P O · P X)2 = 21 |P O|2 |P X|2 , which is equivalent to
((0, 1, −1) · (x, y + 1, −2))2 = 12 |(0, 1, −1)|2 |(x, y + 1, −2)|2 , or (y + 3)2 = x2 + (y + 1)2 + 4. The answer
is x2 − 4y − 4 = 0.
13. Compute the number of pairs of 2011-tuples (x1 , x2 , ..., x2011 ) and (y1 , y2 , ..., y2011 ) such that xk =
x2k−1 − yk−1
2
− 2 and yk = 2xk−1 yk−1 for 1 ≤ k ≤ 2010, x1 = x22011 − y2011
2
− 2, and y1 = 2x2011 y2011 .
Answer: 22011
Define zk = xk + iyk . Then the equations are equivalent to zk+1 = zk 2 − 2, z2012 = z1 . Let α be a
solution of z1 = α + α−1 (which always has two distinct solutions unless z1 = 2 or −2). Then one can
k−1 k−1 2011 2011
check by induction that zk = α2 + α−2 . Since one has z2012 = z1 , α2 + α−2 = α + α−1 .
SMT 2011 Team Test and Solutions February 19, 2011

Set N = 22011 and rewrite the above as α2N + 1 = αN −1 + αN +1 , or (αN +1 − 1)(αN −1 − 1) = 0. Since
N is even, N + 1 and N − 1 are relatively prime. So the equations X N +1 = 1 and X N −1 = 1 have
only the root 1 in common. Therefore there are (N + 1) + (N − 1) − 1 = 2N − 1 possibilities for α.
Meanwhile, any one value of z1 = α + α−1 corresponds to two choices of α except when α = 1 or −1.
So our 2N − 2 choices of α 6= 1 together give N − 1 different solutions for z1 , and α = 1 give a single
solution z = 2. The answer is N = 22011 .
Z 1
ln(x + 1)
14. Compute I = dx.
0 x2 + 1
π ln(2)
Answer: 8
ln(x + 1)
Let I denote the integral we wish to compute. The function f (x) = does not have an
x2 + 1
elementary antiderivative. We will use Taylor series to compute I. We can find the Taylor series for
ln(x + 1)
the function using the following formulas:
x2 + 1
x2 x3
ln(x + 1) = x − + − ...
2 3
1
= 1 − x2 + x4 − . . .
1 + x2
These formulas aren’t good everywhere, but they do hold in (0, 1). We have

x2 x3 x4
 
+ . . . 1 − x2 + x4 − x6 + . . .


f (x) = x − + −
2 3 4
       
1 2 1 3 1 1 4 1 1
=x+ − x + −1 x + − + x + − + 1 x5 + . . .
2 3 4 2 5 3

In particular, an antiderivative is given by

       
1 2 1 1 3 1 1 4 1 1 1 5 1 1 1
F (x) = x + − x + −1 x + − + x + − + 1 x6 + . . .
2 3 2 4 3 5 4 2 6 5 3

The definite integral I is given by F (1), i.e., the sum

       
1 1 1 1 1 1 1 1 1 1 1
I= + − + −1 + − + + − + 1 + ...
2 3 2 4 3 5 4 2 6 5 3

Now we use the facts that

1 1 1 π
1− + − + ... =
3 5 7 4
1 1 1
1 − + − + . . . = ln(2),
2 3 4
from the Taylor series for tan−1 (x) and ln(x + 1) respectively. Notice that in the above sum, every
1
number of the form , where r is even and s is odd, occurs exactly once, with a positive sign if
r·s
r + s ≡ 3 (mod 4) and a negative sign if 1 (mod 4). Therefore, it is clear that
  
1 1 1 1 1 1 1
I = 1 − + − + ... − + − + ...
3 5 7 2 4 6 8
 
π 1 1 1 1
= · 1 − + − + ...
4 2 2 3 4
π ln(2)
= .
8
SMT 2011 Team Test and Solutions February 19, 2011

15. Find the smallest α > 0 such that there exists m > 0 making the following equation hold for all positive
integers a, b ≥ 2:  
1 1
+ (a + b)α ≥ m.
gcd(a, b − 1) gcd(a − 1, b)
1
Answer: 2
Note that both gcd(a, b − 1) and gcd(a − 1, b) divide a + b − 1. Also they are relatively prime, since
gcd(a, b − 1) | a and gcd(a − 1, b) | a − 1. So their product is less than or equal to a + b − 1, and
therefore by the AM-GM inequality we have
s
1 1 1 2
+ ≥2 ≥√ .
gcd(a, b − 1) gcd(a − 1, b) gcd(a, b − 1) · gcd(a − 1, b) a+b−1

Thus α = 12 and m = 2 suffice. To show that there is no such m for smaller α, let b = (a − 1)2 . Then
gcd(a, b − 1) = a and gcd(a − 1, b) = a − 1, so

(2a − 1)(a2 − a + 1)α

 
1 1
+ (a + b)α =
gcd(a, b − 1) gcd(a − 1, b) a(a − 1)

and the limit when a goes to ∞ is zero if α < 12 .

 2011 SMT POWER ROUND: POLES AND POLARS

Time limit: 50 minutes.

Instructions: For this test, you work in teams of eight to solve a multi-part, proof-oriented question.
Answers are to be submitted on clearly labeled sheets of scratch paper. Every problem must be written on
its own sheet. Every submitted page should contain the team number, problem number, and a page number.
No calculators.

1. Definition and Basic Properties

In the Cartesian plane, the polar of the point P = (px , py ) with respect to the unit circle is the line
px x + py y = 1. The pole of the line l : lx x + ly y = 1 with respect to the unit circle is the point (lx , ly ).
1. Graph the polars (or the poles) of the following figures, together with the original figures (separate graph
for each part). Also draw your coordinate axes. You don’t need to justify. [2 pts each]
(1) A(1, 1) (the “A” is just a label for the point (1, 1))
(2) b : y = x + 2
(3) C(−0.6, 0.8)
(4) d : x = 2y + 2
(5) E : the pole of line a, the polar of A
(6) f : the polar of the intersection of lines a and b
2. How many intersections are there between the polar of P and the unit circle, given that P is (1) inside,
(2) on, and (3) outside the unit circle? Justify. [12 pts]
The reciprocation transform takes a point which is not the origin to its polar, and a line which do not go
through the origin to its pole. It is easy to see that reciprocation establishes a bijection between non-origin
points and lines which do not go through the origin, and it is its own inverse: l is the polar of the point P
if and only if P is the pole of the line l.
3. The more common definitions of poles and polars are the following:
Define the inversion of the nonorigin point P with respect to the unit circle as follows: it is the point
P 0 on the ray OP satisfying OP · OP 0 = 1. Then define the polar of P to be the line going through P 0
which is perpendicular to OP . Also for a line l which does not go through the origin, define the pole of l
to be the inversion of the foot of the perpendicular from O to l.
Show that this definition is equivalent to our initial definition. [12 pts]
(For future reference, if you take the inversion of all the points on a circle that goes through the origin,
you get a line. If you take the inversion of all the points on a circle that does NOT go through the origin,
you get another circle.)
4. Generalize the concept of reciprocation to reciprocation around any circle (x − a)2 + (y − b)2 = r2 (think
of the reciprocation we have defined as reciprocation around the unit circle). Find the equation of the
polar of (p, q) (where (p, q) 6= (a, b)) and the coordinates of the pole of c(x − a) + d(y − b) = 1 (where
(c, d) 6= (0, 0)). Your definition should be consistent with the rescaling and translation, but you don’t
need to justify your answer. [12 pts]

2. The Duality Principle

For the remaining problems you don’t need to use generalized reciprocation unless noted.
5. (1) Prove that A is on the polar of B if and only if B is on the polar of A using the definition given in
section 1. (2) Prove the same thing using the definition given in problem 3. [12 pts]
6. The following are corollaries of 5. Let A, B, C be non-origin points, and a, b, c be their respective polars.
Prove the following: [7 pts]
(a) The point A is on b if and only if the point B is on a.
(b) The pole of AB is the intersection of a and b. Conversely, the polar of the intersection of a and b is
AB.
Date: February 19, 2011.
1
 2 2011 SMT POWER ROUND: POLES AND POLARS

(c) Points A, B, C are collinear (on the same line) if and only if lines a, b, c are concurrent (go through
the same point).
In geometry, duality refers to geometric transformations that replace points by lines and lines by points
while preserving incidence properties, the relation of a line going through a point. It is easy to see that
reciprocation is a duality for nonorigin points and lines not going through the origin. This leads to a
general principle called the duality principle: any theorem about incidences between points and lines may be
transformed into another theorem about lines and points, by a substitution of the appropriate words. The
transformed theorem is sometimes called the dual theorem or reciprocal.
7. State the dual theorem of the following. [6 pts]
(Pappus’s theorem) Given A, B, C collinear and D, E, F collinear (not necessarily in that order), the
three intersection points X = BF ∩ CE, Y = AF ∩ CD, and Z = AE ∩ BD are also collinear.
8. Explain how to generalize the duality principle to theorems which also include incidences between a given
circle and some points. Use this generalization to state the dual theorem of the following. Your generalized
reciprocation from problem 4 will be useful. [10 pts]
(Pascal’s theorem) For a cyclic hexagon ABCDEF (not necessarily around the circle in that order),
the three intersections of opposite sides X = AB ∩ DE, Y = BC ∩ EF , Z = CD ∩ F A are collinear.

3. Reciprocation and Cyclic Quadrilaterals

9. Given a circle, let P be a point inside the circle, and AC and BD be two chords of the circle which go
through P . Let Q be the intersection of the lines AD and BC (outside the circle). Then Q is on the
polar of P . Prove this statement following the steps below: [21 pts]
(1) Let X be the point on the segment AD satisfying AX/XD = AQ/QD. Supposing that the circle
is the unit circle, points A and D have coordinates A = (xa , ya ) and D = (xd , yd ), and the ratio
AQ : QD = m : n (where m, n 6= 0), express the coordinates of Q and X in terms of xa , ya , xd , yd , m, n.
We say that X is the harmonic conjugate of the three points A, D, Q.
(2) Show that X is on the polar of Q.
(3) Define Y be the harmonic conjugate of B, C, Q. Show that AC, XY, BD intersect at one point. To
show this, let P1 and P2 be the intersection of AC and XY , BD and XY respectively, and calculate the
ratio P1 X/P1 Y and P2 X/P2 Y using Menelaus’ theorem (if ABC is a triangle and P, Q, R are points
BQ CR
on the lines AB, BC, and CA respectively, then P, Q, R are collinear if and only if PAP B QC RA = −1)
to see that those two coincide.
(4) Conclude that P is on the polar of Q.
10. For cyclic quadrilateral ABCD, we can consider three points P = AC ∩ BD, Q = AD ∩ BC and
R = AB ∩ CD. Show that the polar of each point goes through other two ponts. What can be said about
the orthocenter of P QR? [12 pts]
11. Prove the following “six points on the line” property: for cyclic quadrilateral ABCD we have (1) the
intersection of AC and BD, (2) the intersection of AD and BC, (3) the intersection of the tangent at A
and the tangent at B, (4) the intersection of the tangent at C and the tangent at D, and (5) and (6) the
tangent points of the two tangents from AB ∩ CD to the circle. The property is that all of these points
are on the same line. [12 pts]
12. This is not related to poles and polars, but can you prove that (7) the intersection of the circumcircles of
ADP and BCP (other than P ) is also on the line? [12 pts]

4. Conic Sections
13. Take a circle centered at A with radius r, a point O inside the circle, and a circle centered at O with
radius 1. The poles of tangents to circle A with respect to circle O form a conic section. (1) Is it an
ellipse, parabola, or hyperbola? (2) What is its relationship to the polars of the points on circle A with
respect to circle O? What are the conic sections we get if O is (3) on circle A, or (4) outside? You should
justify each of your choices under the assumption that these loci are indeed conic sections, but you do
not need to check their precise shape. [12 pts]
14. Take circles centered at A and O as in problem 13, with O lying on circle A. Take the associated conic
section, again as in problem 13. Consider the feet of the perpendiculars from O to the tangent lines of
this conic section; find and justify the locus of these feet. [12 pts]
 2011 SMT POWER ROUND: POLES AND POLARS 3

15. Again set your diagram up as in problem 13 (this time, make no assumptions about the placement of
O relative to circle A). Suppose point P lies on the associated conic section (i.e. it is the pole of some
tangent to circle A with respect to circle O). Let a be the polar of A with respect to circle O. Prove that
the distance OP is equal to  times the distance from P to a, where  is a constant that does not depend
on P . (Hint: it may help to consider the point M , defined as the intersection between OA and the polar
of P .) [14 pts]

5. Counting
Taking the dual of certain counting problems can make the counting easier. Apply the idea of duality
to help solve the following problem.
16. For any n > 0, take n distinguishable points P1 , ..., Pn in the plane. For any line ` not going through
any of the points, let the “linear partition of P1 , ..., Pn by `”, which we denote M` (P1 , ..., Pn ), be the
unordered pair {L, R} where L is the set of points on one side of the line and R is the set of points on
the other side of the line. Let C(P1 , ..., Pn ) be the number of distinct unordered pairs M` (P1 , ..., Pn ) as `
ranges over all lines. i.e., C(P1 , ..., Pn ) is the number of different ways we can split P1 , ..., Pn with lines.
Here are some examples. For one point, C = 1 because there is only one way to “split” the set of one
point. For two points, C = 2 because we can have both points on one side of the line or each point on
an opposite side of the line. For three collinear points P1 , P2 , P3 , C = 3 because the linear partitions are
| P1 , P2 , P3 and P1 | P2 , P3 and P1 , P2 | P3 . For three points on the vertices of an equilateral triangle,
C = 4 because we can have all the points on the same side of the line or we can split off any one of the
three points from the other two.
What is the maximum of C(P1 , ..., Pn ) over all choices of n points? Prove that your answer is the
maximum. Duality will be helpful. [22 pts]
 SMT Power Round Solutions : Poles and Polars

February 18, 2011

1 Definition and Basic Properties

1
Note that the unit circles are not necessary in the solutions. They just make the graphs look nicer.

(1)
2.0

1.5

1.0

0.5

-1.0 -0.5 0.5 1.0 1.5 2.0

-0.5

-1.0

1
(2)
3

- 2.0 -1.5 -1.0 -0.5 0.5 1.0

-1

2
(3)
2.0

1.5

1.0

0.5

- 2.0 -1.5 -1.0 -0.5 0.5 1.0

-0.5

-1.0

(4)
1.0

0.5

-1.5 -1.0 -0.5 0.5 1.0 1.5 2.0

-0.5

-1.0

-1.5

(5)
This is the same as in (1).

(6)
It is interesting to note that this line goes through the A and the pole of b.

3
 1.5

1.0

0.5

-2 -1 1 2

-0.5

-1.0

2
The perpendicular through the origin to the polar of P is −py x + px y = 0. This line intersects P at P ∗ = |PP|2 . In
particular, |P ∗ | = |P1 | . So P ∗ is on the other side of the unit circle from P . Since P ∗ is the closest point on the
polar to the origin, the polar intersects the unit circle once when P ∗ is on the unit circle, twice when P ∗ is inside
the unit circle, and zero times when P ∗ is outside the circle. In other words, the polar intersects the unit circle
once when P is on the unit circle, twice when P is outside the unit circle, and zero times when P is inside the unit
circle.

3
Let P 0 = P
|P |2 . This is the inversion of P because it is clearly on the ray OP , and also because it satisfies
0
OP · OP 0 = OP|P·OP
|2 = 1.
p
P is on the polar of P because px ( p2p+p
0 x y
2 ) + py ( p2 +p2 ) = 1. Finally, the polar is perpendicular to OP because
x y x y

OP is pointing in the same direction as the normal (px , py ) to the polar.

4
Let O = (a, b) and apply the definition from problem 3 except replace OP · OP 0 = 1 with OP · OP 0 = r2 . In
particular, let’s find the point P 0 = (p0x , p0y ) on the ray OP such that OP · OP 0 = r2 . To do this, parametrize the
ray OP as R(t) where

R(t) = (a + (px − a)t, b + (py − b)t) ,

for t ≥ 0. Then notice that
OP · OR(t) = px (px − a)t + py (py − b)t
2
r
Setting this to r2 and solving for t gives t = p2 +p2 −ap x −bpy
. Plugging this into our parametrization,
x y

r2 (px − a)r2 (py − b)r2

   
0
P =R = a+ ,b + .
p2x + p2y − apx − bpy |OP |2 |OP |2
Finally, note that the perpendicular to OP through P 0 = (p0x , p0y ) is given by

(px − a)(x − p0x ) + (py − b)(y − p0y ) = 0.

Plugging in our coordinates for P 0 gives the answer, which is

(px − a)x + (py − b)y = (px − a)a + (py − b)b + r2 .

4
2 The Duality Principle
5
Suppose A = (xA , yA ) is on the polar of B = (xB , yB ). The polar of A is xA x + yA y = 1. Since B is on this line,
xA xB + yA yB = 1. This immediately implies that A is on the polar xB x + yB y = 1 of B.
The other direction is completely symmetric.
Alternatively, use the definition from problem 3. Let A0 and B 0 be the inversions of A and B respectively, and
suppose B is on the polar of A. Consider the point B 00 which is the foot of the perpendicular from A to OB. Since
6 AA0 B = 6 AB 00 B = 90◦ , quadrilateral AA0 B 00 B is cyclic, and by Power of a Point on its circumscribed circle and
point O, we have (OB 00 )(OB) = (OA0 )(OA) = r2 where r is the radius of circle O. But then OB 00 = r2 /OB = OB 0 ,
so in fact B 0 = B 00 . Therefore A lies on the perpendicular to OB through B 0 , otherwise known as the polar of B.

6
Let A = (xA , yA ), B = (xB , yB ), and C = (xC , yC ). Then a is xA x + yA y = 1, b is xB x + yB y = 1, and c is
xC x + yC y = 1.

a
This is just a restatement of 5, in our new notation.

b
The intersection of a and b is a point that is on both polars a and b. So by part (a), both A and B lie on the polar
of the intersection of a and b. In other words, the polar of the intersection of a and b is the line AB.
The converse follows immediately from the fact that reciprocation is an involution. In particular, apply re-
ciprocation to the statement “the polar of the intersection of a and b is the line AB” to get the statement “the
intersection of a and b is the polar of AB.”

c
Suppose a, b, c go through the same point. Then, by part (a), the polar of this point goes through A, B, C. In
particular, A, B, C are collinear.
Suppose A, B, C are collinear. Then the pole of the line through A, B, C is on a, b, c. In particular, a, b, c all go
through the same point.

7
Given a, b, c concurrent and d, e, f concurrent, the three lines x = (b∩f )(c∩e), y = (a∩f )(c∩d), and z = (a∩e)(b∩d)
are also concurrent.

8
Take the dual to be the dual around the relevant circle, as described in our generalization in (5). Then incidence
of a point with a circle corresponds with tangency of its polar with the circle.
Let abcdef be sides of a cyclic dual hexagon (not necessarily in that order). Extend them to lines abcedf . Then
the three lines (a ∩ b)(d ∩ e), (b ∩ c)(e ∩ f ), and (c ∩ d)(f ∩ a) are concurrent.

3 Reciprocation and Cyclic Quadrilaterals

9
1
 
n m n m
X= xa + xb , ya + yb
m+n m+n m+n m+n

5
  
−n m −n m
Q= xa + xb , ya + yb
m−n m−n m−n m−n

2
Just plug in and check

     
n m −n m n m −n m
xa + xb xa + xb + ya + yb ya + yb
m+n m+n m−n m−n m+n m+n m−n m−n

−n2 2 2 m2 2 2 −n2 m2
= (x + y ) + (x + y ) = + =1
m2 − n2 a a
m2 − n2 b b
m2 − n2 m2 − n2

3
Apply Menelaus’ theorem to the triangle XY Q and the collinear points AP1 C, giving
XP1 Y C QA XP1 CQ AX
= −1 =⇒ =−
P1 Y CQ AX P1 Y Y C QA
Apply Menelaus’ theorem to the triangle XY Q and the collinear points BP1 D, giving
XP2 Y B QD XP2 BQ DX
= −1 =⇒ =−
P2 Y BQ DX P2 Y Y B QD
AX AQ BQ
The two ratios we just computed are the same because of the equations XD = QD and BY
Y C = QC defining X and
Y respectively.
Therefore P1 = P2 . And since P1 = P2 is both on AC and BD, and P is defined as the intersection of AC and
BD, P = P1 = P2 .

4
X and Y are both on the polar of Q by part (2). So the line XY is the polar of Q. By part (3), P is on XY . So
P is on the polar of Q.

10
What we proved in part (9) is that if we have four points A, B, C, D on a circle, then AD ∩ BC is on the polar of
AC ∩ BD. Just permute the points in this statement around to see that, in fact, each pair of points in P, Q, R is
on the polar of the other point.
The orthocenter of P QR is the center of the circle. In particular, if we are using the unit circle around the
origin, then the orthocenter of P QR is the origin. This is true because each side of the triangle is the polar of the
opposite vertex. So the perpendicular through each side through the opposite vertex goes through the center of
the circle. Ie, all three altitudes intersect at the center of the circle.

11
We already know that (1) and (2) are on the polar of AB ∩ CD, so let’s check the remaining points.
Point (3). Observe the following diagram. We want to show that R, the intersection of the tangent at A and
the tangent at B is on the polar of AB ∩ CD. Equivalently, we need to show that AB ∩ CD is on the polar of R.
To show this, let’s show that AB is the polar of R.
First, note that AB is perpendicular to OR by symmetry. So AB is parallel to the polar of R. To see that AB
is in fact equal to the polar of R, we need only show that OI · OR = 1. One easy way to see this is to note that
|OI| = cos θ and |OR| = cos1 θ .

6
 Point (4). Same as point (3).
Point (5). Call point (5) X. We want to show that X is on the polar of AB ∩ CD, which is equivalent to
showing that AB ∩ CD is on the polar of X. Since X is on the circle, the polar of X is the tangent to the circle
at X. By the definition of X, AB ∩ CD is on this tangent. So we are done.
Point (6). Same as point (5).

12
The line through P and the other intersection of the circumcircles is the radical axis of the two circumcircles. So
it suffices to prove that another of the “six points” also lies on the radical axis. But by Power of a Point on the
circle in which ABCD is inscribed, AD ∩ BC clearly has the same power with respect to both circles. Therefore
the second intersection of the circumcircles indeed lies on the line.

4 Conic Sections
13
(1) It’s an ellipse. Every tangent to circle A is at distance at least r − OA from O, and thus every pole of such a
tangent is at distance at most 1/(r − OA) from O—that is, the locus of these poles is bounded. Therefore it must
be an ellipse.
(2) The polars of the points of circle A are tangents to the ellipse by duality.
(3) The conic section is a parabola, since there is precisely one tangent line (the one at O) which has no finite
pole, so that the conic should go to infinity along AO and be continuous elsewhere.
(4) The conic section is a hyperbola. There are now two tangents to circle A passing through O, neither of
which have a pole. Their points of tangency to A have polars which are tangent to the conic but never touch it. A
conic section with two asymptotes is a hyperbola.

14
As previously explained, each tangent to the parabola is the polar of a point T on circle A. The foot of the
perpendicular from O to the tangent is the inversion of T with respect to circle O. So the locus is the inversion of
circle A about circle O, which is a line. Specifically, it’s the line going through the intersection of the circles.

7
15
Let P be a point on the conic and let T M be its polar where T is the point of tangency to circle A. The inversion
P 0 of P lies on T M . Since M lies on the polar of P , P lies on the polar of M , so that the foot of the perpendicular
from P to line OA is the inversion M 0 of M . Let A0 be the inversion of A and K be the foot of the perpendicular
from P to a. We need to prove that OP = P K. We will use directed distances to avoid casework. We calculate

OA0 − OM 0
 
PK 0 1 1
= = OP −
OP OP OA OM

OP 0 OM
 
AT AM r
= −1 = = .
OM OA AM OA OA
r
We set  = OA . (Note: this section was mostly taken from Geometry Revisited.)

5 Counting
16
First we prove by induction the nice fact that n lines split the plane into a maximum of n(n+1) 2 + 1 regions.
For the base case, notice that 1 line divides the plane into 2 = 1 + 1 regions.
For the inductive step, assume that n−1 lines divide the plane into a maximum of n(n−1) 2 +1 regions. Add an n-
th line that intersects all n−1 lines in points where they are not intersecting each other. This new line splits each of
the n regions it goes through into 2 regions. Ie, this new line adds n regions. So we have n(n−1) 2 + 1 + n = n(n+1)
2 +1
regions.
We still need to show that we can’t get more than n(n+1) 2 regions with n lines. If we do get more regions, then
n(n−1)
either (1) we started with more than 2 + 1 regions before we added the n-th line or (2) we added more than
n regions when we added the n-th line. But (1) contradicts the inductive hypothesis that n − 1 lines give us a
maximum of n(n−1) 2 + 1 regions. And if (2) is the case, then our n-th line split more than n regions, forcing it to
have intersected more than n − 1 old lines. But there are only n − 1 old lines to intersect. So we are done.
Next, we claim that the maximum number of distinct linear partitions is n(n−1) 2 + 1 and we use our lemma to
prove this.
Take any set of n points P1 , ..., Pn and translate them so that Pn is at the origin. Let L1 , ..., Ln−1 be the dual
lines of P1 , ..., Pn−1 and let R1 , ..., Rs be the regions bounded by these lines. Now we will show that there is a
one-to-one correspondence between the regions R1 , ..., Rs and the linear partitions of P1 , ..., Pn .
To get a linear partition from a region Ri , take a point X ∈ Ri . The dual of X is some line ` that gives us a
linear partition of P1 , ..., Pn . This partition is independent of our choice of X ∈ Ri because as we move X around in
Ri we don’t touch any of the dual lines L1 , ..., Ln−1 and therefore ` never goes through any of the points P1 , ..., Pn .
Furthermore, we get every partition because every partition induced by a line ` is achieved when we take X to be
the dual point to `. Finally, each partition we get is different because when we choose a region R on one side of a
line Lj we get Pj being in the same side of the partition as the origin and when we choose a region R on the other
side of the line Lj we get Pj being on the other side of the partition as the origin.
So there are the same number of linear partitions of P1 , ..., Pn as there are regions created by the n − 1 dual
lines to P1 , ..., Pn−1 . By our lemma, we can arrange for this to be n(n−1) 2 + 1 but we can’t have it be any bigger.
n(n−1)
So 2 is the maximum.

8
SMT 2012 Algebra Test and Solutions February 18, 2012

1. Compute the minimum possible value of

(x − 1)2 + (x − 2)2 + (x − 3)2 + (x − 4)2 + (x − 5)2

for real values of x.

Answer: 10
Solution: We know that this expression has to be a concave-up parabola (i.e. a parabola that
faces upwards), and there is symmetry across the line x = 3. Hence, we conclude that the vertex
of the parabola occurs at x = 3. Plugging in, we get 4 + 1 + 0 + 1 + 4 = 10 .
2 −6x+5
2. Find all real values of x such that ( 51 (x2 − 10x + 26))x = 1.
Answer: 1, 3, 5, 7
Solution: Clearly, the above equation holds if if 51 (x2 − 10x + 26) = 1 or x2 − 6x + 5 = 0, from
which we obtain 3, 7 and 1, 5, respectively. To see that these are the only possible values for
x, note that x2 − 10x + 26 = (x − 5)2 + 1 is always positive. Since for positive a, the function
f (y) = ay is strictly increasing, the only solution to ay = 1 is y = 0.
23 −1 3 43 −1 163 −1
3. Express
23 +1
× 333 −1
+1
× 43 +1
× ··· × 163 +1
as a fraction in lowest terms.
91
Answer: 136
Solution: We note
k k k k
! !
Y n3 − 1 Y (n − 1)(n2 + n + 1) Y n−1 Y n2 + n + 1
= = .
n3 + 1 (n + 1)(n2 − n + 1) n+1 n2 − n + 1
n=2 n=2 n=2 n=2

1·2 k2 +k+1 91
Each product telescopes, yielding k·(k+1) · 3 . Evaluating at k = 16 yields .
136
4. If x, y, and z are integers satisfying xyz + 4(x + y + z) = 2(xy + xz + yz) + 7, list all possibilities
for the ordered triple (x, y, z).
Answer: (1, 1, 1), (1, 3, 3), (3, 1, 3), (3, 3, 1)
Solution: Rearranging the given equality yields xyz − 2(xy + xz + yz) + 4(x + y + z) − 8 = −1.
But the left side factors as (x − 2)(y − 2)(z − 2). Since all quantities involved are integral, we
must have each factor equal to ±1. It is easy to verify that the only possibilities for (x, y, z) are
those listed.

5. The quartic (4th-degree) polynomial P (x) satisfies P (1) = 0 and attains its maximum value of
3 at both x = 2 and x = 3. Compute P (5).
Answer: −24
Solution: Consider the polynomial Q(x) = P (x) − 3. Q has roots at x = 2 and x = 3.
Moreover, since these roots are maxima, they both have multiplicity 2. Hence, Q is of the form
a(x − 2)2 (x − 3)2 , and so P (x) = a(x − 2)2 (x − 3)2 + 3. P (1) = 0 =⇒ a = − 34 , allowing us to
compute P (5) = − 43 (9)(4) + 3 = −24 .

6. There exist two triples of real numbers (a, b, c) such that a − 1b , b − 1c , and c − a1 are the roots
to the cubic equation x3 − 5x2 − 15x + 3 listed in increasing order. Denote those (a1 , b1 , c1 ) and
(a2 , b2 , c2 ). If a1 , b1 , and c1 are the roots to monic cubic polynomial f and a2 , b2 , and c2 are the
roots to monic cubic polynomial g, find f (0)3 + g(0)3 .
SMT 2012 Algebra Test and Solutions February 18, 2012

Answer: −14
Solution: By Viéta’s Formulas, we have that f (0) = −a1 b1 c1 and g(0) = −a2 b2 c2 . Additionally,
(a− 1b )(b− 1c )(c− a1 ) = −3 and (a− 1b )+(b− 1c )+(c− a1 ) = 5. Expanding the first expression yields
1
−3 = abc− abc −((a+b+c)−( a1 + 1b + 1c )) = abc− abc
1
−5. This is equivalent to (abc)2 −2(abc)−1 = 0,
√ √ √
so abc = 1 ± 2. It follows that f (0)3 + g(0)3 = −(1 + 2)3 − (1 − 2)3 = −14 .

7. The function f (x) is known to be of the form ni=1 fi (ai x), where ai is a real number and fi (x)
Q
is either sin(x) or cos(x) for i = 1, . . . , n. Additionally, f (x) is known to have zeros at every
integer between 1 and 2012 (inclusive) except for one integer b. Find the sum of all possible
values of b.
Answer: 2047
Solution: The possible values of b are the powers of two not exceeding 2012 (including 20 = 1).
The following proof uses the fact that the zeroes of sine and cosine are precisely numbers of the
form tπ and (t + 1/2)π, respectively, for t an integer.
Suppose b is not a power of 2. Then it can be written as 2m (1 + 2k) for m ≥ 0, k > 0. Since
2m < b, by assumption one of the fi must have a root at 2m . But then this same fi must have
a root at b:
• If fi (x) = sin(ax) and fi (2m ) = 0, then 2m a = tπ for some integer t, so

fi (b) = sin(ba) = sin((1 + 2k)2m a) = sin((1 + 2k)tπ) = 0.

• If fi (x) = cos(ax) and fi (2m ) = 0, then 2m a = (t + 1/2)π for some integer t so

fi (b) = cos(ba) = cos((1 + 2k)2m a) = cos((1 + 2k)(t + 1/2)π) = cos((t + k + 2kt + 1/2)π) = 0

This is a contradiction, so b must be a power of 2.

For each b of the form 2m , we can construct an f that works by using cosine terms to cover
integers preceding b and sine terms thereafter:

m
!  2012 
Y Y
f (x) = cos(πx/2i )  sin(πx/j)
i=1 j=b+1

has a root at every positive integer at most 2012 except b.

Hence, our final answer is 1 + 2 + 4 + ... + 1024 = 2048 − 1 = 2047 .

8. For real numbers (x, y, z) satisfying the following equations, find all possible values of x + y + z.

x2 y + y 2 z + z 2 x = −1
xy 2 + yz 2 + zx2 = 5
xyz = −2
q
3 1
Answer: 2 or 2

Solution 1: Let x/y = a, y/z = b, and z/x = c. Then abc = 1. By dividing the first two
equations by the third equation, we have a+b+c = −5/2 and 1/a+1/b+1/c = ab+bc+ca = 1/2.
So a, b, c are roots of 2X 3 + 5X 2 + X − 2 = 0. By observation, the three roots of this equation
 SMT 2012 Algebra Test and Solutions February 18, 2012

are −2, −1, and 1/2. Without loss of generality, assume that a = −1 and y = −x. If c = −2,
then we have z = −2x, so xyz = 2x3 = −2, and thus x = −1. In this case we have

(x, y, z) = (−1, 1, 2).

√
If c = 1/2, then z = x/2, so xyz = −x3 /2 = −2, or x = 3 4. In this case we have
r !
√
3
√
3 3 1
(x, y, z) = 4, − 4, .
2
r
3 1
It follows that x + y + z can be either 2 or .
2
Solution 2: We have (x + y)(y + z)(z + x) = x2 y + y 2 z + z 2 x + xy 2 + yz 2 + zx2 + 2xyz = −1 + 5 +
2(−2) = 0. Hence one of the quantities x+y, y+z, z+x is 0. Suppose WLOG that x = −y, so that
y 2 z = 2 and y 3 + y 2 z − yz 2 = −1. The latter equation can be rewritten as 2y/z + 2 − 2z/y = −1,
or, setting a = y/z, 2a + 3 − 2/a = 0, giving the quadratic 2a2 + 3a − 2 = (2a − 1)(a + 2) = 0.
Then y = z/2 or y = −2z, giving rise respectively to z 3 = 8 and z 3 = 1/2. Since x + y = 0, we
r
3 1
have x + y + z = z = 2 or .
2

9. Find the minimum value of xy, given that x2 + y 2 + z 2 = 7, xy + xz + yz = 4, and x, y, z are

real numbers.
1
Answer: 4
Solution 1: Note that 4xy = (x + y − z)2 + 2(xy + xz + yz) − (x2 + y 2 + z 2 ). Since (x + y − z)2
is non-negative, it follows that 4xy is at least 2 · 4 − 7 = 1, so xy is at least 1/4.
We now aim to find a solution for which xy = 1/4 (thereby proving our lower bound to be
tight). We have seen that xy = 1/4 implies that x + y − z = 0, or z = x + y. Substituting into
xy + xz + yz = 4, we obtain x2 + 3xy + y 2 = 4. Subtracting xy = 1/4 and 5xy = 5/4 √
from this
2 2 15
equation yields (x + y) = 15/4 and (x − y) = 11/4. Thus, one solution is x + y = 2 , x − y =
√ √ √ √√ √
11 15+ 11 15− 11 15
2 , or x = 4 ,y = 4 ,z = 2 .
Solution 2: We have (x + y + z)2 = x2 + y 2
+ z 2 + 2xy + 2xz + 2yz = 7 + 2 · 4 = 15. Note that
minimizing xy is equivalent to maximizing xz + yz = (x + y)z. Also note that this maximum
value of (x + y)z is certainly positive, and in fact we can assume x + y and z are both positive,
since otherwise we can replace x, y, z with −x, −y, −z without changing any of our quantities.
2
We can then apply AM-GM to find that (x + y)z ≤ x+y+z , or (x + y)z ≤ (x+y+z) = 15
p
2 4 4 .
1
Hence xy ≥ 4 − 154 = 4 and we proceed as in the first solution.

10. Let X1 , X2 , . . . , X2012 be chosen independently and uniformly at random from the interval (0, 1].
In other words, for each Xn , the probability that it is in the interval (a, b] is b − a. Compute
the probability that dlog2 X1 e + dlog4 X2 e + · · · + dlog4024 X2012 e is even. (Note: For any real
number a, dae is defined as the smallest integer not less than a.)
2013
Answer: 4025
Solution: To simplify notation, define Yn = dlog2n Xn e.
We begin by computing the probability that Yn is odd. Yn = −1 if −2 < log2n Xn ≤ −1, or
1 1 1 1
(2n)2
< Xn ≤ 2n . Similarly, Yn = −3 if (2n) 4 < Xn ≤ (2n)3 , and so on. Adding up the lengths
SMT 2012 Algebra Test and Solutions February 18, 2012

P∞ 1 1
of these intervals, we see that the probability that Yn is odd is k=1 (2n)2k−1 − (2n)2k
. This is a
1 1 1
geometric series with first term 2n (1 − 2n ) and ratio (2n)2
, so its sum is

1 1 1
2n (1 − 2n ) 2n 1
1 = 1 = .
1− (2n)2
(1 + 2n )
2n + 1

Armed with this fact, we are now ready to solve the problem. One way to continue would be to
note that the probability that Y1 is even is 2/3, the probability that Y1 + Y2 is even is 3/5, the
probability that Y1 + Y2 + Y3 is even is 4/7 and to show by induction that the probability that
n+1
Y1 + · · · + Yn is even is 2n+1 . Below, we present an alternate approach.
Note that Y1 +Y2 +· · ·+Y2012 is even if and only if (−1)Y1 +Y2 +···+Y2012 = 1. Rewrite (−1)Y1 +Y2 +···+Y2012
as (−1)Y1 (−1)Y2 · · · (−1)Y2012 , and note that because the Yn are independent,

E (−1)Y1 (−1)Y2 · · · (−1)Y2012 = E[(−1)Y1 ]E[(−1)Y2 ] · · · E[(−1)Y2012 ],

 
(1)

where E denotes the expected value of the quantity. But E[Yn ] = (+1) · P (Yn is even) + (−1) ·
1
P (Yn is odd). We computed earlier that the probability that Yn is odd is 2n+1 , so E[Yn ] = 2n−1
2n+1
1 3 4023 1
and product in (1) is 3 · 5 · · · 4025 , which telescopes to yield 4025 . Let p be the probability that
1
Y1 + Y2 + · · · + Y2012 is even. We just found that (+1)(p) + (−1)(1 − p) = 4025 , which we can
2013
solve to obtain p = .
4025
SMT 2012 Algebra Tiebreaker and Solutions February 18, 2012

1. The quadratic x2 − 4x + 2 has two distinct roots, a and b. Compute the sum of the coefficients
of the monic quadratic with roots a2 + b2 and a3 + b3 .
Answer: 429
Solution: We know that a + b = 4 and ab = 2 from Vieta’s. a2 + b2 = (a + b)2 − 2ab = 12,
while a3 + b3 = (a + b)3 − 3ab(a + b) = 40. Therefore, our quadratic is (x − 12)(x − 40), and the
sum of the coefficients of the quadratic can be obtained by evaluating the quadratic at x = 1,
which gives us 429 .

2. Find the minimum value of xy + xz + yz given that x, y, z are real and x2 + y 2 + z 2 = 1.

Answer: − 21
Solution: We note that (x + y + z)2 = (x2 + y 2 + z 2 ) + 2(xy + xz + yz) = 1 + 2(xy + xz + yz).
Since (x + y + z)2 clearly has a minimum value of 0, it follows that xy + xz + yz is at least
√
−1/2 . This occurs, for example, when x = −y = 1/ 2 and z = 0.
√ 1
3. Find all solutions α with 0◦ < α < 90◦ to the equation 1 + 3 tan(60◦ − α) = .
sin α
Answer: 30◦ , 50◦
√
Solution: Multiply both sides by cos(60◦ − α) and write 1 as 2 sin(30◦ ) and 3 as 2 cos(30◦ )
to get
cos(60◦ − α)
2 (sin(30◦ ) cos(60◦ − α) + cos(30◦ ) sin(60◦ − α)) = .
sin(α)
Applying the formula for the sine of a sum, we see that the left-hand-side is 2 · sin(30◦ + 60◦ − α),
or 2 cos(α). Multiplying both sides by sin(α) yields 2 sin(α) cos(α) = cos(60◦ − α), which can be
rewritten
sin(2α) = sin(α + 30◦ ). (1)

Therefore, either 2α = α + 30◦ or 2α + (α + 30◦ ) = 180◦ , from which we obtain α = 30◦ and
α = 50◦ , respectively.
SMT 2012 Calculus Test and Solutions February 18, 2012

R 10
1. What is
0 (x − 5) + (x − 5)2 + (x − 5)3 dx?
Answer: 250
3
Solution: This integral is equal to
5 5
53 (−5)3
Z Z  
2 3 2 250
x + x + x dx = x dx = − = .
−5 −5 3 3 3

2. Find the maximum value of Z 3π/2

sin(x)f (x) dx
−π/2

subject to the constraint |f (x)| ≤ 5.

Answer: 20
Solution: Clearly we want to maximize f (x) when sin (x) ≥ 0 and minimize f (x) when sin (x) <
0. We do this by setting f (x) = 5 in the first case and f (x) = −5 in the second case. Noting
that the bounds of integration cover precisely one full period of sin, we see that the integral
becomes equivalent to twice the integral of 5 sin (x) over the half period where sin (x) ≥ 0. This
results in 20 .

3. Calculate Z 35
1
dx.
25 x − x3/5
5 8
Answer: 2
ln 3
Solution: Note that we can write the integral as
Z 35
1
dx.
25 x3/5 (x2/5 − 1)

We solve via u-substitution. Let u = x2/5 − 1:

2 5
du = x−3/5 dx =⇒ dx = x3/5 du.
5 2
The integral becomes
32 −1 8
x3/5
Z Z
5 5 1
3/5
du = du,
2 22 −1 x ·u 2 3 u
which evaluates to
5 5 8
(ln 8 − ln 3) = ln .
2 2 3
√
4. Compute the x-coordinate of the point on the curve y = x that is closest to the point (2, 1).
√
2+ 3
Answer: 2
Solution: We want to minimize the distance between the points (a2 , a) and (2, 1). We can
equivalently minimize the square of the distance between those two points, which is

(2 − a2 )2 + (1 − a)2 = a4 − 3a2 − 2a + 5.
SMT 2012 Calculus Test and Solutions February 18, 2012

The derivative of this function is 4a3 −6a−2, which

√ can be factored as 2(a+1)(2a2 −2a−1). The
1± 3
roots of this cubic are therefore a = −1, . Two of the roots are negative and therefore
√ √ 2
1+ 3 2+ 3
invalid, so a = and a2 = .
2 2
5. Let
x2 x3 x4 x5
f (x) = x + + + + ,
2 3 4 5
and set g(x) = f −1 (x). Compute g (3) (0).
Answer: 1
Solution 1: The inverse function rule tells us that
−1
g 0 (x) = f 0 (g(x))

.

Using this and the fact that g(0) is clearly equal to zero, this problem can be solved with a
straightforward bash.
Solution 2: We begin by observing that we know by the inverse function rule that we will not
need to know any derivatives of f past its third derivative. Since the first three derivatives of
f at zero agree with those of the function − log (1 − x) (by Taylor series expansion), we can
assume f (x) = − log (1 − x). Now,

y = − log (1 − x) =⇒ −y = log (1 − x) =⇒ e−y = 1 − x =⇒ x = 1 − e−y

so g(y) = 1 − e−y and therefore g (3) (0) = e−0 = 1 .

6. Compute
  1
sin x 1−cos x
lim .
x→0 x

Answer: e−1/3
Solution 1: We take logs and evaluate by L’Hopital’s rule:
"  1 # cos x 1
sin x 1−cos x log(sin x) − log x − x cos x − sin x
lim log = lim = lim sin x x
= lim
x→0 x x→0 1 − cos x x→0 sin x x→0 x sin2 x
−x sin x −x −1 1
= lim 2 = lim = lim =− .
x→0 sin x + 2x sin x cos x x→0 sin x + 2x cos x x→0 cos x + 2 cos x − 2x sin x 3

Therefore, the answer is e−1/3 .

Solution 2: We can approximate sin x and cos x by their Taylor series. Applying substitutions
then yields
1 2 − 22
x2 x2 x2
   
sin x 1−cos x x
lim = lim 1 − = lim 1 +
x→0 x x→0 6 x→0 6
−2x2
1 −x/3
  
1
= lim 1 + 2 = lim 1 + = e−1/3
x→∞ 6x x→∞ x

because e = limx→∞ (1 + x1 )x .
SMT 2012 Calculus Test and Solutions February 18, 2012

7. A differentiable function g satisfies

Z x
(x − t + 1)g(t) dt = x4 + x2
0

for all x ≥ 0. Find g(x).

Answer: 12x2 − 24x + 26 − 26e−x
Solution: First differentiate the equation with respect to x:
Z x
g(x) + g(t) dt = 4x3 + 2x.
0

Differentiate again to obtain

g 0 (x) + g(x) = 12x2 + 2.
The solution 12x2 − 24x + 26 can be found using the method of undetermined coefficients, so
the general solution will be
g(x) = 12x2 − 24x + 26 + Ce−x
for some constant C. By substituting x = 0 into the first equation, we see that g(0) = 0. We
therefore find that C = −26, making the answer 12x2 − 24x + 26 − 26e−x .

8. Compute Z ∞
ln x
dx.
0 x2 +4
π ln 2
Answer: 4
Solution: Substitute x = 2 tan θ to get
Z ∞
1 π/2 1 π/2
Z Z
ln x 1 π
dx = ln(2 tan θ) dθ = · ln 2 + ln(tan θ) dθ.
0 x2 + 4 2 0 2 2 2 0
We will now show that this final integral is zero by substituting u = π/2 − θ to yield
Z π/2 Z π/2  π 
ln(tan θ) dθ = ln tan −θ dθ
0 0 2
Z π/2   Z π/2
1
= ln dθ = − ln(tan θ) dθ,
0 tan θ 0

π ln 2
which gives us what we wanted, so the answer is therefore .
4
√
n2
1!2! · · · n!
9. Find the ordered pair (α, β) with non-infinite β 6= 0 such that lim = β holds.
n→∞ nα
Answer: (1/2, e−3/4 )
Solution 1: Taking the logarithm of 1!2! · · · n!, we find

ln(1!2! · · · n!) = ln 1 + (ln 1 + ln 2) + (ln 1 + ln 2 + ln 3) + · · ·

= n ln 1 + (n − 1) ln 2 + · · · + ln n
1 2 n n(n + 1)
= n ln + (n − 1) ln + · · · + ln + ln n.
n n n 2
SMT 2012 Calculus Test and Solutions February 18, 2012

Dividing this by n2 , we have

n
!
√
n2 n+1 1 X n+1−m m
ln( 1!2! · · · n!) = ln n + ln
2n n n n
m=1

As n goes to infinity, the sum will converge to the integral

Z 1  1
1 2 1 2 3
(1 − x) ln x dx = x − x ln x − x + x ln x = −
0 4 2 0 4

and the first term will approach 12 ln n, so if we subtract 12 ln n then this expression will converge
to − 43 . Finally, by raising e to the power of each side, we have
√
n2
1!2! · · · n!
lim = e−3/4 .
n→∞ n1/2

Solution 2: We take logs and approximate by Stirling’s approximation1 . Here, we will be some-
what sloppy with our approximations, and freely use ∼ to indicate approximations that become
irrelevant in the limit. All of these calculations can be made rigorous, and such justification is
left as an exercise to the reader.
Stirling’s approximation says that log(n!) ∼ n log n − n. Using this, we have
√
n2
! n
1!2! · · · n! 1 1 X
log = log(1!2! · · · n!) − α log n = log(k!) − α log n
nα n2 n2
k=1
n n
1 X 1 X 1 n(n + 1)
∼ 2 (k log k − k) − α log n = 2 (k log k) − 2 − α log n
n n n 2
k=1 k=1

n(n+1) n2
Approximate 2 ∼ 2 and approximate the infinite sum by an integral (and integrate by
parts):
Z n
1 n2 n2 1
 
1 1 1
∼ 2 x log x dx − − α log n = 2 log n − + − − α log n
n 1 2 n 2 4 4 2
1 3
∼ log n − − α log n.
2 4
Therefore,
√
n2
!   
1!2! · · · n! 1 3
lim log = lim − α log n − ,
n→∞ nα n→∞ 2 4

which is finite only when 12 − α = 0, in which case α = 12 and the limit evaluates to − 34 .
Therefore, we wish to compute
√
n2
" √
n2
!#
1!2! · · · n! 1!2! · · · n!
lim = exp lim log = e−3/4 .
n→∞ n1/2 n→∞ nα
1
http://en.wikipedia.org/wiki/Stirling’s approximation
 SMT 2012 Calculus Test and Solutions February 18, 2012

10. Find the maximum of Z 1

f (x)3 dx
0
given the constraints Z 1
−1 ≤ f (x) ≤ 1, f (x) dx = 0.
0

Answer: 1/4
Solution 1: Consider the expression
Z 1  2
1
(f (x) − 1) f (x) + dx.
0 2

Since f (x) ≤ 1 this expression is less than or equal to 0. Meanwhile, expanding the integrand
gives
1 2
 
3 1
(f (x) − 1) f (x) + = f (x)3 − f (x) − ,
2 4 4
so its integral is
1 2 1
3 1 1 1
Z  Z Z Z
1
(f (x) − 1) f (x) − dx = f (x)3 dx −
f (x) dx − dx
0 2 0 4 0 4 0
Z 1
1
= f (x)3 dx − ,
0 4

proving that the answer is at most 1/4. This expression is zero when f (x) = 1 or f (x) = −1/2,
so 1/4 is indeed the maximum value.
Solution 2: Let f+ and f− denote the positive and negative part of f respectively. Define
A+ = {f > 0} and A− = {f < 0}. Then f+ and f− are nonzero only on A+ and A− . Also the
condition on f means
Z 1 Z 1
0 ≤ f+ (x), f− (x) ≤ 1, f+ (x) dx = f− (x) dx = s.
0 0

We will fix s and try to optimize the value f 3 = f+3 − f−3 . For the maximum of f+3 , we
R R R R

have the inequality f+ (x)3 ≤ f+ (x), and integrating it gives

Z 1 Z 1
3
f+ (x) dx ≤ f+ (x) dx = s.
0 0

f−3 , we claim that

R
Equality holds when f+ (x) is either 0 or 1 for all x. For the minimum of
the minimum occurs when f− is constant on A− . From the inequality
3
a3 + b3

a+b
≥
2 2

for 1 > a, b > 0, we speculate that the minimum is achieved when the values of f− are distributed
as close as possible. Then if we denote the length of A− to be l, the minimum occurs when
SMT 2012 Calculus Test and Solutions February 18, 2012

f− (x) = s/l for all x ∈ A− , and the integral will be l · (s/l)3 = s3 /l2 . This can be made rigorous
by applying Jensen’s inequality
R R !
A− φ(f− (x)) dx A− f− (x) dx
≥φ
l l

for the convex function φ(t) = t3 . This gives the minimum as

f− (x)3 dx s3 s3
R Z
≥ 3, f− (x)3 dx ≥ 2 .
l l l
Meanwhile A+ should have length greater than s, since f+ is only nonzero on A+ and the integral
of f+ ≤ 1 over A+ has length at most A+ . So we have 1 − l ≥ s, l ≤ 1 − s. Now our objective
integral is optimized to be the maximum of a single-variable function in s, as follows:
Z 1
s3
f (x)3 dx ≤ s − .
0 (1 − s)2

We differentiate this in s to find the maximum. Since

s3
 
d 1 − 3s
s− = ,
ds (1 − s)2 (1 − s)3

its maximum is obtained at s = 1/3. Thus finally we can calculate our answer

1 (1/3)3 1 1 1
− 2
= − = ,
3 (2/3) 3 12 4

and equality occurs when f (x) = 1 on a set of length 1/3 and f (x) = −s/l = −1/2 elsewhere.
SMT 2012 Calculus Tiebreaker and Solutions February 18, 2012

1. A very tired audience of 9001 attends a concert of Haydn’s Surprise Symphony, which lasts 20
minutes. Members of the audience fall asleep at a continuous rate of 6t people per minute, where
t is the time in minutes since the symphony has begun.
The Surprise Symphony is named such because when t = 8 minutes, the orchestra plays exactly
one very loud note, waking everyone in the audience up. After that note, though, the audience
continues to fall asleep at the same rate as before. Once a member of the audience has fallen
asleep, he or she will stay asleep except for the rude awakening when t = 8 minutes.
How many collective minutes does the audience sleep during the symphony?
Answer: 5696
Solution: Let f (t) = 6t when 0 ≤ t ≤ 20.
Break into two cases: t ≤R8 and t > 8. When 0 ≤ t < 8, the number of people sleeping at any
t
given moment is F1 (t) = 0 6x dx = 3t2 − f (0) = 3t2 , so the total number of minutes slept in
R8 R8
the first eight minutes is 0 F1 (t) dt = 0 3t2 dt = 512.
Rt
When t > 8, the number of people sleeping at any given moment is F2 (t) = 8 6x dx = 3t2 −
R 20
F1 (8) = 3t2 −192, so the total number of minutes slept in the last twelve minutes is 8 F2 (t) dt =
R 20 2
8 3t − 192 dt = 5184. Adding these together, the answer is 5696 .
 2 n
n +n+3
2. Evaluate the limit lim .
n→∞ n2 + 3n + 5

Answer: e−2
Solution: Note that
n n
n2 + n + 3 n2 + 3n + 5 − (2n + 2)
 
lim = lim
n→∞ n2 + 3n + 5 n→∞ n2 + 3n + 5
  n
n+1
= lim 1 − 2
n→∞ n2 + 3n + 5
 n
2
= lim 1 −
n→∞ n

which by a well-known identity is equal to e−2 .

3. Evaluate R∞
(1 + x2 )−2012 dx
R0∞ .
2 −2011 dx
0 (1 + x )
4021
Answer: 4022
Solution: We begin by making the substitution u = tan−1 x:
 
d −1 dx
du = tan x dx = =⇒ (1 + x2 ) du = (1 + tan2 u) du = dx.
dx 1 + x2
Noting that tan−1 0 = 0 and that tan−1 x approaches infinity as x approaches π/2, we have that,
for any positive integer k,
Z ∞ Z π/2
2 −k
(1 + x ) dx = (1 + tan2 u)(1 + tan2 u)−k du
0 0
Z π/2 Z π/2
2 1−k
= (sec u) du = (cos2 u)k−1 du.
0 0
SMT 2012 Calculus Tiebreaker and Solutions February 18, 2012

Note that we can write this as Z 2π

1
(cos2 u)k−1 du
4 0
because this integral is the same over any quadrant of the unit circle. Using the method of
integrating by reduction (see en.wikipedia.org/wiki/Integration by reduction formulae),
we can show that for positive integers m,
Z 2π 2π
2m − 1 2π
Z
2m 1 2m−1
cos2m−2 x dx.

cos x dx = cos x sin x +
0 2m 0 2m 0

Since we are evaluating this on the interval [0, 2π], the first term is zero. In particular, for
m = k − 1, we have
Z ∞
1 2π
Z
(1 + x2 )−k dx = (cos2 u)k−1 du
0 4
0
1 2k − 3 2π
Z 
2k−4
= cos du .
4 2k − 2 0

Carrying out this recursion, we have

Z ∞
3 1 2π
 
1 2k − 3 2k − 5
Z
2 −k
(1 + x ) dx = · ... · 1 dx
0 4 2k − 2 2k − 4 4 2 0
2π (2k − 2)!
= ·
4 (2 · 4 . . . (2k − 4) · (2k − 2))2
π (2k − 2)!
= · 2k−2 .
2 2 (k − 1)!(k − 1)!

Now, our original ratio becomes

R∞
(1 + x2 )−2012 dx π 4022! 2 24020 · 2010! · 2010!
R0∞ = · · ·
2 −2011 dx 2 24022 · 2011! · 2011! π 4020!
0 (1 + x )
4022 · 4021 4021 4021
= = = .
22· 2011 · 2011 2 · 2011 4022
SMT 2012 General Test and Solutions February 18, 2012

1. In preparation for the annual USA Cow Olympics, Bessie is undergoing a new training regime.
However, she has procrastinated on training for too long, and now she only has exactly three
weeks to train. Bessie has decided to train for 45 hours. She spends a third of the time training
during the second week as she did during the first week, and she spends a half of the time training
during the third week as during the second week. How much time did she spend training during
the second week?
Answer: 10
Solution: Let x be the number of hours spent training during the second week. We have that
x
3x + x + = 45. Therefore, x = 10 .
2
2. Nick and Moor participate in a typing challenge. When given the same document to type, Nick
finishes typing it 5 minutes before Moor is done. They compete again using a second document
that is the same length as the first, but now Nick has to type an extra 1200-word document in
addition to the original. This time, they finish at the same time. How fast (in words per minute)
does Nick type? (Assume that they both type at constant rates.)
Answer: 240
Solution 1: Let n and m be Nick and Moor’s typing speeds in words per minute, respectively.
Let w be the number of words in the document, and let t be the time it takes for Moor to
type the document. Then from the first round, we know that d = mt = n(t − 5). From the
second round, we know that d + 1200 = nt, or d = nt − 1200. Equating the two, we obtain
n(t − 5) = nt − 1200, so 5n = 1200, or n = 240 .
Solution 2: From the problem statement, Nick takes 5 minutes to type the extra 1200 words,
so his speed is 1200/5 = 240 words per minute.
3. The Tribonacci numbers Tn are defined as follows: T0 = 0, T1 = 1, and T2 = 1. For all n ≥ 3,
we have Tn = Tn−1 + Tn−2 + Tn−3 . Compute the smallest Tribonacci number greater than 100
which is prime.
Answer: 149
Solution: The first few Tribonacci numbers are 0, 1, 1, 2, 4, 7, 13, 24, 44, 81, 149. 149 is the
smallest Tribonacci number greater than 100, and it also turns out to be prime, so that is our
answer.
4. Steve works 40 hours a week at his new job. He usually gets paid 8 dollars an hour, but if he
works for more than 8 hours on a given day, he earns 12 dollars an hour for every additional
hour over 8 hours. If x is the maximum number of dollars that Steve can earn in one week by
working exactly 40 hours, and y is the minimum number of dollars that Steve can earn in one
week by working exactly 40 hours, what is x − y?
Answer: 96
Solution: In the minimum case, Steve can work 8 hours a day for five days, thereby earning no
overtime pay and earning exactly $320. In the maximum case, Steve works 40 hours without a
single break. This spans two days; there are 16 hours of work at regular pay and 24 hours of
work at overtime pay. Therefore, Steve earns 16 × $8 + 24 × $12 = $416. Therefore, x = 416
and y = 320, so x − y = 96 .
5. There are 100 people in a room. 60 of them claim to be good at math, but only 50 are actually
good at math. If 30 of them correctly deny that they are good at math, how many people are
good at math but refuse to admit it?
SMT 2012 General Test and Solutions February 18, 2012

Answer: 10
Solution: By the principle of inclusion and exclusion, the sum of the number of people who are
good at math and the number of people who claim to be good at math minus the number of
people in both categories gives the number of people who either are good at math or claim they
are good at math. Let x be the number of people in both categories. Then 50+60−x = 100−30,
so x = 40. Thus we are left with 50 − 40 = 10 people who are good at math but refuse to
admit it.

6. A standard 12-hour clock has hour, minute, and second hands. How many times do two hands
cross between 1:00 and 2:00 (not including 1:00 and 2:00 themselves)?
Answer: 119
Solution: We know that the hour and minute hands cross exactly once. Let m be the number
of minutes past one o’clock that this happens. The angle between the minute hand and the 12
◦
must be equal to the angle between the hour hand and the 12. Since 1 minute is 360 60 = 6◦ on
360◦
the clock and 1 hour is 12 = 30◦ , we have 6m = 30(1 + 60 ), so m = 11 = 5 11 . Note that the
m 60 5

second hand is not at the same position at this time, so we do not have to worry about a triple
crossing.
On the other hand, the second hand crosses the hour hand once every minute, for a total of 60
crossings. Also, the second hand crosses the minute hand once every minute except the first
and last, since those crossings take place at 1:00 and 2:00, for a total of 58 crossings. There is a
grand total of 1 + 60 + 58 = 119 crossings.

7. Define a set of positive integers to be balanced if the set is not empty and the number of even
integers in the set is equal to the number of odd integers in the set. How many strict subsets of
the set of the first 10 positive integers are balanced?
Answer: 250
Solution: The set of the

first ten positive integers contains five odd integers and five integers.
Therefore, there are k5 ways to choose k odd integers from five odd integers, and also there are
5 5 2



k ways to choose k even integers from five even integers. Therefore, there are k ways to
pick a balanced subset containing k odd integers and k even integers. Therefore, the answer is
4  2
X 5
= 52 + 102 + 102 + 52 = 250 .
i
i=1

8. At the 2012 Silly Math Tournament, hamburgers and hot dogs are served. Each hamburger
costs $4 and each hot dog costs $3. Each team has between 6 and 10 members, inclusive, and
each member buys exactly one food item. How many different values are possible for a team’s
total food cost?
Answer: 23
Solution: The minimum food cost for a team is 6($3) = $18, and the maximum food cost is
10($4) = $40. Note that all intermediate values can be achieved. Suppose n dollars can be
achieved by purchasing a hamburgers and b hot dogs, where 18 ≤ n < 40. If b > 0, then n + 1
dollars can be achieved by purchasing a + 1 hamburgers and b − 1 hot dogs. If b = 0, then n + 1
dollars can be achieved by purchasing a − 2 hamburgers and b + 3 hot dogs. (This increases
the number of team members by 1.) Repeating this process until $40 is reached, the number
of team members cannot decrease, and since we end up with 10 team members, the number of
team members is always contained within 6 and 10.
 SMT 2012 General Test and Solutions February 18, 2012

Hence the number of different values is 40 − 18 + 1 = 23 .

9. How many ordered sequences of 1’s and 3’s sum to 16? (Examples of such sequences are
{1, 3, 3, 3, 3, 3} and {1, 3, 1, 3, 1, 3, 1, 3}.)
Answer: 277
Solution: Notice that there are 6 sets of 1’s and 3’s that sum to 16. For a given set suppose

n 3’s we have a total of (16 − 3n) + n = 16 − 2n numbers so we want to compute
there are
16−2n
n . Hence the total number of possible sequences is:
5  
X 16 − 2n
= 277 .
n
n=0

10. How many positive numbers up to and including 2012 have no repeating digits?
Answer: 1242
Solution: All one-digit numbers have no repeating digits, so that gives us 9 numbers. For a
two-digit number to have no repeating digits, the first digit must be between 1 and 9, while
the second digit must not be equal to the first, giving us 9 · 9 = 81 numbers. For a three-digit
number to have no repeating digits, the first digit must be between 1 and 9, the second digit
must not be equal to the first, and the third digit must not be equal to either of the two, giving
us 9 · 9 · 8 = 648 numbers. For a four-digit number between 1000 and 1999 to have no repeating
digits, the first digit must be 1, the second digit must not be equal to the first, and so on, giving
us 1 · 9 · 8 · 7 = 504 numbers. Finally, there are no numbers between 2000 and 2012 inclusive
with no repeating digits, so the total is 9 + 81 + 648 + 504 = 1242 .
11. Nikolai and Wolfgang are math professors at a European university, so they enjoy researching
math problems. Interestingly, each is able to do math problems at a constant rate. One day, the
university gives the math department a problem set to do. Working alone, Nikolai can solve all
the problems in 6 hours, while Wolfgang can solve them in 8 hours. When they work together,
they are more efficient because they are able to discuss the problems, so their combined output
is the sum of their individual outputs plus 2 additional problems per hour. Working together,
they complete the problem set in 3 hours. How many problems are on the problem set?
Answer: 48
Solution: Suppose that there are a total of x problems on the problems set. Then Nikolai’s
rate is x6 problems per hour, while Wolfgang’s is x8 . their combined rate (including the efficiency
bonus) is
x x
+ + 2,
6 8
x
which is given to be equal to 3 . Solving for x, we obtain x = 48 .
12. ABC is an equilateral triangle with side length 1. Point D lies on AB, point E lies on AC, and
points G and F lie on BC, such that DEF G is a square. What is the area of DEF G?
√
Answer: 21 − 12 3
Solution: Let x be the length of a side of square DEF G. Then DE = EF = x. Note that
4ADE is equilateral since DE k BC and hence 4ADE ∼ 4ABC, so AE = DE = x, and
consequently EC = 1 − x. Since 4ECF is a 30◦ − 60◦ − 90◦ triangle, we have the proportion
√
EF x 3
= = ,
EC 1−x 2
 SMT 2012 General Test and Solutions February 18, 2012

√
3
√ √
so x = √
2+ 3
= 2 3 − 3. Hence the area of DEF G is x2 = 21 − 12 3 .

13. Define a number to be boring if all the digits of the number are the same. How many positive
integers less than 10000 are both prime and boring?
Answer: 5
Solution: The one-digit boring primes are 2, 3, 5, and 7. The only two-digit boring prime is 11,
since 11 divides all other two-digit boring numbers. No three-digit boring numbers are prime,
since 111 divides all of them and 111 = 3 × 37. No four-digit boring numbers are prime since
they are all divisible by 11. Therefore, there are 5 positive integers less than 10000 which are
both prime and boring.

14. Given a number n in base 10, let g(n) be the base-3 representation of n. Let f (n) be equal to
the base-10 number obtained by interpreting g(n) in base 10. Compute the smallest positive
integer k ≥ 3 that divides f (k).
Answer: 7
Solution: Using brute force, we note that 3, 4, 5, and 6 are invalid, but 7 = 213 . Thus, the
answer is 7 .

15. ABCD is a parallelogram. AB = BC = 12, and ∠ABC = 120◦ . Calculate the area of
parallelogram ABCD.
√
Answer: 72 3
Solution: Since opposite sides of a parallelogram are equal, AB = BC = CD = DA = 12.
Since adjacent angles of a parallelogram are supplementary, ∠BCD = ∠CDA = 60◦ . Therefore,
when we draw diagonal BD, we get two equilateral √ triangles, both with side length 12. The area
s2 3
of an equilateral triangle with side length s is , so therefore the area of the parallelogram
√ 4
122 × 3 √
is 2 × = 72 3 .
4
16. Given a 1962-digit number that is divisible by 9, let x be the sum of its digits. Let the sum of
the digits of x be y. Let the sum of the digits of y be z. Compute the maximum possible value
of z.
Answer: 9
Solution: Let theP1962 digit number be a. First of all note that 9 | x, y, z. This is because
a = P1962 i ≡ 1962
P
i=1 ia 10 i=1 ai mod 9 = x. Now clearly the largest value of x would be obtained if
1962 i
a = i=1 9(10) , hence x ≤ 1962(9) = 17658. It follows that, y ≤ 1 + 7 + 6 + 5 + 8 = 36 and
finally that z ≤ 9. However 9 | z so z = 9 .
_ _
17. A circle with radius 1 has diameter AB. C lies on this circle such that AC / BC= 4. AC divides
the circle into two parts, and we will label the smaller part Region I. Similarly, BC also divides
the circle into two parts, and we will denote the smaller one as Region II. Find the positive
difference between the areas of Regions I and II.
3π
Answer: 10
Solution: Let O be the center of the circle. Note that CO bisects AB, so the areas of 4ACO
and 4BCO are equal. Hence, the desired difference in segment areas is equal to the difference
 SMT 2012 General Test and Solutions February 18, 2012

_
2π
in the areas of the corresponding sectors. The sector corresponding to AC has area 5 , and the
_ 3π
π
sector corresponding to BC has area 10 , so the desired difference is .
10
18. John is on the upper-left corner of a 3 × 3 grid. Once per minute, John randomly chooses a
square that is either horizontally or vertically adjacent to his current square and moves there.
What is the expected number of minutes that John needs to get to the center square?
Answer: 6
Solution: Let a be the expected number of minutes necessary to move from a corner to the
center square. Furthermore, let b be the expected number of minutes necessary to move from
one of the middle squares in any of the leftmost or rightmost rows and columns to the center.
We have that a = b + 1 and b = 2(a+1)
3 + 31 . Solving these two linear equations gives us b = 5
and a = 6 .

19. If f is a monic cubic polynomial with f (0) = −64, and all roots of f are non-negative real
numbers, what is the largest possible value of f (−1)? (A polynomial is monic if it has a leading
coefficient of 1.)
Answer: −125
Solution: If the three roots of f are r1 , r2 , r3 , we have f (x) = x3 −(r1 +r2 +r3 )x2 +(r1 r2 +r1 r3 +
r2 r3 )x − r1 r2 r3 , so f (−1) = −1 − (r1 + r2 + r3 ) − (r1 r2 + r1 r3 + r2 r3 ) − r1 r2 r3 . Since r1 r2 r3 = 64,
the arithmetic mean-geometric mean inequality reveals that r1 + r2 + r3 ≥ 3(r1 r2 r3 )1/3 = 12 and
r1 r2 +r1 r3 +r2 r3 ≥ 3(r1 r2 r3 )2/3 = 48. It follows that f (−1) is at most −1−12−48−64 = −125 .
We have equality when all roots are equal, i.e. f (x) = (x − 4)3 .

20. A permutation of the first n positive integers is quadratic if, for some positive integers a and
b such that a + b = n, a 6= 1, and b 6= 1, the first a integers of the permutation form an
increasing sequence and the last b integers of the permutation form a decreasing sequence, or
if the first a integers of the permutation form a decreasing sequence and the last b integers of
the permutation form an increasing sequence. How many permutations of the first 10 positive
integers are quadratic?
Answer: 1020
Solution: Clearly, either 1 or 10 must be in the middle of the permutation. Assume without
loss of generality that 10 is; we can construct an equivalent permutation with 1 in the middle
by replacing each number i with 11 − i. We can pick any nonempty strict subset of the first 9
positive integers, sort it, place it at the beginning of the permutation, then place 10, then place
the unchosen numbers in decreasing order. There are 29 − 2 = 510 ways to do this. Therefore,
there are 2 × 510 = 1020 quadratic permutations of the first 10 positive integers.

21. In trapezoid ABCD, BC k AD, AB = 13, BC = 15, CD = 14, and DA = 30. Find the area of
ABCD.
Answer: 252
Solution: We can use the standard method of setting up a two-variable system and solving for
the height of the trapezoid. However, since one base is half the length of the other, we may
take a shortcut. Extend AB and CD until they meet at E. Clearly, BC is a midline of triangle
EAD, so we have EA = 2BA = 26 and ED = 2CD = 28. The area of EAD is therefore four
times that of a standard 13-14-15 triangle, which we know is 12 · 14 · 12 = 84 (since the altitude
 SMT 2012 General Test and Solutions February 18, 2012

to the side of length 14 splits the triangle into 9-12-15 and 5-12-13 right triangles). The area of
the trapezoid is 34 the area of EAD by similar triangles, and is therefore 3 · 84 = 252 .
A similar solution draws lines from B and C to the midpoint of AD to form three 13 − 14 − 15
triangles.

22. Two different squares are randomly chosen from an 8 × 8 chessboard. What is the probability
that two queens placed on the two squares can attack each other? Recall that queens in chess
can attack any square in a straight line vertically, horizontally, or diagonally from their current
position.
13
Answer: 36
Solution: All squares that are on the edge of the chessboard can hit 21 squares; there are 28
such squares. Now consider the 6 × 6 chessboard that is obtained by removing these bordering
squares. The squares on the edge of this board can hit 23 squares; there are 20 of these squares.
Now we consider the 12 squares on the boundary of the 4 × 4 chessboard left; each of these
squares can hit 25 squares. The remaining 4 can hit 27 squares. The probability then follows as
21 × 28 + 23 × 20 + 25 × 12 + 27 × 4 13
= .
64 × 63 36
23. Circle O has radius 18. From diameter AB, there exists a point C such that BC is tangent to
O and AC intersects O at a point D, with AD = 24. What is the length of BC?
√
Answer: 18 5
AB AC
Solution: Since ∠ADB = ∠ABC = 90◦ , 4ABC ∼ 4ADB. In particular, = , so
AD AB
AB 2 362
AC = . Therefore, AC = = 54. Since AD = 24, DC = 30. By Power of a Point,
AD √ 24
√
BC = 30 × 54 = 18 5 .

24. The quartic (4th-degree) polynomial P (x) satisfies P (1) = 0 and attains its maximum value of
3 at both x = 2 and x = 3. Compute P (5).
Answer: −24
Solution: Consider the polynomial Q(x) = P (x) − 3. Q has roots at x = 2 and x = 3.
Moreover, since these roots are maxima, they both have multiplicity 2. Hence, Q is of the form
a(x − 2)2 (x − 3)2 , and so P (x) = a(x − 2)2 (x − 3)2 + 3. P (1) = 0 =⇒ a = − 34 , allowing us to
compute P (5) = − 43 (9)(4) + 3 = −24 .
1
25. Compute the ordered pair of real numbers (a, b) such that a < k < b if and only if x3 + 3 = k
x
does not have a real solution in x.
Answer: (−2, 2)
1
Solution: Substitute y = x3 , so now we want to find the values of k such that y + = k has
y
1
no real solutions in y. In particular, since y = x3 is an invertible function, x3 + 3 = k does not
x
1
have a real solution in x if and only if y + = k has no real solutions in y. Therefore, clearing
y
1
denominators of y + = k gives us y + 1 = ky, so y 2 − ky + 1 = 0, so this quadratic equation
2
y
SMT 2012 General Test and Solutions February 18, 2012

has no solutions when the discriminant is negative. The discriminant is k 2 − 4, which is negative
when −2 < k < 2, so therefore the ordered pair is (−2, 2) .
SMT 2012 General Tiebreaker and Solutions February 18, 2012

1. Compute 12 − 22 + 32 − · · · − 182 + 192 − 202 .

Answer: −210
Solution: Note that (2n + 1)2 − (2n + 2)2 = −4n − 3. Therefore, the sum is equivalent to
X9
−2i − 3 = −2(9)(10) − 30 = −210 .
i=0

2. Reimu and Marisa are playing a game with 2012 coins. Reimu flips all 2012 coins, and then is
permitted to flip any subset of the 2012 coins exactly once more. After this, Reimu pays Marisa
$2 for every head on the table, whereas Marisa pays Reimu $1 for every tail on the table. Who
is more likely to earn a profit, and what is the expected profit for that person, in dollars?
Answer: Reimu, $503.
Solution: We compute the expected profit for a game using 1 coin and then multiply this by
2012. We are allowed to do this by linearity of expectation. If Reimu flips the coin and gets
a tail, then she receives $1 guaranteed. If Reimu flips the coin and gets a head instead, she
1 1
then reflips the coin and gets $1 with probability and loses $2 with probability . Therefore,
    2 2
1 1 1
Reimu expects to win 1 + −2 = 0.25 dollars. Multiplying by 2012, Reimu expects
2 4 4
to win $503.

3. We define n to be a squarefree integer if, for every prime p, p2 does not divide n. Let f (n) be the
sum of the reciprocals of all the divisors of n. We define n to be an amazing integer if f (n) = 2.
How many squarefree amazing integers are there?
Answer: 1
k
σ(n) Y pi + 1
Solution: Note that f (n) = = . In order for f (n) to be integral, the only primes
n pi
i=1
that can divide n are 2 and 3, since those are the only consecutive primes. Note that f (6) = 2,
but 1, 2, and 3 do not satisfy this constraint. Therefore, there is exactly 1 squarefree amazing
integer.
SMT 2012 Geometry Test and Solutions February 18, 2012

_ _
1. A circle with radius 1 has diameter AB. C lies on this circle such that AC / BC= 4. AC divides
the circle into two parts, and we will label the smaller part Region I. Similarly, BC also divides
the circle into two parts, and we will denote the smaller one as Region II. Find the positive
difference between the areas of Regions I and II.
3π
Answer: 10
Solution: Let O be the center of the circle. Note that CO bisects AB, so the areas of 4ACO
and 4BCO are equal. Hence, the desired difference in segment areas is equal to the difference
_
2π
in the areas of the corresponding sectors. The sector corresponding to AC has area 5 , and the
_ 3π
π
sector corresponding to BC has area 10 , so the desired difference is .
10
2. In trapezoid ABCD, BC k AD, AB = 13, BC = 15, CD = 14, and DA = 30. Find the area of
ABCD.
Answer: 252
Solution: We can use the standard method of setting up a two-variable system and solving for
the height of the trapezoid. However, since one base is half the length of the other, we may
take a shortcut. Extend AB and CD until they meet at E. Clearly, BC is a midline of triangle
EAD, so we have EA = 2BA = 26 and ED = 2CD = 28. The area of EAD is therefore four
times that of a standard 13-14-15 triangle, which we know is 12 · 14 · 12 = 84 (since the altitude
to the side of length 14 splits the triangle into 9-12-15 and 5-12-13 right triangles). The area of
the trapezoid is 43 the area of EAD by similar triangles, and is therefore 3 · 84 = 252 .
A similar solution draws lines from B and C to the midpoint of AD to form three 13 − 14 − 15
triangles.

3. Let ABC be an equilateral triangle with side length 1. Draw three circles Oa , Ob , and Oc with
diameters BC, CA, and AB, respectively. Let Sa denote the area of the region inside Oa and
outside of Ob and Oc . Define Sb and Sc similarly, and let S be the area of the region inside all
three circles. Find Sa + Sb + Sc − S.
√
3
Answer: 2
Solution: Let x be 1/4 the area of ABC, and let y be the area of a 60 degree sector of Oa
minus x. Note that
Sa = Sb = Sc = 3x + y, S = x + 3y,
√
so Sa + Sb + Sc − S = 8x = 2|4ABC| = 3/2 .

4. Let ABCD be a rectangle with area 2012. There exist points E on AB and F on CD such
that DE = EF = F B. Diagonal AC intersects DE at X and EF at Y . Compute the area of
triangle EXY .
503
Answer: 6
Solution: Let (XY Z) denote the area of triangle XY Z.
After a bit of angle-chasing, we can use SAS congruence to prove that 4DEF ∼ = 4BF E, so
EB ∼= DF and therefore AE ∼ = F C. If we draw altitudes from E and F onto CD and AB,
respectively, we note that 2AE = 2F C = DF = BE, so AE = 13 AB.
Next, note that 4AEX ∼ 4CDX, so CX CD
AX = AE = 3. Also, 4CF Y ∼ 4AEY , so
CY
AY = CF
AE =
1 1
1. Hence, XY = 4 AC =⇒ (EXY ) = 4 (EAC).
SMT 2012 Geometry Test and Solutions February 18, 2012

Finally, AE = 13 AB =⇒ (EAC) = 31 (BAC). Since BAC is half the rectangle and therefore has
503
area 1006, we get (EXY ) = 1006
12 = .
6
5. What is the radius of the largest sphere that fits inside an octahedron of side length 1?
1
Answer: √
6
Solution: It is obvious that the sphere must be tangent to each face, because if not, then it can
be moved so that it is tangent to four faces; now the radius can be increased until the sphere is
tangent to the other four. Additionally, it is clear that the center of the sphere should be in the
center of the octahedron.
Now notice that the sphere must be tangent to the octahedron at the centroid of each face. This
can be seen by symmetry. It is clear that it should be tangent somewhere along the median
from one vertex to the opposite side, and this is true for all three medians, which meet at the
centroid.
Now we can proceed in a few ways. One way is to isolate one half of the octahedron i.e. a
square-based pyramid. Slice this pyramid in half perpendicular to the square base and parallel
to one of the sides of the square base. This slice will go through the medians of two opposite
triangular faces, in addition to the center of the
√ sphere itself. Hence, we get an isosceles triangle
ABC with base BC = 1 and legs of length 3/2. O, the center of the sphere, is the midpoint
of BC. The radius of the sphere is the altitude from O to AB. If this altitude intersects AB at
D, then we have
OD · AB = AO · BO,
√ √
(1/ 2)(1/2)
since both equal twice the area of AOB, and so DO = AO·BO AB = √
3/2
= 1/ 6 .
√
Alternatively, note that our octahedron can be obtained by reflecting the region x+y+z ≤ 1/ 2,
x, y, z ≥ 0 across the xy, yz, and zx planes. The inscribing sphere√has its center at
√ origin, so its
radius is the distance from the origin to the plane x + y + z = 1/ 2, which is 1/ 6.

6. A red unit cube ABCDEF GH (with E below A, F below B, etc.) is pushed into the corner
of a room with vertex E not visible, so that faces ABF E and ADHE are adjacent to the wall
and face EF GH is adjacent to the floor. A string of length 2 is dipped in black paint, and one
of its endpoints is attached to vertex A. How much surface area on the three visible faces of the
cube can be painted black by sweeping the string over it?
√
Answer: 2π 3
+ 3−1
Solution: First, it is clear that all of face ABCD can be painted black. This has area 1.
Now we look at the other two visible faces. By symmetry, we only need to consider one of
these faces, say BCGF . Unfold BCGF along BC so that it is coplanar with ABCD, forming a
rectangle AF 0 G0 D with width 1 and height 2. Now, it is clear that the region that can be painted
on BCGF is precisely the part of BCG0 F 0 that is at most two units away from A. Let a circle
centered at A with radius two intersect DG0 at X. Since AX = 2, AD = 1, and AD ⊥ XD,
we conclude that m∠DAX = π3 =⇒ ∠F 0 AX = π6 . Letting (P1 P2 . . . Pn ) denote the area of the
n-gon with vertices P1 , . . . , Pn , we can write the desired area as
√
0 22 π 3
area of sector F AX + (AXD) − (ABCD) = + − 1.
12 2
SMT 2012 Geometry Test and Solutions February 18, 2012

Putting all this together, we get our final answer to be

√ !
3 2π √
1 + 2 π/3 + −1 = + 3−1 .
2 3

7. Let ABC be a triangle with incircle O and side lengths 5, 8, and 9. Consider the other tangent
line to O parallel to BC, which intersects AB at Ba and AC at Ca . Let ra be the inradius of
triangle ABa Ca , and define rb and rc similarly. Find ra + rb + rc .
√
6 11
Answer: 11
Solution: We claim that the answer is equal to the inradius in general. Let Ta = ABa Ca ,
Tb = Ab BCb , Tc = Ac Bc C be the smaller triangles cut by the tangents drawn to O. Also let
D, E, and F be the points of tangency between O and BC, CA, and AB respectively. By
considering the fact that tangents to O from the same point should have the same length, we
have ABa + Ba Ca + Ca A = AE + AF . If we sum this over all vertices, then we can see that
the sum of the perimeters of Ta , Tb , and Tc equals the perimeter of A. Then, the Principle of
Similarity gives ra + rb + rc = r where r is inradius of ABC. The inradius can be calcultaed by
Heron’s Formula as
p √ √
s(s − a)(s − b)(s − c) 11 · 6 · 3 · 2 6 11
r= = = .
s 11 11

Alternatively, let ha denote the height of the altitude from A to BC, and let r be the inradius
of ABC. Since 4ABC ∼ 4ABa Ca and since the altitude from A to Ba Ca has length ha − 2r,
we get
ra ha − 2r
= .
r ha
Noticing that
(ABC) aha
r= 1 = ,
2 (a
+ b + c) a + b+c
we get
a
ra = r − 2r .
a+b+c
Applying the same reasoning to ra and rb , we can compute

ra + rb + rc = 3r − 2r = r.

8. Let ABC be a triangle with side lengths 5, 6, and 7. Choose a radius r and three points outside
the triangle Oa , Ob , and Oc , and draw three circles with radius r centered at these three points.
If circles Oa and Ob intersect at C, Ob and Oc intersect at A, Oc and Oa intersect at B, and all
three circles intersect at a fourth point, find r.
√
35 6
Answer: 24
Solution: Let the point where all three circles intersect be denoted as X.
First, note that AOb XOc , BOc XOa , and COa XOb are all rhombi. This helps us easily prove
that AOb k BOa . Since these segments are also congruent, we get that ABOa Ob is a parellogram,
and hence AB ∼ = Oa Ob . Similarly, BC ∼
= Ob Oc and CA ∼= Oc Oa .
 SMT 2012 Geometry Test and Solutions February 18, 2012

Since r is clearly the circumradius of Oa Ob Oc , this equals the circumradius of ABC, which is
computed as
√
abc 5·6·7 35 6
= √ = .
4(ABC) 4 9·4·3·2 24

Alternatively, consider a homothety of ratio 2 around X. Let A0 X, B 0 X, and C 0 X be diameters

of circles Oa , Ob , and Oc , respectively. Then the homothety takes Oa to A0 , Ob to B 0 , and Oc
to C 0 ; furthermore, since Oa Ob perpendicularly bisects XC, the midpoint of Oa Ob is taken to
C—that is, A0 B 0 has midpoint C, and similarly A0 C 0 has midpoint B and B 0 C 0 has midpoint A.
Therefore, ABC is the midpoint triangle of A0 B 0 C 0 , from which we conclude that A0 BC, AB 0 C,
and ABC 0 are each congruent to ABC. But r is the circumradius of each of these triangles,
hence is the circumradius of ABC.

9. In quadrilateral ABCD, m∠ABD ∼

= m∠BCD and ∠ADB = ∠ABD + ∠BDC. If AB = 8 and
AD = 5, find BC.
39
Answer: 5
Solution: Note that ∠ADB and ∠CBD are supplementary. Therefore, we can extend AD past
D to a new point C 0 such that 4DBC ∼
= 4BDC 0 (alternatively, consider flipping 4DBC over
∼
the altitude to BD). Since ∠ABD = ∠AC 0 B, we have 4ABD ∼ 4AC 0 B, and so

AC 0 AB AB 2 64
= =⇒ AC 0 = = .
AB AD AD 5

39
Since AC 0 = AD + DC 0 , we get DC 0 = BC = 64
5 −5= .
5
p
10. A large flat plate of glass is suspended 2/3 units above a large flat plate of wood. (The glass √
is infinitely thin and causes no funny refractive effects.) A point source of light is suspended 6
units above the glass plate. An object rests on the glass plate of the following description. Its
base is an isosceles trapezoid ABCD with ABkDC, AB = AD = BC = 1, and DC = 2. The
point source of light is directly above
√ the midpoint of CD. The object’s upper face is a triangle
EF G with EF = 2, EG = F G = 3. G and AB lie on opposite sides of the rectangle EF CD.
The other sides of the object are EA = ED = 1, F B = F C = 1, and GD = GC = 2. Compute
the area of the shadow that the object casts on the wood plate.
√
Answer: 4 3
Solution: We have ∠A = ∠B = 120◦ and ∠C = ∠D = 60◦ at the base, and the three “side”
faces – ADE, BCF , and CDG – are all equilateral triangles. If those faces are folded down to
the glass plate, they will form a large equilateral triangle of side length 3. Let E0 , F0 , and G0
be the vertices of this equilateral triangle corresponding to E, F and G, respectively; the large
triangle can be folded up along AD, CD, and BD respectively to form the three side faces of
the object.
Observe that M , the midpoint of CD, is the centroid of E0 F0 G0 . As side ADE is folded along
AD, which is perpendicular to E0 M , the projection of E onto the glass plate still lies on EM .
This also holds for the projections of F and G, so projections E1 , F1 , and G1 of E, F , and G
lie on E0 M , F0 M and G0 M respectively.
Since EF CD is a rectangle, E1 F1 CD is as well. Thus E1 D is perpendicular to EA. From E0 E1
being perpendicular to AD we can conclude that E1 should be the center of triangle ADE0 .
SMT 2012 Geometry Test and Solutions February 18, 2012

Symmetry gives AE = DE = E0 E, so AE0 DE should be a regular tetrahedron. A similar

argument applies to BF0 CF .
√ √
The next step is to figure out the location of G. As EG = 3 and DG = 2, it follows that
∠DEG is right. Similarly ∠CF G is also right, so plane EF G should be perpendicular to plane
EF CD.
Now we cut the whole object along the perpendicular bisector plane of AB and consider its cross-
section along the plane. It will cut AB and EF along their midpoints N and P respectively.
As ABM P forms √ a regular tetrahedron
√ of side length 1 and N is midpoint of AB, we have
N M = N P = 3/2.√Also M G = 3 and ∠M P G is right. Let Q be the midpoint of M G;
then P Q = M Q = 3/2, since right triangles are inscribed in semicircles. It follows √ that
N P M and QM√ P are congruent and N P and M G are parallel. From M G = M G0 = 3 and
N P = N M = 3/2, this gives similarity between N M P and M G0 G, and GG0 = 2P M = 2.
Therefore DCGG0 also forms a regular tetrahedron.
Since AE0 DE , BF0 CF , and CG0 DG are all regular tetrahedrons, we have three lines E0 E,
F0 F , and G0 G meeting at a point X where E0 F0 G0 X forms a regular tetrahedron of side length
3. Thus we finally demystified our object completely: it was obtained by cutting the regular
tetrahedron E0 F0 G0 X along planes EF G, ADE, BCF , CDG. Moreover we find that X is
actually our point source, as it is also
√ directly above M - both the midpoint of CD and the
center of E0 F0 G0 - and its height is 6, the same as that of point source. So the projection
of the object to the glass plate will be exactly E0 F0 G0 , an equilateral triangle of side length 3.
Hence the projection down to the wood plate will give an equilateral triangle of side length 4,
√
and our answer is its area, 4 3 .
SMT 2012 Geometry Tiebreaker and Solutions February 18, 2012

1. In 4ABC, the altitude to AB from C partitions 4ABC into two smaller triangles, each of
which is similar to 4ABC. If the side lengths of 4ABC and of both smaller triangles are all
integers, find the smallest possible value of AB.
Answer: 25
Solution: Let the altitude from C to AB intersect AB at D. Note that ∠A and ∠B must be
acute since AD partitions ABC into two triangles. Since triangles ADC and ADB each contain
right angles, we conclude that ∠ACB must be right.
Now, we must have 4ABC ∼ 4ACD ∼ 4CBD. We now have enough information to determine
AD, AB, and CD in terms of AB, BC, and CA, which we denote as c, a, and b, respectively.
We have
b c b2
= =⇒ AD = ,
AD b c
a c a2
= =⇒ BD = ,
BD a c
1 1 ab
CD · c = ab =⇒ CD = .
2 2 c
Obviously, (a, b, c) has to be in the form (kx, ky, kz) where (x, y, z) is a Pythagorean triple with
no common factors, and k is a positive integer. Note that in particular x, y, and z must be
pairwise coprime, because of the constraint x2 + y 2 = z 2 .
Given this, we need to find k such that

kz | k 2 y 2 ⇐⇒ z | ky 2 ⇐⇒ z | k,

so the smallest such k is when k = z. This choice makes BD and CD integral as well, so given
any Pythagorean triple (x, y, z) with pairwise coprime entries, the minimum k required equals
z, and the minimum possible value of c is z 2 . The smallest such Pythagorean triple is (3, 4, 5),
so report 25 .

2. Four points O, A, B, and C satisfy OA = OB = OC = 1, ∠AOB = 60◦ , and ∠BOC = 90◦ . B

is between A and C (i.e. ∠AOC is obtuse). Draw three circles Oa , Ob , and Oc with diameters
OA, OB, and OC, respectively. Find the area of region inside Ob but outside Oa and Oc .
√
Answer: (1 + 3)/8
Solution: Let D, E, and F be the midpoints of BC, CA, and AB, respectively. Observe that
Oa goes through E and F , Ob goes through D and F , and Oc goes through D and E; the radii
to these points are all midlines of some triangle (either AOB, AOC, or BOC) and are parallel
to sides of length 1. Hence, the region Ob \ (Oa ∪ Oc ) has four vertices D, E, F , and B: along
DE and EF the boundary is concave with the shape of an arc of radius 1, while along F B and
BD it is convex. But note that DEF B is a parallelogram, so DE = BF . This implies that two
arcs–one from D to E coming from Oc and the other from B to F coming from Ob –are congruent,
so the convex region outside segment F B can be fit into the concave region inside segment DE.
Thinking similarly for EF and DB we have that the area of  Ob \(O
√
a ∪Oc ) is precisely
 the area of
√
1 1 3 1 ◦
parallelogram DEF B. Thus, the answer is |ABC|/2 = 2 2 + 4 − 2 sin 150 = (1 + 3)/8 .
This problem can also be solved by use of the Principle of Inclusion-Exclusion: in particular,
the desired region has area of

Ob − (Oa ∩ Ob + Oc ∩ Ob − Oa ∩ Ob ∩ Oc ),
SMT 2012 Geometry Tiebreaker and Solutions February 18, 2012

where it is clear that Oa ∩ Ob ∩ Oc = Oa ∩ Oc and is thus easily computable.

3. Circles with centers O1 , O2 , and O3 are externally tangent to each other and have radii 1, 12 ,
and 41 , respectively. Now for i > 3, let circle Oi be defined as the circle externally tangent to
circles Oi−1 and Oi−2 with radius 21−i that is farther from Oi−3 . As n approaches infinity, the
area of triangle O1 O2 On approaches the value A. Find A.
√
14
Answer: 6
Solution: First, note that all triangles Oi Oi+1 Oi+2 are similar (for i ≥ 1). In particular,
this implies that for all such i, m∠Oi+1 Oi+3 Oi+2 ∼
= m∠Oi Oi+2 Oi+1 and m∠Oi+3 Oi+1 Oi+2 ∼ =
m∠Oi+4 Oi+2 Oi+3 . Hence, ∠Oi Oi+2 Oi+4 = π i.e. the points are collinear.
From here, there are many ways to proceed. One way is to note that the collinearity of all O2n
and all O2n+1 implies that the desired area is simply the infinite sum of areas ∞
P
i=1 (Oi Oi+1 Oi+2 ).
O1 O2 O3 has side lengths 32 , 54 , and 43 , so its area is 16
1
the area of a 3-5-6 triangle, which by
√ √
Heron’s Formula is 7 · 4 · 2 · 1 = 2 14. Furthermore, for all i > 1, Oi Oi+1 Oi+2 has 41 the area
√
14
of Oi−1 Oi Oi+1 , so the desired sum is geometric with first term 8 and common ratio 41 . Hence,
√ √
14 14
report 1−8 1 = .
4 6
SMT 2012 Advanced Topics Test and Solutions February 18, 2012

1. Define a number to be boring if all the digits of the number are the same. How many positive
integers less than 10000 are both prime and boring?
Answer: 5
Solution: The one-digit boring primes are 2, 3, 5, and 7. The only two-digit boring prime is 11,
since 11 divides all other two-digit boring numbers. No three-digit boring numbers are prime,
since 111 divides all of them and 111 = 3 × 37. No four-digit boring numbers are prime since
they are all divisible by 11. Therefore, there are 5 positive integers less than 10000 which are
both prime and boring.

2. Find the sum of all integers x, x ≥ 3, such that 201020112012x (that is, 201020112012 interpreted
as a base x number) is divisible by x − 1.
Answer: 32
Solution: Note that x ≡ 1 (mod x − 1), and so xn ≡ 1 (mod x − 1) for all positive integers n.
Hence, the number 201020112012 in base x is congruent to the sum of its digits = 12 (mod x−1).
Therefore, we simply need to find all x ≥ 3 such that 12 ≡ 0 (mod x − 1) ⇐⇒ (x − 1) | 12, so
x − 1 = 1, 2, 3, 4, 6, 12 =⇒ x = 3, 4, 5, 7, 13 (since x ≥ 3). Hence, our answer is 32 .

3. Given that log10 2 ≈ 0.30103, find the smallest positive integer n such that the decimal repre-
sentation of 210n does not begin with the digit 1.
Answer: 30
Solution: Observe that 210n begins with the digit 1 when the fractional part of log10 210n =
10n log10 2 ≈ 3.0103n is < log10 2. Therefore, we want 0.0103n > log10 2 ≈ 0.30103 ⇒ n ≥ 30 .

4. Two different squares are randomly chosen from an 8 × 8 chessboard. What is the probability
that two queens placed on the two squares can attack each other? Recall that queens in chess
can attack any square in a straight line vertically, horizontally, or diagonally from their current
position.
13
Answer: 36
Solution: All squares that are on the edge of the chessboard can hit 21 squares; there are 28
such squares. Now consider the 6 × 6 chessboard that is obtained by removing these bordering
squares. The squares on the edge of this board can hit 23 squares; there are 20 of these squares.
Now we consider the 12 squares on the boundary of the 4 × 4 chessboard left; each of these
squares can hit 25 squares. The remaining 4 can hit 27 squares. The probability then follows as
21 × 28 + 23 × 20 + 25 × 12 + 27 × 4 13
= .
64 × 63 36
5. A short rectangular table has four legs, each 8 inches long. For each leg Bill picks a random
integer x, 0 ≤ x < 8 and cuts x inches off the bottom of that leg. After he’s cut all four legs,
compute the probability that the table won’t wobble (i.e. that the ends of the legs are coplanar).
43
Answer: 512
Solution: We can describe a table by a, b, c, d (1 ≤ a, b, c, d ≤ 8), giving the final lengths
of each of the four legs in clockwise order. How much a table is tipped north to south will
be determined by the difference between the lengths a, b and c, d, and east to west by the
difference between the lengths a, c and b, d. Hence, for the table to not wobble we must have
a − c = b − d ⇐⇒ a − b = c − d ⇐⇒ a + d = b + c.
SMT 2012 Advanced Topics Test and Solutions February 18, 2012

We can therefore split into cases based on S = a + d = b + c. The number of ordered pairs (x, y)
such that x + y = S and 1 ≤ x, y ≤ 8 is TS = 8 − |S − 9| (similar to adding the values on two
8-sided dice). The number of choices for (a, d) is therefore TS and the number of choices for
(b, c) is TS , so the number of choices for (a, b, c, d) is TS2 .
Summing over all possible values of S this is

T22 + . . . T16
2
= (8 − |2 − 9|)2 + · · · + (8 − |16 − 9|)2
= 12 + 22 + · · · + 72 + 82 + 72 + · · · + 22 + 12
= 2 12 + · · · + 72 + 82

7 · 8 · 15
= 2· + 82
6
= 7 · 8 · 5 + 82
= 8 (7 · 5 + 8) .

Hence, the probability is

8 (7 · 5 + 8) 7·5+8 43
= = .
84 83 512

6. Two ants are on opposite vertices of a regular octahedron (an 8-sized polyhedron with 6 vertices,
each of which is adjacent to 4 others), and make moves simultaneously and continuously until
they meet. At every move, each ant randomly chooses one of the four adjacent vertices to move
to. Eventually, they will meet either at a vertex (that is, at the completion of a move) or on an
edge (that is, in the middle of a move). Find the probability that they meet on an edge.
2
Answer: 11
Solution: If the two ants are not on the same vertex, they can either be on opposite vertices
or on adjacent vertices. Let x and y be the probabilities that the ants will eventually meet on
an edge when starting out from opposite vertices and from adjacent vertices, respectively. From
opposite vertices, one of the ants must move to one of the remaining four vertices, which are
all equivalent with respect to the other ant. That ant can either meet the first ant at a vertex,
become adjacent to it (two ways to do this), or again become opposite from it. So
1 1
x = x + y.
4 2

If the two ants are adjacent, the cases become slightly more complicated. If the first ant moves
towards the second ant, the second ant can move towards it (meeting on an edge); otherwise
they will be adjacent. If the first ant moves away from the second ant, they will become adjacent
no matter what the second ant does. If the first ant moves to the side (two ways to do this),
they will be opposite if the second ant chooses the other direction, and will meet at a vertex if
it chooses the same direction. Otherwise they will be adjacent. So
1 11 1
y = x+ y+ .
8 16 16

2
This system of equations is easily solved to obtain x = .
11
 SMT 2012 Advanced Topics Test and Solutions February 18, 2012

7. Determine the greatest common divisor of the elements of the set {n13 − n | n is an integer}.
Answer: 2730
Solution: Let D be the desired greatest common divisor. By Fermat’s Little Theorem we have:
n13 ≡ (n6 )2 (n) ≡ (n3 )2 (n) ≡ n4 ≡ n2 ≡ n mod 2.
Hence 2 | (n13 − n) for all n, so 2 | D. Similarly we can show that p | D for p ∈ {3, 5, 7, 13}.
Since these are all prime, their product, 2730, divides D.
213 − 2 = 8190 = 3(2730), so D is either 2730 or 3(2730). As 313 − 3 = 3(312 − 1) is not divisible
by 9, D = 2730 .

8. We say that a set of positive integers S, all greater than

  1, covers
  an integer x if for every pair
k l
of integers k and l such that 2 ≤ k < l ≤ x, we have 6= for at least one integer y in S.
y y
How many numbers are in the smallest set S which covers 30?
Answer: 10
Solution: We claim that the set of all primes less than 30 is the smallest set S which covers
30. We first prove that no smaller set can exist; assume that one does exist. This smaller set
cannot contain some prime p < 30. Note that, therefore, p and p + 1 are indistinguishable. This
is a contradiction. We must now show that the set of all primes less than 30 is valid. Since two
consecutiveintegers
 are relatively prime, if some prime p divides k, it does not divide k + 1.
k+1 k
Therefore, = + 1. Therefore, the set of all primes less than k will always cover k. It
p p
remains to compute the number of primes less than 30. There are 10 primes less than 30.

9. Evaluate ∞
P 1 P∞ 1 2
n=1 n3 (n+1)3 . You can use Euler’s result n=1 n2 = π /6.

Answer: 10 − π 2
Solution: We use partial fractions repeatedly to obtain that
 3
1 1 1
= −
n (n + 1)3
3 n n+1
 
1 1 1 1 1
= − −3 −
n3 (n + 1)3 n(n + 1) n n + 1
 2
1 1 1 1
= − −3 −
n3 (n + 1)3 n n+1
1 1 3 3 6
= 3
− 3
− 2− 2
+
n (n + 1) n (n + 1) n(n + 1)
1 1 3 3 6 6
= 3
− 3
− 2− 2
+ − .
n (n + 1) n (n + 1) n n+1

Then by taking sums and using the property of telescoping sums we have
∞
π2 π2
 
X 1
= 1 − 3 −3 − 1 + 6 = 10 − π 2 .
n3 (n + 1)3 6 6
n=1

10. We say that two polynomials F (x) and G(x) are equivalent mod 5 if and only if F (x) − G(x) =
5 · H(x) for some integer polynomial H(x). We say that F (x) has n as a root mod 5 if and only
SMT 2012 Advanced Topics Test and Solutions February 18, 2012

if 5 | F (n). How many inequivalent integer polynomials mod 5 of degree at most 3 do not have
any integer roots mod 5?
Answer: 204
Solution: Observe that a polynomial

Ia (X) = 1 − (X − a)p−1

takes value 1 at a and 0 elsewhere in mod p, by Fermat’s little theorem. Thus for any polynomial
F mod p, we have
p−1
X
F (n) = F (a)Ia (n) (mod p)
a=0

for all n. Now the polynomial of degree ≤ p − 1

p−1
X
F (X) − F (a)Ia (X)
a=0

has 0, 1, · · · , (p − 1) as roots, thus it should be zero mod p. This means that polynomials mod
p of degree less than p have one-to-one correspondence to p-tuples of (F (0), F (1), · · · , F (p − 1))
mod p. Since F not having any roots is equivalent to that none of F (a) is zero, there are (p − 1)p
ways to choose (F (0), F (1), · · · , F (p − 1)). This gives the answer to the first part.
Pp−1
For the second part, note that coefficient of X p−1 in
P
a=0 F (n)Ia (X) is − F (a), so it is
equivalent toP find number of p-tuples (F (0), F (1), · · · , F (p − 1)) satisfying F (a) 6= 0 (mod p)
for all a and F (a) = 0 (mod p). We define

An = the number of n tuples (a1 , · · · , an ) satisfying

1 ≤ ai ≤ p − 1, p | a1 + · · · + an

and the problem is to find Ap . We establish the recurrence relation on An . For the initial
condition we have A1 = 0 and A2 = p − 1. For n > 2, note that if (a1 , a2 , · · · , an ) is counted in
An , then an is uniquely chosen to be ≡ −(a1 + · · · + an−1 ) (mod p) unless a1 + · · · + an−1 is not
divisible by p. This is equivalent to say that An is same as the number of (n − 1)-tuples with
their sum not divisible by p. This gives the recurrence

An = (p − 1)n−1 − An−1

and by solving it we have

An = (p − 1)n−1 − (p − 1)n−2 + (p − 1)n−3 − · · · + (−1)n−2 (p − 1).

(p−1)np +(−1)p−1 (p−1) (p−1)p −(p−1) 45 −4
So the answer is Ap = (p−1)+1 = p . Evaluating at p = 5, we get 5 =
1020
5 = 204 .
SMT 2012 Advanced Topics Tiebreaker and Solutions February 18, 2012

1. Let p(n) be the smallest digit that is part of the decimal writing of a natural number n.
Compute p(100) + p(101) + p(102) + . . . + p(998) + p(999).
Answer: 2025
Solution:

p(000) + p(001) + ... + p(999) = (103 − 93 ) × 0 + (93 − 83 ) × 1 . . . + (23 − 13 ) × 8 + (13 − 03 ) × 9

= 93 + 83 + 73 + 63 + 53 + 43 + 33 + 23 + 13
= 2025 .

2. We define n to be a squarefree integer if, for every prime p, p2 does not divide n. Let f (n) be the
sum of the reciprocals of all the divisors of n. We define n to be an amazing integer if f (n) = 2.
How many squarefree amazing integers are there?
Answer: 1
k
σ(n) Y pi + 1
Solution: Note that f (n) = = . In order for f (n) to be integral, the only primes
n pi
i=1
that can divide n are 2 and 3, since those are the only consecutive primes. Note that f (6) = 2,
but 1, 2, and 3 do not satisfy this constraint. Therefore, there is exactly 1 squarefree amazing
integer.

3. There are 7 cages in a row in an animal shelter and an ample supply of three different kind of
animals: dogs, cats, and golden bears. Since golden bears do not like each other, they cannot be
placed in adjacent cages. Cats also do not like each other (but not as much so as golden bears),
so there cannot be more than two cats in a row. How many ways are there to fill the cages with
animals?
Answer: 1002
Solution: Let a1n denote the number of ways to fill up n cages with the last cage being a golden
bear. Let b1n denote the number of ways with the last cage being a cat and the next-to-last not
a cat. Let b2n denote the number of ways with the last two cages being both cats. Let cn denote
the number of ways with the last cage being a dog. Then we have the recurrence



 a1n = b1n−1 + b2n−1 + cn−1
b1 = a1 + c

n n−1 n−1
b 2 = b1
n n−1



c = a1 + b1 + b2 + c

n n−1 n−1 n−1 n−1 .

Initial values are a1 = b1n = cn = 1 while b2n = 0. Following the recurrence until n = 7, we obtain
a17 = 276, b17 = 250, b27 = 95, c7 = 381, so the total is 1002 .
SMT 2012 Team Test and Solutions February 18, 2012

1. How many functions f : {1, 2, 3, 4, 5} → {1, 2, 3, 4, 5} take on exactly 3 distinct values?

Answer: 1500
Solution: There are 53 possibilities for the range, so the answer is 10N where N is the

number of surjective functions from {1, 2, 3, 4, 5} to a given 3-element
set. The total number
5 3
of functions {1, 2, ..., 5} → {1, 2, 3} is 3 , from which we subtract 2 (the number of 2-element
subsets of {1, 2, 3}) times 25 (the number of functions mapping into
that subset), but then
3
(according to the Principle of Inclusion-Exclusion) we must add back 1 (the number of functions
mapping into a 1-element subset of {1, 2, 3}). Thus: N = 35 − 32 (25 ) + 31 (15 ) = 150. So

10N = 10(150) = 1500 .

2. Let i be one of the numbers 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11. Suppose that for all positive integers
n, the number nn never has remainder i upon division by 12. List all possible values of i.
Answer: 2, 6, 8, 10
Solution: The table below gives the value of k n (mod 12) for k = 0, 1, . . . , 11 and n = 1, 2, 3.
(Note that when n = 1, this is just the value of k.)
k n=2 n=3
0 0 0
1 1 1
2 4 8
3 9 3
4 4 4
5 1 5
6 0 0
7 1 7
8 4 8
9 9 9
10 4 4
11 1 11
All values of k have period 1 or 2 (the rows for 2 and 8 continue 4, 8, 4, 8, etc.) We can see that
nn cannot be congruent to 2, 6, 10 when divided by 12 for n > 1, and hence cannot be congruent
to 2, 6, 10 at all. For nn to be congruent to 8 (mod 12), we would need either n ≡ 2 (mod 12)
and n ≡ 1 (mod 2) or n ≡ 8 (mod 12) and n ≡ 1 (mod 2); this is impossible since a number
which is 2 or 8 (mod 12) must be even. All other remainders indeed occur; this can be checked
by inspection, with the help of the Chinese Remainder Theorem. So our answer is 2, 6, 8, 10 .

3. A card is an ordered 4-tuple (a1 , a2 , a3 , a4 ) where each ai is chosen from {0, 1, 2}. A line is an
(unordered) set of three (distinct) cards {(a1 , a2 , a3 , a4 ), (b1 , b2 , b3 , b4 ), (c1 , c2 , c3 , c4 )} such that
for each i, the numbers ai , bi , ci are either all the same or all different. How many different lines
are there?
Answer: 1080
Solution: There are 34 = 81 different cards. Any choice of two cards determines a unique line,
since if we know ai and bi , ci must equal ai , bi if ai = bi and must equal neither if ai 6= bi . This
81


produces 2 lines, but for each line there are 3 pairs of cards which generate that line, so our
final answer is 13 81


2 = 27 · 40 = 1080 .
SMT 2012 Team Test and Solutions February 18, 2012

4. We say that the pair of positive integers (x, y), where x < y, is a k-tangent pair if we have
1 1 1
arctan = arctan + arctan . Compute the second largest integer that appears in a 2012-
k x y
tangent pair.
Answer: 811641
Solution: By taking tangents of both sides we have

1 1/x + 1/y y+x

= = ,
k 1 − 1/xy xy − 1

so xy = k(x + y) + 1, (x − k)(y − k) = k 2 + 1. For k = 2012 the second largest factor of k 2 + 1

is (k 2 + 1)/5 = 809629 and thus the second largest integer y is k + 809629 = 811641 .

5. Regular hexagon A1 A2 A3 A4 A5 A6 has side length 1. For i = 1, . . . , 6, choose Bi to be a point

on the segment Ai Ai+1 uniformly at random, assuming the convention that Aj+6 = Aj for all
integers j. What is the expected value of the area of hexagon B1 B2 B3 B4 B5 B6 ?
√
9 3
Answer: 8
Solution 1: By symmetry, E[(Bi Ai+1 Bi+1 )] = E[(Bj Aj+1 Bj+1 )] for all integers i and j. There-
fore, applying linearity of expectation, the expected area of B1 B2 B3 B4 B5 B6 is equal to the area
of A1 A2 A3 A4 A5 A6 minus six times the expected area of B1 A2 B2 . Since the lengths of B1 A2 and
B2 A2 are independent, this expectation is equal to ( 21 sin 120◦ )E[B1 A2 ]E[B2 A2 ]. It is easy to
√
3
see that E[B1 A2 ] = E[B2 A2 ] = 1/2, so E[(A1 B2 A2 )] = 16 . The area of a unit regular hexagon
√
√ √
3
√
3 9 3
is 6( 3/4), so our answer is 6( 4 − 16 ) = .
8
Solution 2: Since the area of B1 B2 B3 B4 B5 B6 is linear in the location of Bi for each i, and the
Bi are all independent, we can argue that the average case comes when each Bi is a midpoint
of Ai Ai+1 .

6. Evaluate
∞ X
∞
X 1
.
nm(n + m + 1)
n=1 m=1

Answer: 2
Solution 1: Using the partial fraction
 
1 1 1 1
= −
m(n + m + 1) n+1 m n+m+1

we can sum the series in m first to get

∞ X
∞ ∞  
X 1 X 1 1 1 1
= + + ···
nm(n + m + 1) n(n + 1) 1 2 n+1
n=1 m=1 n=1
X 1
= .
kn(n + 1)
k≤n+1
SMT 2012 Team Test and Solutions February 18, 2012

For k = 1, this sum has a term for each n ≥ 1, and for larger values of k, it has a term for each
n ≥ k − 1. Then we can rewrite this as
∞ ∞ ∞  
X 1 X 1 X 1 1
= + −
n(n + 1) k n n+1
n=1 k=2 n=k−1
∞
X 1
=1+ =1+1= 2 .
k(k − 1)
k=2

Solution 2: Consider the Taylor expansion

∞
x2 x3 X xn
− log(1 − x) = x + + + ··· = .
2 3 n
n=1

We square this equation to obtain

∞
X xn+m
2
(log(1 − x)) = .
nm
n,m=1

Integrating both sides from 0 to 1 will give the answer. We make the substitution

x = 1 − ey , 0 ≥ y > −∞

to get Z 1 Z −∞ Z ∞
2
(log(1 − x)) dx = y 2
(log e ) (−e dy) = y
y 2 e−y dy = 2
0 0 0
upon integrating by parts twice.

7. A plane in 3-dimensional space passes through the point (a1 , a2 , a3 ), with a1 , a2 , and a3 all
positive. The plane also intersects all three coordinate axes with intercepts greater than zero
(i.e. there exist positive numbers b1 , b2 , b3 such that (b1 , 0, 0), (0, b2 , 0), and (0, 0, b3 ) all lie on
this plane). Find, in terms of a1 , a2 , a3 , the minimum possible volume of the tetrahedron formed
by the origin and these three intercepts.
9
Answer: a a a
2 1 2 3
Solution: Let the x, y, and z intercepts of the plane be b1 , b2 , and b3 , respectively. The
tetrahedron in question has volume 16 b1 b2 b3 . The equation of our plane is bx1 + by2 + bz3 = 1, since
these three intercepts determine the plane. Therefore, we are minimizing 16 b1 b2 b3 subject to the
constraint ab11 + ab22 + ab33 = 1. By AM-GM, we get that
 a1 a2 a3 3
a1 a2 a3 b1 + b2 + b3 1
≤ = .
b1 b2 b3 3 27
Therefore, b1 b2 b3 ≥ 27a1 a2 a3 with equality if
a1 a2 a3 1
= = = ,
b1 b2 b3 3
27a1 a2 a3
which is attained by choosing bi = 3ai . Hence, the desired minimum volume is 6 =
9
a1 a2 a3 .
2
SMT 2012 Team Test and Solutions February 18, 2012

8. The left end of a rubber band e meters long is attached to a wall and a slightly sadistic child
holds on to the right end. A point-sized ant is located at the left end of the rubber band at time
t = 0, when it begins walking to the right along the rubber band as the child begins stretching
it. The increasingly tired ant walks at a rate of 1/(ln(t + e)) centimeters per second, while the
child uniformly stretches the rubber band at a rate of one meter per second. The rubber band
is infinitely stretchable and the ant and child are immortal. Compute the time in seconds, if it
exists, at which the ant reaches the right end of the rubber band. If the ant never reaches the
right end, answer +∞.
100
Answer: ee −e
Solution: Let x(t) be the position of the ant at time t, l(t) be the length of the rubber band at
1
time t, and va (t) = 100 log(t+e) be the ant’s walking speed relative to the rubber band. We have
dx
dt = va (t) + x · dl/dt
l(t) , where the second term comes from the stretching of the rubber band (the
rate of change in length divided by the length gives the rate of stretching of an infinitesimal piece
of rubber band; multiplying by x(t) gives the contribution to the ant’s forward motion from the
dx d
part of the rubber band behind the ant). This is just  dt = va (t) + x dt (log l(t)), and multiplying
− log l(t) d x va (t) x(t) R t va (s)
both sides by e and rearranging, we have dt l(t) = l(t) , or l(t) = 0 l(s) ds (applying
R t a (s) Rt
initial conditions). We need 0 vl(s) 1
ds = 100 1 1 t
0 (s+e) log(s+e) ds = 100 [log log(s + e)]0 to equal 1,
100
which occurs at time t = ee −e .

9. We say that two lattice points are neighboring if the distance between them is 1. We say that a
point lies at distance d from a line segment if d is the minimum distance between the point and
any point on the line segment. Finally, we say that a lattice point A is nearby a line segment
if the distance between A and the line segment is no greater than the distance between the line
segment and any neighbor of A. Find the number of lattice points that are nearby the line
segment connecting the origin and the point (1984, 2012).
Answer: 1989
Solution: For notational convenience, let ` be the line passing through the origin and (1984, 2012).
First, note that a point P can only be nearby the given line segment if its x-coordinate is between
0 and 1984, inclusive. If the x-coordinate of P is negative, its distance to any point on the line
segment is less than the distance between that point on the line segment and the point one unit
to the right of P ; if P has x-coordinate greater than 1984, take the point one unit to the left.
This inequality holds due to the creation of a right or obtuse angle between P , the point next
to P , and any point on the line segment (the edge case where the three points are collinear
remains, but this is easily checked separately).
Now, fix a value x ∈ {0, 1, . . . , 1984}, and let S be the set of lattice points with that x coordinate.
Let Q be the point on the line segment with this x-coordinate. Note that for any P ∈ S, the
distance between P and the line segment is either the distance from P to ` or the distance from
P to one of the endpoints of the line segment. In the latter case, since there always exists a
cardinal direction to move closer to a given point, P is not nearby our segment. Now consider
the former case. By similar triangles, the distance between P and ` is proportional to P Q, and
so the only P which could possibly be nearby the segment are the P closest to Q. There are
two such points if the y-coordinate of Q has fractional part 1/2, and one such point otherwise.
Finally, we show that all such points are in fact nearby: this relies on the fact that the slope of `
is greater than 1. Consider a point P that is no further from ` than any other lattice point with
 SMT 2012 Team Test and Solutions February 18, 2012

the same x-coordinate. We already know that its distance to Q, the point on the line with same
x-coordinate, is less than or equal to 1/2. Now draw the horizontal line through P , intersecting
` at R. Since the slope of ` is greater than 1, we have P R < P Q ≤ 1/2, and so P is, out of all
lattice points with the same y coordinate, the closest one to `. Hence, it is a nearby point.
Now, we just need to count the number of nearby points. There are 1985 different valid choices
of x-coordinate, and we must double-count all the ones for which the point on ` with that
x-coordinate has y-coordinate with real part 1/2. Since ` is given by
2012 503
y= x= x,
1984 496
this condition holds when x ≡ 248 (mod 496), so there are 4 such values in the relevant interval.
Hence, report 1989 .
 
i
10. A permutation of the first n positive integers is valid if, for all i > 1, i comes after in
2
the permutation. What is the probability that a random permutation of the first 14 integers is
valid?
1
Answer: 31752
1
Solution 1: 1 must be the first number in the permutation; this happens with probability .
14
1
2 must come before 4, 5, 8, 9 10, and 11; this happens with probability .
7
1
3 must come before 6, 7, 12, 13, and 14; this happens with probability .
6
1
4 must come before 8 and 9; this happens with probability .
3
1
5 must come before 10 and 11; this happens with probability .
3
1
6 must come before 12 and 13; this happens with probability .
3
1
7 must come before 14; this happens with probability .
2
All these events are independent, so the answer is the product of the above probabilities, or
1
.
31752
Solution 2: Create a directed graph with vertices labeled 1 through 14, with an arrow from
vertex a to vertex b if b has to come after a in a valid permutation. The graph looks like this: 1
points to 2 and 3, 2 points to 4 and 5, 3 points to 6 and 7, 4 points to 8 and 9, 5 points to 10 and
11, 6 points to 12 and 13, 7 points to 14. We want to count the number of ways to merge this
graph
into a line. There are two ways of ordering 12 and 13 WRT 6. Once this is fixed, there
are 53 = 10 ways of ordering the 6 − 12 − 13 group and the 7 − 14 group WRT 3. Similarly,
there are 2 ways of ordering 8, 9 WRT 4, 2 ways of ordering 10, 11 WRT 5, and 63 = 20 ways

of ordering the 4 − 8 − 9 group and the 5 − 10 − 11 group WRT 2. Finally, there are 13


6
ways
of ordering the 2 . . . group and the 3 . . . group WRT 1. This gives 2 · 10 · 2 · 2 · 20 · 13
6 valid
permutations, and dividing by 14! gives the answer.
 SMT 2012 Team Test and Solutions February 18, 2012

11. Given that x, y, z > 0 and xyz = 1, find the range of all possible values of

x3 + y 3 + z 3 − x−3 − y −3 − z −3
.
x + y + z − x−1 − y −1 − z −1

Answer: (27, +∞)

Solution: We make the following modifications to the numerator. Since xyz = 1, we may
multiply x−3 , y −3 , z −3 by x3 y 3 z 3 , and also add −1 + x3 y 3 z 3 . The numerator then factors as
−1 + x3 + y 3 + z 3 − x3 y 3 − x3 z 3 − y 3 z 3 + x3 y 3 z 3 = (x3 − 1)(y 3 − 1)(z 3 − 1). Similarly, the
denominator factors as (x − 1)(y − 1)(z − 1), so that the expression can be rewritten as (x2 + x +
1)(y 2 + y + 1)(z 2+ z + 1). Dividing
 this by xyz and writing z = 1/(xy), we find that this is in
1 1 1
fact 1 + x + x 1 + y + y 1 + xy + xy . We have from (a − 1)2 ≥ 0 that a + 1/a ≥ 2, so the

product has value at least (1 + 2)(1 + 2)(1 + 2) = 27. Equality cannot occur: this would require
x = 1/x, y = 1/y, or x = y = z = 1, making the original denominator zero. Everything larger
than 27 can occur, however: we
simply consider
the special case y = x, when the expression
1 1 2 1


reduces to 1 + x + x 1 + x + x 1 + x + x2 , a continuous function of x which is unbounded
as x → +∞. Thus we answer (27, +∞) .
√ √ √ √
12. A triangle has sides of length 2, 3 + 3, and 2 2 + 6. Compute the area of the smallest
regular polygon that has three vertices coinciding with the vertices of the given triangle.
√
Answer: 12 + 6 3
Solution: We begin by computing the angles of the triangle; repeated application of the Law
of Cosines gives us that the angles of the triangle are 15◦ , 60◦ , and 105◦ . Therefore, since the
greatest common divisor of the angles of the triangle is 15◦ , a dodecagon is the smallest regular
√
polygon that satisfies our current constraints. In particular, the dodecagon has side length 2,
and it πremains
 √ to compute the area of the dodecagon. The area of a regular dodecagon is
√
3 cot 2
( 2) = 12 + 6 3 .
12
13. How many positive integers n are there such that for any natural numbers a, b, we have n |
(a2 b + 1) implies n | (a2 + b)? (Note: The symbol | means “divides”; if x | y then y is a multiple
of x.)
Answer: 20
Solution: Let P represent the property of n such that n | a2 b + 1 ⇒ n | a2 + b for all a, b ∈ N.
Let Q represent the property of n such that (a, n) = 1 ⇒ n | a4 − 1 for all a ∈ N. We shall prove
that they are equivalent.
Proof that P ⇒ Q: Let a be a positive integer with (a, n) = 1. By Bézout’s identity, we can find
b ∈ N such that n | a2 b + 1. By P, n | a2 + b. Then a4 − 1 = a2 (a2 + b) − (a2 b + 1), so n | a4 − 1.
Proof that Q ⇒ P: Let a, b be positive integers with n | a2 b + 1. Clearly (a, n) = 1, so n | a4 − 1.
Then a2 (a2 + b) = (a4 − 1) + (a2 b + 1). Since a and n are relatively prime, n | a2 + b.
Now we wish to find all n with property Q. If a is odd, we have a4 − 1 = (a2 − 1)(a2 + 1), a2 ≡ 1
(mod 8), and a2 + 1 is even, so 16 | a4 − 1. If (a, 3) = 1, we have a2 ≡ 1 (mod 3), so 3 | a4 − 1.
If (a, 5) = 1, we have 5 | a4 − 1 by Fermat’s Little Theorem. This argument shows that n | 240
is sufficient.
To show n | 240 is necessary, suppose n has property Q, and let n = 2a · k, where k is odd. If
k > 1, then (k − 2, n) = 1, so by Q we conclude that n | (k − 2)4 − 1. Then k | (k − 2)4 − 1, but
 SMT 2012 Team Test and Solutions February 18, 2012

(k − 2)4 ≡ (−2)4 ≡ 16 (mod k), so k | 15. Now, since (11, n) = 1, n | 114 − 1, so 2a | 114 − 1,
resulting in a ≤ 4. Thus n | 240 is also necessary.
The number of natural numbers n such that property P holds is simply the number of positive
integer divisors of 240, which is (4 + 1)(1 + 1)(1 + 1) = 20 .

14. Find constants α and c such that the following limit is finite and nonzero:

n
e 1 − n1 − 1
c = lim .
n→∞ nα
Give your answer in the form (α, c).
Answer: (−1, −1/2)
Solution 1: Take x = 1/n and let F (x) = e(1 − x)1/x . Then we have

x x2 x3
   
1/x log(1 − x)
F (x) = e(1 − x) = exp 1 + = exp − − − ···
x 2 3 4

since log(1 − x) has Taylor expansion −x − x2 /2 − x3 /3 − · · · . Especially observe that F (0) = 1,

so when α = −1 the limit can be represented as a derivative of F , as follows:

n
e 1 − n1 − 1 F (x) − F (0) 1
lim = lim = F 0 (0) = − .
n→∞ 1/n x→0 x 2

Solution 2: Observe that if α ≥ 0 then the limit would be zero, so therefore α < 0. This
suggests that we should replace n1 = x, so the limit becomes

e(1 − x)1/x − 1
c = lim .
x→0 x−α
Applying L’Hôpital’s Rule yields
 
1 ln(1−x)
e(1 − x)1/x x2 −x
− x2
c = lim .
x→0 −αx−α−1
2
Observe that ln(1 − x) ≈ −x − x2 , and that limx→0 e(1 − x)1/x , so therefore the numerator is
   
1/x 1 ln(1 − x) 1 ln(1 − x)
lim e(1 − x) − = lim −
x→0 x2 − x x2 x→0 x2 − x x2
2
!
x + x2
   
1 1 1 1 1 1 1
= lim + = lim + + = lim + =− .
x→0 x2 − x x2 x→0 x2 − x x 2 x→0 x − 1 2 2

This is nonzero, so therefore our desired α is α = −1, with limit −1/2, and our final answer is
(−1, −1/2) .

15. Sean thinks packing is hard, so he decides to do math instead. He has a rectangular sheet that
he wants to fold so that it fits in a given rectangular box. He is curious to know what the optimal
size of a rectangular sheet is so that it’s expected to fit well in any given box. Let a and b be
positive reals with a ≤ b, and let m and n be independently and uniformly distributed random
variables in the interval (0, a). For the ordered 4-tuple (a, b, m, n), let f (a, b, m, n) denote the
SMT 2012 Team Test and Solutions February 18, 2012

ratio between the area of a sheet with dimension a×b and the area of the horizontal cross-section
of the box with dimension m × n after the sheet has been folded in halves along each dimension
until it occupies the largest possible area that will still fit in the box (because Sean is picky, the
sheet must be placed with sides parallel to the box’s sides). Compute the smallest value of ab
that maximizes the expectation f .
√ 2
Answer: e−1+ 1+2 ln(2)
Solution: First, note that for fixed a, b, m, and n, f (a, b, m, n) = f (a, b, m/2, n) = f (a, b, m, n/2)
because we can go from an optimal sheet folding in one case to an optimal sheet folding in an-
other case by either folding or unfolding the sheet in half, which scales the sheet’s folded area
by the same amount as the box’s cross-sectional area. This implies that we can tile the region
m ∈ (0, a), n ∈ (0, a) with an infinite number of rectangles for which the expectation of f is the
same inside each rectangle. In particular, the set R of these rectangles is the set of all rectangles
bounded by points of the form ( 2ax , 2ay ) and ( 2x+1
a a
, 2y+1 ) in mn-space for nonnegative integers x
and y. The expectation of f over the whole region is the same as the expectation of f inside
any one of these rectangles. Let us choose to examine the rectangle where we have m ∈ [a/2, a),
n ∈ [a/2, a). Clearly, it is optimal to fold the sheet exactly once in the dimension where the
sheet has length a. For the other dimension, we may assume that b ∈ [a, 2a) because if b ≥ 2a,
we are forced to fold it anyways until b < 2a. Now, we only have to fold in the b dimension
once if b/2 < max(m, n), and otherwise we must fold twice. Therefore, the expected value of f
is equal to
Z a Z a
4 (a/2)(b/2)(1/2)H(b/2−max(m,n))
dm dn,
a2 a/2 a/2 mn

where H(x) is the Heaviside Step Function (defined as H(x) = 1 if x ≥ 0 and 0 otherwise).
We can compute this integral by noticing that it is equal to
!
Z a Z a Z b/2 Z b/2
4 ab/4 ab/8
dm dn − dm dn .
a2 a/2 a/2 mn a/2 a/2 mn

Hence, this problem relies on evaluating

Z dZ d Z d
1 ln(d) − ln(c)
dx dy = dy = (ln(d) − ln(c))2 = ln(d/c)2 .
c c xy c y

Plugging in, we get that the original integral equals

4 b
2
((ab/4) ln(2)2 − (ab/8) ln(b/a)2 ) = (2 ln(2)2 − ln(b/a)2 ).
a 2a

Let k = b/a, so that the expectation we are maximizing is g(k) = k2 (2 ln(2)2 − ln(k)2 ) over the
domain k ∈ [1, 2). The derivative of g is
  
1 2 2 −2 ln(k) k 1
(2 ln(2) − ln(k) ) + = − (ln(k)2 + 2 ln(k) − 2 ln(2)2 ),
2 k 2 2

which we set to zero, getting

p
−2 ± 4 + 8 ln(2)2 p
ln(k) = = −1 ± 1 + 2 ln(2)2 .
2
SMT 2012 Team Test and Solutions February 18, 2012

We now need to check two final things. First, we must see if either of these solutions to g 0 (k) = 0
are in the interval (1, 2). Only the positive solution to the above quadratic could possibly result
in k being in greater than 1 to begin with. The easiest way to see that this gives us a value in
the interval (1, 2) is by noticing that g(1) = g(2) = ln(2)2 (since the k = 1 and k = 2 cases both
result in the folded sheet having the same area, namely a/2 × a/2, for all choices of m and n),
so we are guaranteed a point with zero derivative in the interval (1, 2) by Rolle’s Theorem.
Additionally, we must check that this is a local maximum and not a minimum. We claim here
that g(k) is concave down on the interval (1, 2), so what we have found is a local maximum.
First, ln(k)2 is concave up on (1, 2) because

d2 2 − 2 ln(k)
2
ln(k)2 = ,
dk k2
which is positive when ln(k) < 1 ⇐⇒ k < e. Hence, 2 ln(2)2 − ln(k)2 is concave down in the
same interval. It is also clearly decreasing. Finally, we have for generic functions f and g that if
f (x) = xg(x), then f 0 (x) = g(x) + xg 0 (x) and f 00 (x) = 2g 0 (x) + xg 00 (x), so on an interval where g
is decreasing and concave down and x is positive, then f is guaranteed to also be concave down.
This scenario holds for our function g(k), so it is concave down.
√ 2
Hence, report e−1+ 1+2 ln(2) .
SMT 2012 Power Round and Solutions February 18, 2012

Introduction
Shuffling a deck of playing cards is a very important life skill. The standard riffle shuffle goes like this: you
take a stack of cards, split it into two piles, hold one pile in your left hand and the other in your right, and
drop cards from each hand onto a common pile on the table in random order. After repeating this several
times, your card deck is hopefully fairly well-mixed and ready for a game or maybe a magic trick. (Of course,
in practice, the easiest way to perform the riffle shuffle is to allow the cards to interleave without dropping
them, but this is mathematically equivalent.)
People often perform only three or four riffle shuffles in a row before using a standard deck of 52 playing
cards. One might ask if this is really enough. It’s not hard to see that after one riffle shuffle, many orderings
of the cards are impossible to reach and many others are much more likely than they ought to be in a
uniformly random probability distribution. How many shuffles in a row do you really need to approximate
uniform randomness? This Power Round builds up some of the basic ideas you need to answer this question.
In order to understand how to mathematize shuffling, we first discuss the basic concept of a permutation
and some specific properties that we will need. Next, we give a mathematically precise definition of the
Gilbert-Shannon-Reeds riffle shuffle, which has been shown in experiments to be a good model for how real
people shuffle, and develop the theory of the probabilities it generates. Unfortunately, actually computing
the necessary number of shuffles for approximate uniform randomness is beyond the scope of this test,
but hopefully you will come to believe that such a number is indeed computable. Finally, we develop the
mathematics of the perfect shuffle, a deterministic shuffle very useful in magic tricks to those able to perform
it. Have fun!
SMT 2012 Power Round and Solutions February 18, 2012

Permutation Enumeration
One of the key tools that we will use to analyze shuffles is the permutation. A permutation of a set S is
defined as a listing of elements of S in some order (with each element appearing precisely once); for example,
permutations of S = {1, 2, 3, 4, 5} include (4, 2, 3, 5, 1) or (3, 2, 1, 4, 5).

1. (a) [2] List all permutations of {1, 2, 3}.

(b) [2] Give an expression for the number of permutations of {1, 2, 3, . . . , n} in terms of n. Compute
the number for n = 5.
Solution to Problem 1:

(a) (1, 2, 3), (1, 3, 2), (2, 1, 3), (2, 3, 1), (3, 1, 2), (3, 2, 1)
(b) n!. 120.

We can also think of a permutation as an operation we perform on some ordered listing to get another
ordered listing. For instance, with S = {1, 2, 3}, we can think of the permutation (2, 1, 3) as the operation
of swapping the first and second elements in an ordered listing of three elements, and leaving the third in
place. In general, we interpret a permutation (σ(1), σ(2), . . . , σ(n)) as the operation that sends the σ(1)th
element to the first position, the σ(2)th element to the second position, and so on. The listings we previously
wrote down are just what we get when we apply the permutation to (1, 2, . . . , n). Note that the permutation
whose listing is itself (1, 2, . . . , n) corresponds to doing nothing at all—for this reason, we call it the identity
permutation, and write it as 1.
Given this interpretation, we define the notions of composition and inverse. The composition σ ◦ τ of
two permutations σ and τ is the operation of performing τ first, then σ. The inverse σ −1 of a permutation
σ is the permutation such that σ −1 ◦ σ = 1.

2. (a) [2] Compute the composition σ ◦ τ of permutations σ = (1, 5, 4, 3, 6, 2) and τ = (2, 4, 6, 3, 1, 5).
(b) [2] Compute the inverse of (3, 1, 4, 2) and the inverse of (2, 4, 6, 3, 1, 5).
(c) [2] Show that (σ ◦ τ )−1 = τ −1 ◦ σ −1 for all permutations σ and τ of {1, 2, . . . , n}.
Solution to Problem 2:

(a) (2, 1, 3, 6, 5, 4)
(b) (2, 4, 1, 3), (5, 1, 4, 2, 6, 3).
(c) We can show that τ −1 ◦ σ −1 is the inverse of (σ ◦ τ ) as follows:

(τ −1 ◦ σ −1 ) ◦ (σ ◦ τ ) = τ −1 ◦ (σ −1 ◦ σ) ◦ τ = τ −1 ◦ τ = 1.

When talking about shuffles, both of these interpretations of permutations have a natural meaning.
The listing interpretation corresponds to a state of the deck, and the process interpretation corresponds to
shuffling the deck form one state to another.

3. [5] Suppose that a process shuffles a deck of σ into τ . Which permutation will be produced when
(1, 2, . . . , n) is shuffled by that process? Justify.
Solution to Problem 3: τ ◦ σ −1 . Since a permutation µ changes σ to µ ◦ σ, µ should be τ ◦ σ −1 in
order for µ ◦ σ to be τ .

So far, we’ve talked about permutations as deterministic processes that always produce the same result
on the same input. But in real shuffling, people usually don’t produce the same result every time (unless
they’re trained magicians!). Thus, we will mainly focus on shuffling processes that are random processes,
i.e. different outcomes occur with certain probabilities. The probability that a random shuffle turns a deck
σ into a deck τ is called the transition probability from σ to τ .
SMT 2012 Power Round and Solutions February 18, 2012

4. [5] For any random shuffle, show that the transition probability from σ to τ is the same as the transition
probability from 1 to τ ◦ σ −1 .
Solution to Problem 4: The probability that a random shuffle takes σ to µ ◦ σ depends only on
µ, since changing the labels on the cards does not affect the shuffle. So we can replace σ by 1 and,
according to problem 3, therefore replace τ by µ = τ ◦σ −1 without changing the transition probability.
Now, let’s return to permutations. An ascent of a permutation σ of {1, 2, . . . , n} is any position 1 ≤ i < n
such that σ(i) < σ(i + 1). For example, the permutation (2, 7, 1, 3, 5, 4, 8, 6) has ascents at positions 1, 3, 4, 6.
Similarly, a descent of a permutation σ is any position where σ(i) > σ(i + 1). In our example, the descents
occur at positions 2, 5, 7. Note that every position i < n is either an ascent or a descent.
5. (a) [2] List the ascents and descents of (9, 2, 7, 6, 3, 1, 8, 4, 5).
(b) [2] Compute the number of permutations of {1, 2, 3} with exactly one descent.
(c) [3] There are 11 permutations of {1, 2, 3, 4} with exactly two ascents. List them.
No explanations required.
Solution to Problem 5:
(a) Ascents at 2, 6, 8; descents at 1, 3, 4, 5, 7.
(b) 4. They are (1, 3, 2), (2, 1, 3), (2, 3, 1), (3, 1, 2).
(c) (1, 2, 4, 3), (1, 3, 2, 4), (1, 3, 4, 2), (1, 4, 2, 3), (2, 1, 3, 4), (2, 3, 1, 4), (2, 3, 4, 1), (2, 4, 1, 3), (3, 1, 2, 4),
(3, 4, 1, 2), (4, 1, 2, 3).
Define the Eulerian number nk as the number of permutations of {1, 2, . . . , n} with k ascents. For



example, as given in the preceding problem, 42 = 11.



6. [5] Prove the symmetry property of Eulerian numbers:

   
n n
= .
k n−k−1

Solution to Problem 6: We find a bijection (one-to-one correspondence) between the permu-

tations with k ascents and the permutations with n − k − 1 ascents. Indeed, if a permutation
σ = (σ(1), σ(2), . . . , σ(n)) has k ascents, then it has n − k − 1 descents because each position i < n is
either an ascent or a descent. Reversing the permutation swaps ascents and descents, so therefore the
permutation (σ(n), σ(n − 1), . . . , σ(1)) has n − k − 1 ascents and k descents.
7. [8] Prove that the Eulerian numbers satisfy the recurrence
     
n n−1 n−1
= (k + 1) + (n − k) .
k k k−1

Solution to Problem 7: Consider any permutation σ of {1, 2, . . . , n} with k ascents. We have

σ(i) = n for some 1 ≤ i ≤ n, and removing this σ(i) yields a permutation σ 0 of {1, 2, . . . , n − 1} with
either k or k − 1 ascents.
Every permutation of {1, 2, . . . , n} with k ascents is therefore built from a permutation of {1, 2, . . . , n −
1} with k or k − 1 ascents by inserting n. There are now two cases.
Given a permutation of {1, 2, . . . , n − 1} with k − 1 ascents, we gain an ascent by inserting n only when
we do so at a descent or at the end of the permutation. There are n − k − 1 descents, so each such
permutation produces n − k permutations of {1, 2, . . . , n} with k ascents.
Similarly, given a permutation of {1, 2, . . . , n − 1} with k ascents, we want to preserve the number
of ascents when inserting n. To do this, the insertion must happen at one of the k ascents, or at
the beginning of the permutation. Each such permutation therefore produces k + 1 permutations of
{1, 2, . . . , n} with k ascents.
Combining these two cases yields the desired recurrence.
SMT 2012 Power Round and Solutions February 18, 2012

n
8. [5] Using the recurrence for Eulerian numbers, compute a table of Eulerian numbers. Include k for
1 ≤ n ≤ 6, 0 ≤ k ≤ 6.
Solution to Problem 8:

n 
n 
n 
n 
n 
n
n
n 0 1 2 3 4 5 6
1 1 0
2 1 1 0
3 1 4 1 0
4 1 11 11 1 0
5 1 26 66 26 1 0
6 1 57 302 302 57 1 0

9. [12] Prove Worpitzky’s Identity:

n   
X n x+k
xn = .
k n
k=0

To ensure that the binomial coefficient makes sense, assume that x is an integer and x ≥ n.1
Solution

0 x
to Problem 9: We prove this by induction. Firstly, notice that in the case n = 0, we have
0 0 = 1 = x0 .
Pn

Now, assume that xn = k=0 nk x+k


n . We will use this to prove Worpitzky’s identity for n + 1. We
compute using the recurrence for Eulerian numbers:
n+1
X   n+1    n+1   
n+1 x+k X n x+k X n x+k
= (k + 1) + (n − k + 1)
k n+1 k n+1 k−1 n+1
k=0 k=0 k=0
n    n   
X n x+k X n x+k+1
= (k + 1) + (n − k)
k n+1 k n+1
k=0 k=0
n     
X n x+k x+k−n x+k+1
= (k + 1) + (n − k)
k n n+1 n+1
k=0
n 
X n x+k     n   
xn + x X n x+k
= =x = x · xn = xn+1 .
k n n+1 k n
k=0 k=0

This completes the induction.

A rising sequence of a permutation σ is a maximal sequence of consecutive numbers appearing as a

subsequence of (not necessarily adjacent entries of) σ. (Here, “maximal” means that we cannot add more
numbers to the rising sequence; (1, 2) and (1, 2, 3) cannot both be rising sequences.) Every permutation
decomposes into disjoint rising sequences. For example, the permutation (6, 1, 2, 4, 7, 5, 3) decomposes into
three rising sequences: (1, 2, 3), (4, 5), and (6, 7). Here, (1, 2, 3) is a rising sequence of (6, 1, 2, 4, 7, 5, 3)
because the numbers 1, 2, 3 appear in order in σ but 1, 2, 3, 4 do not.

10. [8] Recall the definition of the inverse of a permutation from the text before problem 2. Show that
the number of rising sequences of a permutation σ is equal to one more than the number of descents
of σ −1 . That is, show

#{rising sequences of σ} = #{descents of σ −1 } + 1.

Solution to Problem 10: If k and k + 1 are in the same rising sequence, then their positions σ −1 (k)
and σ −1 (k + 1) must satisfy σ −1 (k) < σ −1 (k + 1). This means that k + 1 starts a new rising sequence
if σ −1 (k) > σ −1 (k + 1), i.e. if σ −1 has a descent at position k. Also, there is an additional rising
sequence that starts at 1; this does not correspond to a descent. Hence, the number of rising sequences
of σ is one more than the number of descents of σ −1 .
1 This
`a´
actually works in greater generality. We can define generalized binomial coefficients b
for any real number a and
nonnegative integer b, and Worpitzky’s identity holds in this more general context.
SMT 2012 Power Round and Solutions February 18, 2012

The Gilbert-Shannon-Reeds shuffle

As remarked in the introduction, the Gilbert-Shannon-Reeds (GSR) shuffle is a mathematical model which
has been shown in experiments to fit the way real people shuffle real card decks. Here we will develop this
model and a number of its interesting properties.
First we introduce some standard notation. Let j1 , j2 , . . . , ja be nonnegative integers so that j1 + j2 +
· · · + ja = n. We define  
n n!
= .
j1 , j2 , . . . , ja j1 !j2 ! · · · ja !
This number is called a multinomial
coefficient. Of course, when a = 2, this is just a binomial coefficient,
which we will usually refer to as jn1 .

11. Compute (no explanations required):

7


(a) [2] 3,2,2 ,
8


(b) [2] 2,2,2,2 , and
100


(c) [2] 99,1,0,0,0 .
Solution to Problem 11:
(a) 210
(b) 2520
(c) 100

The standard GSR shuffle works like this. Take a deck of n cards and cut it into a left pile and a right
pile containing the bottom x cards and top y cards
respectively (so x + y = n), in such a manner that the
probability of putting x cards into the left pile is nx /2n . Drop cards from the bottom of either the left or
the right pile one at a time, in such a manner that if at any point you’re holding X cards on the left and Y
cards on the right, the probability that the next card dropped comes from the left is X/(X + Y ).

12. Take a stack of three cards labeled 1, 2, 3 from bottom to top and apply the GSR shuffle once. Consider
the resulting pile, from bottom to top, as a permutation of 1, 2, 3.
(a) [2] Are any permutations impossible to get? If so, list them.
(b) [2] Compute the probability of putting (i) 0, (ii) 1, (iii) 2, (iv) 3 cards into the left pile during
the cut.
(c) [2] Compute the probability of the final permutation being (i) 3, 1, 2, (ii) 1, 2, 3.
No explanations required.
Solution to Problem 12:
(a) Yes, 3, 2, 1 only.
(b) (i) 1/8, (ii) 3/8, (iii) 3/8, (iv) 1/8.
(c) (i) 1/8, (ii) 1/2.

13. (a) [3] In the general case with n cards, why do the given probabilities of cutting 0, 1, . . . , n cards
(n) (n) (n)
into the left pile always actually add up to 1? That is, show that 20n + 21n + · · · + 2nn = 1.
(b) [3] Take a standard deck of 52 cards and perform one GSR shuffle. Show that the probability of
cutting 0 cards into one of the piles is less than one in one trillion (10−12 ).
Solution to Problem 13:
(a) Standard. Can be done by either quoting binomial theorem or providing combinatorial explana-
tions.
SMT 2012 Power Round and Solutions February 18, 2012

2 1 1 1
(b) 252 < 248 < 1612 < 1012 .

Now we’re ready to describe the GSR a-shuffle, which is exactly like the standard GSR shuffle except
with a piles. That is, take your deck of n cards, cut it into piles of size j1 , . .
. , ja with j1 + · · · + ja = n
n 1
so that the probability of getting precisely those sizes in that order is j1 ,...,j a a
n (we will refer to this as

the cutting stage), and drop cards from the piles, one at a time, so that whenever you are holding piles of
size J1 , . . . , Ja respectively, the probability of dropping the next card from the kth pile is Jk /(J1 + · · · + Ja )
(this is the dropping stage). In the future, we will consistently assume the following: 1. The cards start out
numbered 1 to n from bottom to top. 2. The order of the cards after the dropping stage, from bottom to
top, will be considered as a permutation of 1, 2, . . . , n.

14. (a) [2] Take a 4-card deck and perform one 3-shuffle. Compute the probability that after the cutting
stage, the pile sizes will be 1, 1, 2 in some order.
(b) Now suppose the same 4-card deck has already been cut into piles of size 1, 1, 2 from left to right
(so the leftmost pile has the card numbered 1, the middle pile has card 2, and the rightmost pile
has cards 3 and 4). Perform the dropping stage.
(i) [2] How many permutations of 1, 2, 3, 4 are possible results?
(ii) [2] Compute the probability (given this initial cut) that the final permutation is 2, 3, 4, 1.
(iii) [2] Compute the probability that it is 3, 2, 4, 1.
No explanations required.
Solution to Problem 14:
(a) 4/9
(b) (i) 12 (ii) 1/12 (iii) 1/12

15. (a) [5] Prove that the probabilities we’ve given for every possible way to cut the cards during the
cutting stage really do add up to 1.
(b) [5] Take an n-card deck which has already been cut into a piles of size j1 , . . . , ja . After the
dropping stage, how many permutations of 1, . . . , n are possible? Justify.
(c) [7] Prove that, given this initial cut, every permutation of 1, . . . , n which is possible after the
dropping stage occurs with equal probability. Show that therefore every possible path of operation,
from deck to cut piles to final final permutation, occurs with probability exactly 1/an . Conclude
that the transition probability of the GSR a-shuffle from 1 to σ is the same as the number of
paths leading to σ divided by an . Refer to the definitions from after problem 4.
Solution to Problem 15:
(a) Standard. Possible solutions are: quoting multinomial theorem, explaining combinatorial mean-
ing, or working with induction on a.
n


(b) j1 ,...,j a
. The order of the cards within each pile cannot change, so they can be considered
indistinguishable.
(c) Induct on n. Clear when all piles 0. Given some possible permutation, suppose WLOG the first
card comes from pile 1. The probability of this permutation is then jn1 times the probability of
the permutation with the first card
removed given piles
n−1 n

of size j1 − 1, j2 , . . . , ja , which by the
inductive hypothesis is 1/ j1 −1,...,j a
. This is 1/ j1 ,...,ja as desired.
(Alternatively, directly compute that regardless of dropping order, the numerator must be j1 ! · · · ja !
and the denominator must be n!.)

We will now describe a few apparently different shuffles which turn out to be the GSR a-shuffle in disguise,
or related. The diversity of these descriptions shows just how mathematically rich the GSR shuffle is!

16. (a) [6] A “maximum entropy a-shuffle” is any shuffle in which you cut an n-card deck into a (possibly
empty) piles and then drop cards from the piles one by one, with the stipulation that every
possible path from deck to piles to final permutation should be equally likely. Prove that
SMT 2012 Power Round and Solutions February 18, 2012

(i) the GSR a-shuffle satisfies this property and

(ii) the only way to satisfy this property is to use the same probabilities as in the GSR a-shuffle.
(b) [6] A “sequential a-shuffle” works as follows. First you cut an n-card deck into a piles according
to
 the GSR probability distribution (i.e. getting piles of size j1 , . . . , ja occurs with probability
n
). Then you shuffle pile 1 and 2 together using the dropping stage of the standard
j1 , . . . , ja
GSR 2-shuffle. Having done this, you shuffle the combined pile with pile 3, take the result and
shuffle with pile 4, and so on until you have only one pile left. Prove that the probability of
getting any particular permutation at the end is the same as with the standard a-shuffle.
(c) An “inverse a-shuffle” works as follows. Take your n-card deck and, dealing from the bottom, place
each card on one of a piles uniformly at random (that is, choose each pile with probability 1/a).
Once you’re done, stack the piles together in order from left to right.
(i) [2] Prove that any possible path of operation reachable by an inverse a-shuffle—from deck
to randomly dealt piles to final permutation—appears with probability 1/an .
(ii) [2] Show that inverse a-shuffle is not equivalent to the standard a-shuffle in general by
exhibiting a permutation of 4 cards reachable by an inverse 2-shuffle which is not reachable
by a standard 2-shuffle. You do not need to justify.
(iii) [7] Show that the transition probability from σ to τ of the inverse a-shuffle is the same as
the transition probability τ → σ of the standard a-shuffle. Refer to the definitions from after
problem 4.
Solution to Problem 16:
(a) (i) Clear from earlier calculations.
n


(ii) Given a deck cut into piles of size j1 , . . . , ja , there are necessary j1 ,...,j a
outcomes which
must be equally likely.
Therefore the probability of getting piles of size j1 , . . . , ja must be
n
proportional to j1 ,...,j a
. Or it is just enough to note that probability of each individual
path determines the probability for each middle stage of process tree.
(b) It is enough to show that the sequential a-shuffle also satisfies the property of 15(c): every possible
permutation occurs with equal probability given the initial cut, as it characterizes the maximum
entropy a-shuffle.
Consider the relative location of cards in pile 1 and 2. After shuffling pile 1 and 2 together, order
of cards within those piles do not change anymore. So shuffling of pile 1 and 2 together should be
uniquely determined if final permutation is given. Similarly we can show that in each 2-shuffles
in the sequence should follow certain path to reach the final permutation. Thus the probability
for all permutations should be same.
(c) (i) Clear.
(ii) 2413 is unique answer.
(iii) Both for the standard a-shuffle and inverse a-shuffle, we showed that the transition proba-
bility of a permutation is the number of possible paths to the permutation divided by an .
Hence it suffices to show that the path from σ to the set of piles P to τ exists under the
inverse a-shuffle if and only if the path from τ to P to σ exists under the standard a-shuffle.
In an inverse shuffle, the path (σ, P, τ ) is possible if and only if merging P gives the target
permutation τ and the cards in each pile of P are in order within σ. But in the standard
shuffle, the path (τ, P, σ) is possible if and only if merging P gives the original permutation
τ and the cards in each pile of P are in order within σ. The two conditions coincide exactly.
17. (a) [7] Prove that an inverse a-shuffle followed by an inverse b-shuffle gives rise to permutations with
the same probabilities as an inverse ab-shuffle. (This is called the product rule.)
(b) [15] Explain why this property of an a-shuffle followed by a b-shuffle being the same as an ab-
shuffle must also hold when carrying out the standard (AKA maximal entropy) and sequential
forms of the GSR shuffle. Justify rigorously.
Solution to Problem 17:
SMT 2012 Power Round and Solutions February 18, 2012

(a) Label each card with two numbers according to the piles it landed in during the a-shuffle and the
b-shuffle. Those cards with the same label form a pile in an ab-shuffle.
(b) Let Pa (σ → τ ) and P a (σ → τ ) be the transition probabilities from σ to τ for the standard a-
shuffle and inverse a-shuffle respectively. The probability of obtaining τ from σ after an a-shuffle
and a b-shuffle is given by X
Pa (σ → µ)Pb (µ → τ )
µ

where the sum is taken over all permutations µ. Meanwhile problem 16(c) gives Pa (σ → µ) =
P a (µ → σ), so this sum is the same as µ P a (µ → σ)P b (τ → µ). Now this can be interpreted
P
as the probability of obtaining σ from τ after an inverse b-shuffle and inverse a-shuffle, and
according to 17(a) this is the same as P ba (τ → σ). Applying problem 16(c) again gives P ba (τ →
σ) = Pab (σ → τ ), so we are done.

18. (a) [7] Suppose σ is a permutation with r rising sequences. Prove that the transition probability from
1 to σ for GSR a-shuffle of an n-card deck is
a+n−r


n
.
an
(b) [3] Use this to give another proof of Worpitzky’s identity.
(c) [7] Use part a of this problem and Problem 17 to show that if we repeat an a-shuffle k times on the
same deck, the probability of any one permutation σ appearing after the last shuffle approaches
1/n! as k approaches infinity.
Solution to Problem 18:
(a) We need to count the number of different ways to cut the deck into piles which have σ as a possible
resulting permutation. We will make a − 1 cuts which can be in any of n + 1 locations in the
deck. The rising sequences determine r − 1 of these cuts, but the remaining a − r can be assigned
arbitrarily. This gives a+n−r


n ways in which the deck can be cut. There are an possible final
permutations (counting repeats), giving the desired probability.
(b) Immediate.
ak +n−r 1


(c) The probability is n akn
where r is the number of rising sequences of σ. This is
 k   k 
a +n−r a +1−r 1
··· ·
ak ak n!

and each multiplicative factor can be rewritten as 1 + n−i−r

ak
, which approaches 1 as k approaches
∞. Thus the whole expression approaches 1/n!.

Using the GSR model and some analysis too advanced to explain here, one can show that an n-card
deck must be shuffled at least 23 log2 n times before the probability distribution of the resulting permutations
begins to approach uniformly random. This number is 7 for a 52-card normal playing deck and 9 for an
81-card SET deck. We here at the Stanford Math Tournament consider it very important that you take this
knowledge into account the next time you play a card game!

Perfect shuffles
So far, we’ve been discussing a method of shuffling which aims to make the resulting permutation random.
For certain people, magicians for example, it is more important to discuss forms of shuffling which are
perfectly predictable. Here we analyze the interesting mathematics behind one such shuffle.
In a slight departure from the notation of the previous section, we will work with a deck of 2n cards
which starts out numbered 0, 1, . . . , 2n − 1 from bottom to top. (We will refer to the location of the bottom
card as the 0th position in the deck, and so on.) There are two perfect riffle shuffles, the out-shuffle O
and the in-shuffle I. In both shuffles, you cut the deck exactly in half and alternate dropping a card from
SMT 2012 Power Round and Solutions February 18, 2012

each half. O drops the original bottom card first and leaves the original top card on the top, whereas I
drops the original bottom card second and leaves the card originally at the top second from the top. For
example, applying O to 0, 1, . . . , 2n − 1 gives 0, n, 1, n + 1, 2, n + 2, . . . , n − 1, 2n − 1, whereas applying I to
0, 1, . . . , 2n − 1 gives n, 0, n + 1, 1, n + 2, 2, . . . , 2n − 1, n − 1. We will refer to the permutations obtained by
applying O and I to a0 , a1 , . . . , a2n−1 as O(a0 , a1 , . . . , a2n−1 ) and I(a0 , a1 , . . . , a2n−1 ) respectively; thus we
might say that O(0, 1, . . . , 2n − 1) = 0, n, 1, n + 1, 2, n + 2, . . . , n − 1, 2n − 1. We will refer to the permutation
obtained by applying O to a0 , a1 , . . . , a2n−1 k times as Ok (a0 , a1 , . . . , a2n−1 ), and similarly for I.
The order of a shuffle on 2n cards is the least positive number of times you must apply that shuffle to
0, 1, . . . , 2n − 1 before getting 0, 1, . . . , 2n − 1 back.

19. Compute (no explanation needed)

(a) [2] I(O(I(0, 1, 2, 3, 4, 5))),

(b) [2] the order of O on 8 cards, and
(c) [2] Ok (0, 1, 2, 3, 4, 5, 6, 7) for all k ≥ 1.

Solution to Problem 19:

(a) 5, 3, 4, 1, 2, 0
(b) 3
(c) k ≡ 0 (mod 3) : 0, 1, 2, 3, 4, 5, 6, 7
k ≡ 1 (mod 3) : 0, 4, 1, 5, 2, 6, 3, 7
k ≡ 2 (mod 3) : 0, 2, 4, 6, 1, 3, 5, 7

20. (a) [3] Prove that after one out-shuffle of 2n cards, the card numbered j has moved to position 2j
(mod 2n − 1).
(b) [3] Prove that the order of the in-shuffle on 2n cards is the same as the order of the out-shuffle
on 2n + 2 cards.
(c) [2] Prove that the order of the out-shuffle on 2n cards is the least positive integer k such that
2k ≡ 1 (mod 2n − 1).
(d) [2] Compute the order of the out-shuffle on 52 cards.
Solution to Problem 20:

(a) For j ≤ n − 1, j moves to 2j, immediately before j + n. That is, j + n moves to 2j + 1 =

2(j + n) − (2n − 1) ≡ 2(j + n) (mod 2n − 1).
(b) Add a card to the beginning and end; now each in-shuffle is just an out-shuffle with two ghost
cards that never move.
(c) From part a, it is evident that after k out-shuffles, the card numbered j has moved to position 2k j
(mod 2n − 1). Thus, the order of the out-shuffle on this deck is the least positive integer k so that
j ≡ 2k j (mod 2n − 1) for all j, i.e. the least positive integer k such that 2k ≡ 1 (mod 2n − 1).
(d) We have a deck of 52 = 2n cards, where n = 26 (so 2n − 1 = 51). Therefore the order of the
out-shuffle on this deck is the least positive integer k such that 2k ≡ 1 (mod 51), and we can
quickly check that this integer k is 8.

21. (a) [3] Take a deck of 2m cards and number them as usual. Prove that if a card’s number has
binary representation am−1 am−2 . . . a0 , then after one out-shuffle, that card has moved to position
am−2 . . . a0 am−1 .
(b) [3] What do m in-shuffles do to 2m cards? Justify.
Solution to Problem 21:

(a) This follows directly from j → 2j (mod 2m − 1).

SMT 2012 Power Round and Solutions February 18, 2012

(b) Performing m in-shuffles on a deck of 2m cards will reverse the order of the cards in the deck. We
can think of these in-shuffles as out-shuffles on 2m + 2 cards, where the top and bottom cards are
“dummy” cards that never move. Then, the jth card goes to 2m j ≡ −j (mod 2m + 1).

22. [8] Given a deck of 2n cards numbered as usual and k ∈ {0, 1, . . . , 2n−1}, state and prove an algorithm
consisting only of in- and out- shuffles for bringing the card numbered 0 to the kth position in the
deck. (Hint: consider the binary expansion of k.)
Solution to Problem 22: Interpret 1 as in and 0 as out in the aforementioned binary expansion and
perform the resulting operations from left to right. You start with the card numbered 0 and the first
in-shuffle takes it to position 1. Note that if your card is at position j, for j not too large, an out-shuffle
takes it to 2j and an in-shuffle takes it to 2j + 1—so each operation pushes the binary expansion of j to
the left and adds the appropriate 0 or 1 in the units digit. This process will never result in a position
greater than 2n − 1, so you need not worry about modding out by 2n − 1.