We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 148
102
COMBINATORIAL
PROBLEMS
FROM THE TRAINING OF
THE USA IMO TEAM
Titu Andreescu
and
Zuming FengPreface
This book contains one hundred and two highly selected problems
used in the training and testing of the USA International Math-
ematical Olympiad (IMO) team. It is not a collection of very
difficult, impenetrable questions. Instead, the book gradually builds
students’ combinatorial skills and techniques. This work aims to
broaden students’ view of mathematics and better prepare them
for possible participation in various mathematical competitions. It
provides in-depth enrichment in important areas of combinatorics
by reorganizing and enhancing students’ problem-solving tactics and
strategies. The book further stimulates students’ interest for future
study of mathematicsIntroduction
In the United States of America, the selection process leading
to participation in the International Mathematical Olympiad (IMO)
consists of a series of national contests called the American Math-
ematics Contest 10 (AMC 10), the American Mathematics Contest:
12 (AMC 12), the American Invitational Mathematics Examination
(AIME), and the United States of America Mathematical Olympiad
(USAMO). Participation in the AIME and the USAMO is by in-
vitation only, based on performance in the preceding exams of the
sequence. The Mathematical Olympiad Summer Program (MOSP)
is a four-week intensive training program for around one hundred
very promising students who have risen to the top in the American
Mathematics Competitions. The six students representing the United
States of America in the IMO are selected on the basis of their
USAMO scores and further testing that takes place during MOSP
Throughout MOSP, full days of classes and extensive problem sets
give students thorough preparation in several important areas of
mathematics. These topics include combinatorial arguments and
identities, generating functions, graph theory, recursive relations,
sums and products, probability, number theory, polynomials, theory
of equations, complex numbers in geometry, algorithmic proofs, com-
binatorial and advanced geometry, functional equations, and classical
inequalities.
Olympiad-style exams consist of several challenging essay problems.
Correct solutions often require deep analysis and careful argument
Olympiad questions can seem impenetrable to the novice, yet most
can be solved with elementary high school mathematics techniques,
vvi INTRODUCTION
cleverly applied.
Here is some advice for students who attempt the problems that
follow.
¢ Take your time! Very few contestants can solve all the given
problems
¢ Try to make connections between problems. An important theme
of this work is: all important techniques and ideas featured in the
book appear more than once!
.
Olympiad problems don’t “crack” immediately. Be patient. Try
different approaches. Experiment with simple cases. In some
cases, working backward from the desired result is helpful.
.
Even if you can solve a problem, do read the solutions. They
may contain some ideas that did not occur in your solutions,
and they may discuss strategic and tactical approaches that can
be used elsewhere. The solutions are also models of elegant
presentation that you should emulate, but they often obscure
the torturous process of investigation, false starts, inspiration,
and attention to detail that led to them. When you read the
solutions, try to reconstruct the thinking that went into them.
Ask yourself, “What were the key ideas?” “How can I apply these
ideas further?”
.
Go back to the original problem later, and see if you can solve it
in a different way. Many of the problems have multiple solutions,
but not all are outlined here.
.
Meaningful problem solving takes practice. Don’t get discouraged
if you have trouble at first. For additional practice, use the books
on the reading listAcknowledgments
‘Thanks to Po-Ling Loh, Po-Ru Loh, and ‘Tim Perrin who helped with
typesetting, proofreading and preparing solutions.
Many problems are either inspired by or adapted from mathematical
contests in different countries and from the following journals:
High-School Mathematics, China
Revista Matematica Timigoara, Romania
We did our best to cite all the original sources of the problems in the
solution section. We express our deepest appreciation to the original
proposers of the problems.
viiiAbbreviations and Notations
Abbreviations
AHSME American High School Mathematics Examination
AIME American Invitational Mathematics Examination
AMC10 American Mathematics Contest 10
AMC12 American Mathematics Contest 12, which replaces AHS)
ARML American Regional Mathematics League
IMO International Mathematical Olympiad
USAMO United States of America Mathematical Olympiad
MOSP Mathematical Olympiad Summer Program
Putnam The William Lowell Putnam Mathematical Competition
St. Petersburg St. Petersburg (Leningrad) Mathematical Olympiad
for Numerical Sets and Fields
the set of integers
the set of integers modulo n
the set of positive integers
the set of nonnegative integers
the set of rational numbers
the set of positive rational numbers
the set of nonnegative rational numbers
the set of n-tuples of rational numbers:
the set of real numbers
the set of positive real numbers
the set of nonnegative real numbers
the set of n-tuples of real numbers
the set of complex numbers
the coefficient of the term 2” in the polynomial p(«)ABBREVIATIONS and NOTATIONS
xi
Notations for Sets, Logic, and Geometry
Al
ACB
ACB
A\B
ANB
AUB
aceA
the number of elements in set A
Ais a proper subset of B
Ais asubset of B
A without B
the intersection of sets A and B
the union of sets A and B
the element a belongs to the set Axii
Contents
PREFACE ....
INTRODUCTION .......
ACKNOWLEDGEMENTS
ABBREVIATIONS AND NOTATIONS .
1. INTRODUCTORY PROBLEMS .
2. ADVANCED PROBLEMS ....
3. SOLUTIONS TO INTRODUCTORY PROBLEMS . 22
4. SOLUTIONS TO ADVANCED PROBLEMS .
GLOSSARY
FURTHER READING1
Introductory ProblemsIntroductory Problems
. Mr. and Mrs. Zeta want to name their baby Zeta so that its
monogram (first, middle, and last initials) will be in alphabetical
order with no letters repeated. How many such monograms are
possible?
The student lockers at Olympic High are numbered consecutively
beginning with locker number 1. ‘The plastic digits used to
number the lockers cost two cents apiece. Thus, it costs two cents
to label locker number 9 and four cents to label locker number
10. If it costs $137.94 to label all the lockers, how many lockers
are there at the school?
Let n be an odd integer greater than 1. Prove that the sequence
(1). @)()
contains an odd number of odd numbers
How many positive integers not exceeding 2001 are multiples of
3 or 4 but not 5?
. Let
a = .123456789101112...998999,
where the digits are obtained by writing the integers 1 through
999 in order. Find the 1983"¢ digit to the right of the decimal
point.
. Twenty five boys and twenty five girls sit around a table. Prove
that it is always possible to find a person both of whose neighbors
are girls.
. At the end of a professional bowling tournament, the top 5
bowlers have a play-off. First #£5 bowls #4. The loser receives
5¢” prize and the winner bowls #3 in another game. The loser
of this game receives receives 4" prize and the winner bowls #2.
The loser of this game receives 3°¢ prize and the winner bowls
#1. The winner of this game gets 1* prize and the loser gets 2°
prize. In how many orders can bowlers #1 through #5 receive
the prizes?
A spider has one sock and one shoe for each of its eight legs.
In how many different orders can the spider put on its socks and
shoes, assuming that, on each leg, the sock must be put on before
the shoe?Introductory Problems 3
9.
10
ll
13
14
16.
. In a room containing }
A drawer in a darkened room contains 100 red socks, 80 green
socks, 60 blue socks and 40 black socks. A youngster selects
socks one at a time from the drawer but is unable to see the color
of the socks drawn, What is the smallest number of socks that
must be selected to guarantee that the selection contains at least
10 pairs? (A pair of socks is two socks of the same color. No sock
may be counted in more than one pair.)
Given a rational number, write it as a fraction in lowest terms
and calculate the product of the resulting numerator and denom-
inator. For how many rational numbers between 0 and | will 20!
be the resulting product?
Determine the number of ways to choose five numbers from the
first eighteen positive integers such that any two chosen numbers
differ by at least 2
people, N > 3, at least one person has
not shaken hands with everyone else in the room. What is the
maximum number of people in the room that could have shaken
hands with everyone else?
Find the number of ordered quadruples (1,
odd integers that satisfy a; + 2%2+¢3 +a4=
rg, a4) of positive
8.
Finitely many cards are placed in two stacks, with more cards in
the left stack than the right. Each card has one or more distinct
names written on it, although different cards may share some
names. For each name, we define a shuffle by moving every card
that has that name written on it to the opposite stack. Prove
that it is always possible to end up with more cards in the right
stack by picking several distinct names, and doing in turn the
shuffle corresponding to each name
For how many pairs of consecutive integers in the set
{1000, 1001, 1002, ..., 2000}
is no carrying required when the two integers are added?
Nine chairs in a row are to be occupied by six students and
Professors Alpha, Beta, and Gamma. These three professors
arrive before the six students and decide to choose their chairs
so that each professor will be between two students. In how
many ways can Professors Alpha, Beta, and Gamma choose their17.
18
19
21
22.
23
24
Introductory Problems
chairs?
Prove that among any 16 distinct positive integers not exceeding
100 there are four different ones, a,b, ¢,d, such that a+b = c+,
A child has a set of 96 distinct blocks. Each block is of one
of 2 materials (plastic, wood), 3 sizes (small, medium, large), 4
colors (blue, green, red, yellow), and 4 shapes (circle, hexagon,
square, triangle). How many blocks in the set are different from
the “plastic medium red circle” in exactly two ways? (The “wood
medium red square” is such a block.)
Call a 7-digit telephone number djdgd3 — dadsdgd7 memorable if
the prefix sequence didedg is exactly the same as either of the
sequences dadsdg or dsdedz (possibly both). Assuming that each
d; can be any of the ten decimal digits 0,1,2,...,9, find the
number of different memorable telephone numbers
Two of the squares of a 7 x 7 checkerboard are painted yellow,
and the rest are painted green. Two color schemes are equivalent
if one can be obtained from the other by applying a rotation in
the plane of the board. How many inequivalent color schemes are
possible?
In how many ways can one arrange the numbers 21, 31, 41, 51, 61,
71, and 81 such that the sum of every four consecutive numbers
is divisible by 3?
Let S be a set with six elements. In how many different ways can
one select two not necessarily distinct subsets of S' so that the
union of the two subsets is S? The order of the selection does
not matter; for example the pair of subsets {a,c}, {b,c,d,e, f}
represents the same selection as the pair {b,c,d,e, f}, {a,c}.
A set of positive numbers has the triangle property if it has
three distinct elements that are the lengths of the sides of a
triangle whose area is positive. Consider sets {4,5,6,...,n} of
consecutive positive integers, all of whose ten-element subsets
have the triangle property. What is the largest possible value
of n?
Let A and B be disjoint sets whose union is the set of natural
numbers. Show that for every natural number n there exist
distinct a,b > n such that
{a,batb}CA or {aba +b} CB.Introductory Problems 5
25. The increasing sequence 1,3,4,9,10,12,13,... consists of all
re
those positive integers which are powers of 3 or sums of distinct
powers of 3. Find the 100" term of this sequence (where | is the
1 term, 3 is the 2"¢ term, and so on)
Every card in a deck has a picture of one shape — circle, square,
or triangle, which is painted in one of three colors — red, blue,
or green. Furthermore, each color is applied in one of three
shades — light, medium, or dark. The deck has 27 cards, with
every shape-color-shade combination represented. A set of three
cards from the deck is called complementary if all of the following
statements are true:
(a) Hither each of the three cards has a different shape or all
three of the cards have the same shape.
(b) Either each of the three cards has a different color or all three
of the cards have the same color.
(c) Either each of the three cards has a different shade or all
three of the cards have the same shade.
How many different complementary three-card sets are there?
At a math camp, every m students share exactly one common
friend, m > 3. (If A is a friend of B, then B is a friend of A.
Also, A is not his own friend.) Suppose person P has the largest
number of friends. Determine what that number is.
Suppose that 7 boys and 13 girls line up in a row. Let S
be the number of places in the row where a boy and a girl
are standing next to each other. For example, for the row
GBBGGGBGBGGGBGBGGBGG we have S = 12. Find the
average value of S (if all possible orders of these 20 people are
considered)
A bored student walks down a hall that contains a row of closed
lockers, numbered 1 to 1024. He opens the locker numbered 1, and
then alternates between skipping and opening each closed locker
thereafter. When he reaches the end of the hall, the student
turns around and starts back. He opens the first closed locker
he encounters, and then alternates between skipping and opening
cach closed locker thereafter. The student continues wandering
back and forth in this manner until every locker is open. What
is the number of the last locker he opens?30.
31.
32.
35.
38.
39.
Introductory Problems
Let n = 2813!°. How many positive integer divisors of n? are less
than n but do not divide n?
In an arena, each row has 199 seats. One day, 1990 students are
coming to attend a soccer match. It is only known that at most
39 students are from the same school. If students from the same
school must sit in the same row, determine the minimum number
of rows that must be reserved for these students.
Let T = {9* | k is an integer, 0 < k < 4000}. Given that 94°
has 3817 digits and that its first (leftmost) digit is 9, how many
elements of T have 9 as their leftmost digit?
For what values of n > 1 do there exist a number m that
can be written in the form a, +--+ + dp (with a; € {1},a. €
{1,2},...,an € {1,.-.,n}) in (n — 1)! or more ways?
Let the sum of a set of numbers be the sum of its elements. Let
S be a set of positive integers, none greater than 15. Suppose no
two disjoint subsets of S have the same sum. What is the largest
sum a set S with these properties can have?
There are at least four candy bars in n (n > 4) boxes. Each time,
Mr. Fat. is allowed to pick two boxes, take one candy bar from
each of the two boxes, and put those candy bars into a third box.
Determine if it is always possible to put all the candy bars into
one box.
Determine, with proof, if it is possible to arrange 1,2,..., 1000
in a row such that the average of any pair of distinct numbers is
not located in between the two numbers.
Let A) Ay... Ao be a regular dodecagon with O as its center
Triangular regions OA; Aj41, 1 < i < 12 (and Ais = Ai) are to
be colored red, blue, green, or yellow such that adjacent regions
are colored in different colors. In how many ways can this be
done?
There are 2n people at a party. Each person has an even number
of friends at the party. (Here friendship is a mutual relationship.)
Prove that there are two people who have an even number of
common friends at the party.
How many different 4 x 4 arrays whose entries are all 1’s and
—I's have the property that the sum of the entries in each row is
0 and the sum of the entries in each column is 07Introductory Problems 7
40
Al.
42
44.
45.
A square of dimensions (n — 1) x (n — 1) is divided into (n — 1)?
unit squares in the usual manner. Each of the n? vertices of these
squares is to be colored red or blue. Find the number of different
colorings such that each unit square has exactly two red vertices.
(Two coloring schemes are regarded as different if at least one
vertex is colored differently in the two schemes.)
Sixty-four balls are separated into several piles. At each step we
are allowed to apply the following operation. Pick two piles, say
pile A with p balls and pile B with q balls and p > q, and then
remove q balls from pile A and put them in pile B. Prove that it
is possible to put all the balls into one pile
A game of solitaire is played with a finite number of nonnegative
integers. On the first move the player designates one integer as
large, and replaces another integer by any nonnegative integer
strictly smaller than the designated large integer. On subsequent
steps play is similar, except that integer replaced must be the
one designated as large on the previous play. Prove that in some
finite number of steps play must. end.
. Given $C {1,2,...,n}, we are allowed to modify it in any one
of the following ways:
(a) if 1 ¢ S, add the element 1;
(b) ifne
(c) for 1 {2} 4 --- > {n},
starting with @ and ending with {n}, in which each of the 2”
subsets of {1,2,...,n} appears exactly once. Prove that n =
2™ — 1 for some m
There are 2001 coins on a table. For i = 1,2,...,2001 in
succession, one must turn over exactly i coins. Prove that it
is always possible either to make all of the coins face up or to
make all of the coins face down, but not both
For {1,2,...,n} and each of its nonempty subsets a unique
alternating sum is defined as follows: Arrange the numbers in the46
47.
48
Introductory Problems
subset in decreasing order and then, beginning with the largest,
alternately add and subtract successive numbers. (For example,
the alternating sum for {1,2,4,6,9} is9-6+4-2+1=6 and
for {5} it is simply 5.) Find the sum of all such alternating sums
for n
7
In a game of Chomp, two players alternately take “bites” from
a 5-by-7 grid of unit squares. To take a bite, the player chooses
one of the remaining squares, then removes (“eats”) all squares
found in the quadrant defined by the left edge (extended upward)
and the lower edge (extended rightward) of the chosen square.
For example, the bite determined by the shaded square in the
diagram would remove the shaded square and the four squares
marked by x
(The squares with two or more dotted edges have been removed
from the original board in previous moves.) The object of the
game is to make one’s opponent take the last bite. The diagram
shows one of the many subsets of the set of 35 unit squares that
can occur during the game of Chomp. How many different subsets
are there in all? Include the full board and the empty board in
your count.
Each square of a 1998 x 2002 chess board contains either 0 or
1 such that the total number of squares containing | is odd in
each row and each column. Prove that the number of white unit
squares containing 1 is even
Let S be a subset of {1,2,3,... , 1989} such that no two members
of S differ by 4 or 7. What is the largest number of elements 5
can have?
A class of fifteen boys and fifteen girls is seated around a round
table. Their teacher wishes to pair up the students and hand out
fifteen tests—one test to each pair.Introductory Problems 9
As the teacher is preparing to select the pairs and hand out the
tests, he wonders to himself, “How many seating arrangements
would allow me to match up boy/girl pairs sitting next to each
other without having to ask any student to change his or her
seat?” Answer the teacher’s question. (Two seating arrangements
are regarded as being the same if one can be obtained from the
other by a rotation.)
Two squares on an 8 x 8 chessboard are called touching if they
have at least. one common vertex. Determine if it is possible for
a king to begin in some square and visit all the squares exactly
once in such a way that. all moves except. the first are made into
squares touching an even number of squares already visited
A total of 119 residents live in a building with 120 apartments.
We call an apartment overpopulated if there are at least 15 people
living there. Every day the inhabitants of an overpopulated
apartment have a quarrel and each goes off to a different apart-
ment in the building (so they can avoid each other). Is it true
that. this process will necessarily be completed someday?2
Advanced Problems12 Advanced Problems
1. In a tournament each player played exactly one game against
each of the other players. In each game the winner was awarded
| point, the loser got 0 points, and each of the two players earned
1/2 point if the game was a tie. After the completion of the
tournament, it was found that exactly half of the points earned
by each player were earned in games against the ten players with
the least number of points. (In particular, each of the ten lowest
scoring players earned half of her/his points against the other
nine of the ten). What was the total number of players in the
tournament?
2. Let n be an odd integer greater than 1. Find the number of
permutations p of the set {1,2,...,n} for which
Ip) = 1] + [p(2) = 2| +++ p(n) = nfs
3. In a sequence of coin tosses one can keep a record of the
number of instances when a tail is immediately followed by
a head, a head is immediately followed by a head, etc. We
denote these by TH, HH, etc. For example, in the sequence
HHTTHHHHTHHTTTT of 15 coin tosses we observe that
there are five HH, three HT, two TH , and four TT subse-
quences. How many different sequences of 15 coin tosses will
contain exactly two HH, three HT, four TH and five TT
subsequences?
4, Let A = (a1,a2,...,@2001) be a sequence of positive integers.
Let m be the number of 3-element subsequences (a;,aj;,a,) with
4, for which one can choose
four different numbers a,b,c,d from any n distinct integers such
that a +b —c—d is divisible by 20
. A mail carrier delivers mail to the nineteen houses on the east
side of Elm Street. The carrier notes that no two adjacent housesAdvanced Problems 13
Fro)
10
ll
12
13
ever get mailon the same day, but that there are never more than
two houses in a row that get no mail on the same day. How many
different patterns of mail delivery are possible?
For i= 1,2,..., 11, let M; be a set. of five elements, and assume
that for every 1< i1 satisfies the conditions a; =0,a2 = 1,
1 1
ay = 5 dyn + 500 = Lata (-1)" (1 7Advanced Problems 15
19
20.
24.
n> 3. Determine the explicit form of
n n
fn = an +2(T)n+3( Jans
n n
tohin=D(,%,)ete(,” Jo
For a set A, let |A| and s(A) denote the number of the elements
in A and the sum of elements in A, respectively. (If A = 0, then
|A| = s(A) =0.) Let S be a set of positive integers such that
(a) there are two numbers x,y € S with ged(x,y) = 1;
(b) for any two numbers #,y€S,a+yES.
Prove that
Let T be the set of all positive integers not in
s(T) < |T|? < 0.
In a forest each of 9 animals lives in its own cave, and there
is exactly one separate path between any two of these caves.
Before the election for Forest Gump, King of the Forest, some of
the animals make an election campaign. Each campaign-making
animal—FGC (Forest Gump candidate)—visits each of the other
caves exactly once, uses only the paths for moving from cave to
cave, never turns from one path to another between the caves,
and returns to its own cave at the end of the campaign. It is also
known that no path between two caves is used by more than one
FGC. Find the maximum possible number of FGC’s.
For a sequence Aj,...,An of subsets of {1,...,n} and a permu-
tation m of S = {1,...,n}, we define the diagonal set
Dr(A1, Az,.--5An) = (6 S| ig Any}.
What is the maximum possible number of distinct sets which can
occur as diagonal sets for a single choice of A1,..., An?
A subset M of {1,2,3,...,15} does not contain three elements
whose product is a perfect square. Determine the maximum
number of elements in M
Find all finite sequences (#9,21,...,@n) such that for every j,
0 2, is colored one of two colors (red or blue). Show that
there must be a monochromatic n-term sequence a) < a2 <9 <
ay satisfying
= a1 Sg = 2 S++ S dy = ya
The set {1,2,...,3n} is partitioned into three sets A, B, and CAdvanced Problems Ww
33
34.
35.
36
38.
with each set containing n numbers. Determine with proof if it
is always possible to choose one number out of each set so that
one of these numbers is the sum of the other two.
Assume that each of the 30 MOPpers has exactly one favorite
chess variant and exactly one favorite classical inequality. Each
MOPper lists this information on a survey. Among the survey
responses, there are exactly 20 different favorite chess variants
and exactly 10 different favorite inequalities. Let n be the number
of MOPpers M such that the number of MOPpers who listed M’s
favorite inequality is greater than the number of MOPpers who
listed
c
favorite chess variant. Prove that n > 11
Starting from a triple (a,b,c) of nonnegative integers, a move
consists of choosing two of them, say # and y, and replacing one
of them by either x + y or |x — |. For example, one can go from
(3,5, 7) to (3,5, 4) in one move. Prove that there exists a constant
r > 0 such that whenever a,b,c,n are positive integers with
a,b,¢ <2", there is a sequence of at most rn moves transforming
(a,b,c) into (a’,b',c’) with a’b’c! = 0
A rectangular array of numbers is given. In each row and each
column, the sum of all the numbers is an integer. Prove that each
nonintegral number zx in the array can be changed into either
[x] or |x] so that the row-sums and the column-sums remain
unchanged. (Note that [2] is the least integer greater than or
equal to x, while [a is the greatest integer less than or equal to
x)
A finite set of (distinct) positive integers is called a DS-set if each
of the integers divides the sum of them all. Prove that every finite
t.
set of positive integers is a subset of some DS-s
. Twelve musicians M,, Mz,--- , Miz gather at a week-long cham-
ber music festival. Each day, there is one scheduled concert and
some of the musicians play while the others listen as members
of the audience. For i = 1,2,...,12, let tj be the number of
concerts in which musician M; plays, and let =) +l2+++-+lio.
Determine the minimum value of t such that it is possible for each
musician to listen, as a member of the audience, to all the other
musicians.
An m x n array is filled with the numbers {1,2,...n}, each used
exactly m times. Show that one can always permute the numbers18
40
41
42
43
Advanced Problems
within columns to arrange that each row contains every number
{1,2,...,n} exactly once.
Let set U = {1,2,...,n}, where n > 3. A subset S of U is said to
be split by an arrangement of the elements of U if an element not
in S occurs in the arrangement somewhere between two elements
of S. For example, 13542 splits {1,2,3} but not {3,4,5 }. Prove
that for any n — 2 subsets of U, each containing at least 2 and at
most n—1 elements, there is an arrangement of the the elements
of U which splits all of them
A pile of n pebbles is placed in a vertical column. This config-
uration is modified according to the following rules. A pebble
can be moved if it is at the top of a column which contains at
least two more pebbles than the column immediately to its right
(If there are no pebbles to the right, think of this as a column
with 0 pebbles.) At each stage, choose a pebble from among
those that can be moved (if there are any) and place it at the
top of the column to its right. If no pebbles can be moved, the
configuration is called a final configuration. For each n, show
that, no matter what choices are made at each stage, the final
configuration obtained is unique. Describe that configuration in
terms of n.
Let By be the set of all binary strings of length n. Given two
strings (aj), and (b;)#1, define the distance between the strings
as
d((a;), (65) = Yo ar = 64
k=1
Let C, be a subset of B,. The set Cy is called a perfect
error correcting code (PECC) of length n and tolerance m if for
each string (b;) in B, there is a unique string (¢;) in C, with
d((b;), (ci)) < m. Prove that there is no PECC of length 90 and
tolerance 2.
Determine if it is possible to arrange the numbers 1,1,2,2,...,n,
n such that there are j numbers between two j’s, 1 s; + s; for all nonnegative
integers i,j. Suppose {s,} and {t,} are two superadditive
sequences, and let {u;} be the nondecreasing sequence with the
property that each integer appears in {u,} as many times as in
{sn} and {t,} combined. Show that {u,} is also superadditive.
The numbers from 1 to n?, n > 2, are randomly arranged in the
cells of an nxn unit square grid. For any pair of numbers situated
on the same row or on the same column, the ratio of the greater
number to the smaller one is calculated. The characteristic of
the arrangement is the smallest of these n?(n — 1) fractions.
Determine the largest possible value of the characteristic.
For a set 5, let |,S| denote the number of elements in S. Let A be
a set with |A] =n, and let Ay, Ao,..., An be subsets of A with
Aj] > 2, 1 2k. Let 5 be a nonempty
set of k-clement subsets of {1,...,n} with the property that
every (k + L)-element subset of {1,...,n} contains exactly m
elements of S. Prove that S must contain every k-element subset
of {1,...,n}
A set T is called even if it has an even number of elements. Let n
be a positive even integer, and let $1, .S2,...,S, be even subsets
of the set S = {1,2,...,n}. Prove that there exist i and j,
1 \ n=
(7). ()»
contains an odd number of odd numbers
Solution: The sum of the numbers in the given sequence equals
s[()+Q)e+(tJete-sert
which is an odd number and the conclusion follows.
[AMC12 2001] How many positive integers not exceeding 2001
are multiples of 3 or 4 but not 5?
Solution: For integers not exceeding 2001, there are [2001/3] =
667 multiples of 3 and [2001/4] = 500 multiples of 4. The total,
1167, counts the [2001/12] = 166 multiples of 12 twice, so there
are 1167 — 166 = 1001 multiples of 3 or 4. From these we exclude
the [2001/15] = 133 multiples of 15 and the [2001/20] = 100
multiples of 20, since these are multiples of 5. However, this
excludes the [2001/60] = 33 multiples of 60 twice, so we must
re-include these. The number of integers satisfying the conditions
is 1001 — 133 — 100 + 33 = 801.
. [AHSME 1983] Let
& = .123456789101112...998999,
where the digits are obtained by writing the integers 1 through
999 in order. Find the 1983"¢ digit to the right of the decimal
point.
Solution: Look at the first 1983 digits, and let z denote
the 1983"¢ digit. We may break this string of digits into threeSolutions to Introductory Problems 25
segments:
123456789 1011...9899 lO0101...2.
Eee ee
A B Cc
There are 9 digits in A, 2-90 = 180 in B, hence 1983-189 = 1794
in C. Dividing 1794 by 3 we get 598 with remainder 0. Thus
C consists of the first 598 3-digit integers. Since the first 3+
digit integer is 100 (not 101 or 001), the 598'" 3-digit integer is
598 + 99 = 697. Thus z =7
‘Twenty five boys and twenty five girls sit around a table. Prove
that it is always possible to find a person both of whose neighbors
are girls.
First Solution: For the sake of contradiction we assume that
there is a seating arrangement such that there is no one sitting in
between two girls. We call a block any group of girls(boys) sitting
next to each other and sandwiched by boys(girls) from both sides.
By our assumption, each girl block has at most 2 girls and there
are at least 2 boys in the gap between two consecutive girl blocks.
Hence there are at least [25/2] = 18 girl blocks and at least 2x 13
boys sitting in between the 13 gaps between girls blocks. But we
only have 25 boys, a contradiction. Therefore our assumption was
wrong and it is always possible to find someone sitting between
two girls.
Second Solution: We again approach indirectly by assuming
that there is a seating arrangement such that no one is sitting
in between two girls. We further assume that they are sitting
is positions a), qa2,...,a50 in a counterclockwise order (so aso is
next to a). Now we split them into two tables with seating
orders (a1, @3,@5,...,d49) and (a2,@4,@s,...,@50), each in coun-
terclockwise order. Then by our assumption, no girls are next to
each other in the resulting two seating arrangements. Then there
are at most 12 girls sitting around each new table for a total of
at most 24 girls, a contradiction. Therefore our assumption was
wrong and it is always possible to find someone sitting in between
two girls.
. [AHSME 1988] At the end of a professional bowling tournament,
the top 5 bowlers have a play-off. First #5 bowls #4. The loserSolutions to Introductory Problems
receives 5¢/
loser of this game receives receives 4‘ prize and the winner bowls
#2. The loser of this game receives 3" prize and the winner bowls
#1. The winner of this game gets 1** prize and the loser gets 2Ӣ
prize. In how many orders can bowlers #1 through #5 receive
the prizes?
prize and the winner bowls #3 in another game. The
Solution: There are 4 games in every play-off, and each game
has 2 possible outcomes. For each sequence of 4 outcomes, the
prizes are awarded in a different way. Thus there are 24 = 16
possible orders.
. [AMC 12 2001] A spider has one sock and one shoe for each of its
eight legs. In how many different orders can the spider put on
its socks and shoes, assuming that, on each leg, the sock must be
put on before the shoe?
Solution: Number the spider’s legs from 1 through 8, and let
ay and by denote the sock and shoe that will go on leg k. A
possible arrangement of the socks and shoes is a permutation
of the sixteen symbols a,,b1,...,ag,bs, in which a, precedes
by for 1 < k < 8. There are 16! permutations of the sixteen
symbols, and a; precedes 6; in exactly half of these, or 16!/2
permutations. Similarly, ay precedes bo in exactly half of those,
or 16!/2? permutations. Continuing, we can conclude that a,
precedes bj, for 1 < k < 8 in exactly 16!/2° permutations.
. [AHSME 1986] A drawer in a darkened room contains 100 red
socks, 80 green socks, 60 blue socks and 40 black socks. A
youngster selects socks one at a time from the drawer but is
unable to see the color of the socks drawn. What is the smallest
number of socks that must be selected to guarantee that the
selection contains at least 10 pairs? (A pair of socks is two socks
of the same color. No sock may be counted in more than one
pair.)
First Solution: For any selection, at most one sock of each
color will be left unpaired, and this happens if and only if an odd
number of socks of that color is selected. Thus 24 socks suffice:
at most 4 will be unpaired, leaving at least 20 in pairs. However,Solutions to Introductory Problems 27
10.
ll.
23 will do! Since 23 is not the sum of four odd numbers, at most
3 socks out of 23 will be unpaired. On the other hand, 22 will
not do: if the numbers of red, green, blue, and black socks are
5,5,5,7, then four are unpaired, leaving 9 pairs. Thus 23 is the
minimum.
Second Solution: Proceed inductively. If we require only one
pair, then it suffices to select 5 socks. Moreover, selecting 4 socks
doesn’t guarantee a pair since we might select one sock of each
color,
If we require two pairs, then it suffices to select 7 socks: any
set of 7 socks must contain a pair; if we remove this pair, then
the remaining 5 socks will contain a second pair as shown above.
On the other hand, 6 socks might contain 3 greens, | black, 1 red
and 1 blue — hence only one pair. Thus 7 socks is the smallest
number to guarantee two pairs.
Similar reasoning shows that we must draw 9 socks to guarantee
3 pairs, and in general, 2p +3 socks to guarantee p pairs. This
formula is easily proved by mathematical induction. Thus 23
socks are needed to guarantee 10 pairs.
[AIME 1991] Given a rational number, write it as a fraction in
lowest terms and calculate the product of the resulting numerator
and denominator. For how many rational numbers between 0 and
1 will 20! be the resulting product?
Solution: For a fraction to be in lowest terms, its numerator
and denominator must be relatively prime. Thus any prime factor
that occurs in the numerator cannot occur in the denominator,
and vice-versa. There are eight prime factors of 20!, namely
2,3,5,7, 11, 13,17, and 19. For each of these prime factors, one
must decide only whether it occurs in the numerator or in the
denominator. These eight decisions can be made in a total of
28 = 256 ways. However, not all of the 256 resulting fractions
will be less than 1. Indeed, they can be grouped into 128 pairs
of reciprocals, each containing exactly one fraction less than 1
Thus the number of rational numbers with the desired property
is 128.
Determine the number of ways to choose five numbers from the28
12
Solutions to Introductory Problems
first eighteen positive integers such that any two chosen numbers
differ by at least 2.
Solution: — Let a; < ay < a3 < ay < as be the five chosen
numbers. Consider the numbers (61, 62,63, ba,b5) = (a1,42 —
l,ag3 — 2,a4 — 3,45 — 4). Then b;,b2,b3,b4,b5 are five distinct
numbers from the first fourteen positive integers. Conversely,
from any five distinct numbers b; < be < b3 < ba < b5 we can
reconstruct (a1, 42,43, 44,45) = (bi, bz + 1,63 + 2,b4 + 3, bs +4)
to obtain five numbers satisfying the conditions of the problem.
Thus we found a one-to-one mapping between the set of five
numbers satisfying the given conditions and the set. of five distinct
numbers from the first fourteen positive integers. Therefore the
answer is (14) = 2002
[AHSME 1978] In a room containing N people, N > 3, at least
one person has not shaken hands with everyone else in the room.
What is the maximum number of people in the room that could
have shaken hands with everyone else?
Solution: Label the people Ai, As,... , Ay in such a way that
A, and Ay are a pair that did not shake hands with each other
Possibly every other pair of people shook hands, so that only Ay
and Ay did not shake with everyone else. Therefore, at most
N = 2 people shook hands with everyone else.
. [AIME 1998] Find the number of ordered quadruples (#1, #2, 3,
x4) of positive odd integers that satisfy 2, + 2 -+ 23 +24 = 98
Solution: Each x; can be replaced by 2y; — 1, where y; is a
positive integer. Because
4 4
m= en -0)=2 (Sn) -4
i=l i=)
it follows that 51 = 4, yz. Each such quadruple (91, y2, 93, Ya)
corresponds in a one-to-one fashion to a row of 51 ones that has
been separated into four groups by the insertion of three zeros.
For example, (17,5, 11, 18) corresponds to
LLLDLALILTTTTLTALOLITLTOULLITLILITLOLIIIT TALI.Solutions to Introductory Problems 29
14,
15.
There are (°') = 19600 ways to insert three zeros into the fifty
spaces between adjacent ones
(USSR, 1968] Finitely many cards are placed in two stacks, with
more cards in the left stack than the right. Each card has one or
more distinct names written on it, although different cards may
share some names. For each name, we define a shuffle by moving
every card that has that name written on it to the opposite stack.
Prove that it is always possible to end up with more cards in the
right stack by picking several distinct names, and doing in turn
the shuffle corresponding to each name.
Solution: — (By Oaz Nir) We will prove the statement by
induction on n, the number of distinct names present. Call the
left stack L and the right stack R. For case n = 1, one shuffle
will do the job.
We now assume that we have proven the statement for n names
(for some positive integer n), and consider the case with n+ 1
names. Call the first n names a1, a,..., an, and let the new name
be a. There are two cases.
e Case 1: The number of F cards containing only the name a
is less than or equal to the number of R cards containing only
the name a. We can ignore the name a and use the induction
hypothesis to perform required shuffles using some subsets
of the n names a), d2,...,d, and we will be done: there are
now more of these remaining cards in stack R than in stack
L, and since there were at least as many “only a” cards in R
than in L, the final configuration has more cards in R than
in L.
e Case 2: The number of L cards containing only the name
a is greater than the number of R cards containing only the
name a. Then we perform one shuffle with the name a, we
end up at the beginning of Case 1 and we are done.
In either case, we can finish our inductive step and our proof is
complete.
[AIME 1992] For how many pairs of consecutive integers in the
set
{1000, 1001, 1002, ..., 2000}30
16
17
Solutions to Introductory Problems
is no carrying required when the two integers are added?
Solution: Let n have a decimal representation labc. If one of
a,b, or ¢ is 5,6,7, or 8, then there will be carrying when n and
n+ 1are added. If6=9 and c #9, or if a = 9 and either b 49
or c #9, there will also be carrying when n and n+ 1 are added
If n is not one of the integers described above, then n has one
of the forms:
labe 1ab9 1a99 1999,
where a,b,c € {0,1,2,3,4}. For such n, no carrying will be
needed when n and n+1 are added. There are 5°+5?+5+1 = 156
such values of n
[AHSME 1994] Nine chairs in a row are to be occupied by six
students and Professors Alpha, Beta, and Gamma. These three
professors arrive before the six students and decide to choose their
chairs so that each professor will be between two students. In how
many ways can Professors Alpha, Beta, and Gamma choose their
chairs?
First Solution: The two end chairs must be occupied by
students, so the professors have seven middle chairs from which
to choose, with no two adjacent. If these chairs are numbered
from 2 to 8, the three chairs can be:
(2,4,7), (2,4,8), (2,5,7), (25,8)
(2,6,8), (3,5,7), (3,5,8), (3,6,8), (4,6,8)
Within each triple, the professors can arrange themselves in 3!
ways, so the total number is 10 x 6 = 60
Second Solution: Imagine the six students standing in a row
before they are seated. There are 5 spaces between them, each
of which may be occupied by at most one of the 3 professors
Therefore, there are P(5,3) = 5 x 4 x 3 = 60 ways the professors
can select their places
Prove that among any 16 distinct positive integers not exceeding
100 there are four different ones, a,b, c,d, such that a+b =e+d.Solutions to Introductory Problems 31
18.
Solution: Let a, < a) < +++ < aie denote the 16 numbers.
Consider the difference of each pair of those integers. There are
(18) = 120 such pairs.
Let (a;,aj) denote a pair of numbers with a; > aj. If we
have two distinct pairs of numbers (a;,,a;,) and (a;,,a;,) such
that a;, — a;, = ai, — ai,, then we get the desired quadruple
(a,b,¢,d) = (ai,,@,,4;,4;,) unless a;, = a;,. We say a is bad
for the pair of pairs (a;,,a) and (a,a;,) if aj, - a = a— aj, (or
2a = a;, + 4;,). Note that we are done if a number a is bad
for two pairs of pairs of numbers. Indeed, if a is bad for (a;,,),
(a,a;,) and (a;,,@), (a,a;,), then a;, +a;, = 2a = a;, + ai,
Finally, we assume that each a; is bad for at most one pair of
pairs of numbers. For each such pair of pairs of numbers, we take
one pair of numbers out of consideration. Hence there are no bad
numbers anymore. Then we still have at least 120 — 16 = 104
pairs of numbers left. The difference of the numbers in each
remaining pair ranges from 1 to 99. By the Pigeonhole Principle,
some of these differences have the same value. Assume that
in = dj, —4j,, then (aj, ,a;,, @;,,;,) satisfies the conditions
of the problem.
[AHSME 1989] A child has a set of 96 distinct blocks. Each
block is of one of 2 materials (plastic, wood), 3 sizes (small,
medium, large), 4 colors (blue, green, red, yellow), and 4 shapes
(circle, hexagon, square, triangle). How many blocks in the set
are different from the “plastic medium red circle” in exactly two
ways? (The “wood medium red square” is such a block.)
aj, -4,
Solution: For a block to differ from the given block, there is
only 1 choice for a different material, 2 choices for a different
size, 3 choices for a different color, and 3 choices for a different
shape. There are (3) = 6 ways a block can differ from the block
in exactly two ways:
(l
(2
(3
(
(
Material and size: 1-2 = 2 differing blocks.
Material and color: 1-3 = 8 differing blocks.
Material and shape: 1-3 = 3 differing blocks.
Size and color; 2-3 = 6 differing blocks.
Size and shape: 2-3 = 6 differing blocks.
4
532
19
20.
Solutions to Introductory Problems
(6) Color and shape: 3-3 = 9 differing blocks
Thus, 2+3+3+6+6+9 = 29 blocks differ from the given block
in exactly two ways.
[AHSME 1998] Call a 7-digit telephone number dydyd3—dadsdgd7
memorable if the prefix sequence didodg is exactly the same
as either of the sequences d4ad5dg or dsd6d7z (possibly both)
Assuming that each d; can be any of the ten decimal digits
0,1,2,...,9, find the number of different memorable telephone
numbers.
First Solution: There are 10,000 ways to write the last four
digits dadsded7, and among these there are 10000 — 10 = 9990
for which not all the digits are the same. For each of these, there
are exactly two ways to adjoin the three digits didodg to obtain a
memorable number. There are ten memorable numbers for which
the last four digits are the same, for a total of 2-9990+10 = 19990.
Second Solution: Let A denote the set of telephone numbers
for which djdodg is the same as dadsdg and let B be the set
of telephone numbers for which didodg coincides with dgdgdz. A
telephone number dy dod3—dydsdedz belongs to ANB if and only
if dy = dy = ds = dy = dg = dg = dy. Hence, n(ANB) = 10
Thus, by the Inclusion-Exclusion Principle,
n(AU B) = n(A) + n(B) - n(ANB)
= 10°-1-10+ 10-10 -1— 10 = 19990
[AIME 1996] Two of the squares of a 7 x 7 checkerboard are
painted yellow, and the rest are painted green. Two color schemes
are equivalent if one can be obtained from the other by applying a
rotation in the plane of the board. How many inequivalent color
schemes are possible?
Solution: There are (’) = 1176 ways to select the positions
of the yellow squares. Because quarter-turns can be applied
to the board, however, there are fewer than 1176 inequivalent
color schemes. Color schemes in which the two yellow squares
are not diametrically opposed appear in four equivalent forms
Color schemes in which the two yellow squares are diametricallySolutions to Introductory Problems 33
49.
opposed appear in two equivalent forms, and there are 49>! = 24
such pairs of yellow squares. Thus the number of inequivalent
color schemes is
1176-24 24
HTT 300,
7 12
[ARML 1999] In how many ways can one arrange the numbers
21, 31, 41, 51, 61, 71, and 81 such that the sum of every four
consecutive numbers is divisible by 3?
Solution: Since we only need to consider the problem modulo
3, we rewrite the numbers 21, 31, 41, 51, 61, 71, 81 as 0, 1, 2,
0, 1,2, 0. Suppose that a),ay,...,a7 is a required arrangement.
We observe that 0 = (a, + a2 +43 + a4) + (aq + 45 + a6 + 7) =
(a) + ay +++ +47) +445 (0414240414240) +045 ay
(mod 3). Thus aj,a2,a3 must be an arrangement of 0,1,2 as
ay+a2+a3 = a,+a2+a3+a4 = 0 (mod 3). Since aj+a2+a3+¢4 =
ay+a3-+a4-+a5 = 0 (mod 3), we have a, = as (mod 3). Similarly,
we can prove that the order of a5,a¢,a7 is uniquely determined
by a1, a2,a3. Thus we have 3 x 2° x 3!= 144 arrangements.
{AIME 1993] Let S be a set with six elements. In how many
different ways can one select two not necessarily distinct subsets
of S so that the union of the two subsets is S? The order
of the selection does not matter; for example the pair of sub-
sets {a,c}, {b,¢,d,e, f} represents the same selection as the pair
{b,c,d,e, f}, {a,c}
Solution: In order that AU B = S, for each element s of S
exactly one of the following three statements is true:
s€Aands¢B s¢AandseB s€AandseB
Hence if $ has n elements, there are 3” ways to choose the sets A
and B. Except for pairs with A = B, this total counts each pair
of sets twice. Since AU B = S with A= B occurs if and only if
A= B =S, the number of pairs of subsets of S' whose union is
Sis
Bead
2
+1,
which is 365 when n= 6.34
23
24
Solutions to Introductory Problems
[AIME 2001] A set of positive numbers has the triangle property
if it has three distinct elements that are the lengths of the sides
of a triangle whose area is positive. Consider sets {4,5,6,...,n}
of consecutive positive integers, all of whose ten-element subsets
have the triangle property. What is the largest possible value of
n?
Solution: The set {4,5,9, 14, 23, 37, 60,97, 157, 254} is a ten-
element subset of {4,5,6,...,254} that does not have the triangle
property. Let N’ be the smallest integer for which {4,5,6,..., N}
has a ten-element subset that lacks the triangle property. Let
{a1,a2,43,...,d19} be such a subset, with a) < a2 < a3 < +--+ <
ayo. Because none of its three-element subsets define triangles,
the following must be true:
N > ao > do + ag > (ag +47) + ag
= 2ag + a7 > 2(az +45) + 47 = 3a7 4 Qag
> 3(ag +45) + 2ag = 5ag + 3a5 > 8a5 + 5a4
> 1344 + 8a3 > 2las + 13az > 34a2 + lay
> 34-5421-4= 254
Thus the largest. possible value of n is N — 1 = 253. This is yet
another application of the Fibonacei sequence.
[MOSP 1997] Let A and B be disjoint sets whose union is the set
of natural numbers. Show that for every natural number n there
exist distinct a,b > n such that
{a,b,a+b}CA or {a,b,a +0} CB.
Solution: — We shall construct numbers a,b > n such that
a+b is in the same set as a and 6. First assume that |A| is
finite and that m is its largest element. Then n + 1,n+ 2, and
2n+3 = (nt1)+(n+2) are all in B for all n > m. Consequently,
we assume that both A and B are infinite sets.
We approach indirectly, Assume that there is a positive integer
nsuch that for any a,b > n, {a,b,a+} ¢ Aand {a,b,a+6} ¢ B
We now choose x,y, and z in A such that 2 > y > z > n and
y—2 > n. Thisis possible since A is infinite and thus unboundedSolutions to Introductory Problems 35
25.
Then {x2 +y,y+2,2+2}C B. But then y— z has no place to
go. Hence our assumption was wrong and we are done.
[AIME 1986] The increasing sequence 1,3, 4,9, 10,12, 13,... con-
sists of all those positive integers which are powers of 3 or sums of
distinct powers of 3. Find the 100" term of this sequence (where
Lis the 1% term, 3 is the 2°¢ term, and so on).
First Solution: If we use only the first six non-negative integral
powers of 3, namely 1,3,9,27,81 and 243, then we can form only
63 terms, since
()+Qe+G)
Consequently, the next highest power of 3, namely 729, is also
needed.
After the first 63 terms of the sequence the next largest ones
will have 729 but not 243 as a summand. There are 32 of these,
since (3) + (7) +---+ (2) = 32, bringing the total number of
terms to 95. Since we need the 100" term, we must next include
243 and omit 81. Doing so, we find that the 96", 97", ..., 100%"
terms are: 7294243, 729+243+1, 729+243+3, 729+243+3+1,
and 729 + 243 +9 = 981
1 = 63.
Second Solution: Note that a positive integer is a term of this
sequence if and only if its base 3 representation consists only of
0’s and 1's. Therefore, we can set up a one-to-one correspondence
between the positive integers and the terms of this sequence by
representing both with binary digits (0’s and 1's), first in base 2
and then in base 3:
T= la = I~ al
2= 102) = lg) =3
3= lla) = Hg) =4
4 = 100(2) => 100(3) =9
5 = 101(2) => 1013) = 1036
26.
Solutions to Introductory Problems
This is a correspondence between the two sequences in the order
given, that is, the k*" positive integer is made to correspond to
the k* sum (in increasing order) of distinct powers of 3. ‘This
is because, when the binary numbers are written in increasing
order, they are still in increasing order when interpreted in any
other base. (If you can explain why this is true when interpreted
in base 10, you should be able to explain it in base 3 as well.)
Therefore, to find the 100'" term of the sequence, we need only
look at the 100" line of the above correspondence:
100 = 11001002) => 11001003) = 981.
[AIME 1997] Every card in a deck has a picture of one shape —
circle, square, or triangle, which is painted in one of three colors
— red, blue, or green. Furthermore, each color is applied in one
of three shades — light, medium, or dark. The deck has 27 cards,
with every shape-color-shade combination represented. A set of
three cards from the deck is called complementary if all of the
following statements are true:
(a) Either each of the three cards has a different shape or all
three of the cards have the same shape.
(b) Either each of the three cards has a different color or all three
of the cards have the same color.
(c) Either each of the three cards has a different shade or all
three of the cards have the same shade
How many different complementary three-card sets are there?
Solution: Consider any pair of cards from the deck. We show
that there is exactly one card that can be added to this pair to
make a complementary set. If the cards in the pair have the same
shape, then the third card must also have this shape, while if the
cards have different shapes, then the third card must have the
one shape that differs from them. In eith , the shape on
the third card is uniquely determined. Similar reasoning shows
that the color and the shade on the third card are also uniquely
determined. The third card, determined by the first two, is never
one of the first two cards. Thus we can count the number of
complementary sets by counting the number of pairs of cards and
then dividing by 3, because each complementary set. is countedSolutions to Introductory Problems 37
~
three times by this procedure. The number of complementary
sets is
[China 1990] At a math camp, every m students share exactly
one common friend, m > 3. (If A is a friend of B, then B is a
friend of A. Also, A is not his own friend.) Suppose person P
has the largest number of friends. Determine what that number
is.
First Solution: — First note that every student has a friend.
Assume that students A, A2,...,Ag are friends of each other,
where k is a positive integer, 2 << k < m. Then there is a
student A;,4, who is a common friend to all of the students A;,
1 3, student
C has at least 2 friends among the students Aj, Ay,... Am4i-
But this impossible by our argument in Case 1
Overall, we showed that this camp only has the m+ 1 students
Ay, Ao,...;Am41 and they are all friends of each other. Hence
the desired number is m38
28.
Solutions to Introductory Problems
Second Solution: First we observe that P must have at least
m friends, since for any set of m students, their common friend
has at least m friends (namely, those m students). Now we prove
that P cannot have more than m friends. Assume the contrary.
Let S be the set of P’s friends, and let n = |S]. We have by
assumption n > m+ 1. We claim that for each (m — 1)-element
subset S’ of S, there exists a unique person Qs: € S who is a
common friend of all the people in 9’.
Consider any such subset 5’. Adding P to this set gives a set of
size m, and thus by the given there exists a unique person Q who
is a friend of P and all of the members of S’. We claim that this
Q is the Qs: that we want. Indeed, Q € S' since by definition, S
is the set of all friends of P
Now we claim that for any two distinct (m—1)-element subsets
Sy and So of S, Qs, # Qs,. Assume for a contradiction that this
is not the case, that is, there exist $1, 52 C S with Qs, = Qs,
Take any m-element. subset of S; U Sy. Then the people in this
set have two mutual friends, Qs, and P, contradicting the given
It follows that each (m— 1)-element subset S’ corresponds to a
different person Qs. Now, the number of m — 1-element subsets
of Sis
n\n).
m-1)=\2)7™
since n >m-+1 and m > 3. But n= |S
be the same, a contradiction.
[AHSME 1989] Suppose that 7 boys and 13 girls line up in a row
Let S be the number of places in the row where a boy and a
girl are standing next to each other. For example, for the row
GBBGGGBGBGGGBGBGGBGG we have S$ = 12. Find the
average value of S (if all possible orders of these 20 people are
considered)
, so two of the Q’s must
First Solution: Suppose that John and Carol are two of the
people. For ¢ = 1,2,...,19, let J; and C; be the numbers of
orderings (out of all 20!) in which the i” and (i+ 1)** persons
are John and Carol, or Carol and John, respectively. Then
J; = Cj = 18! is the number of orderings of the remaining persons
For i = 1,2,...,19, let Nj be the number of times a boy-girlSolutions to Introductory Problems 39
i
or girl-boy pair occupies positions i and i+ 1. Since there are 7
boys and 13 girls, Nj = 7-13 -(J; + Cj). Thus the average value
of S is
Ny +No+N3+...+Nig _ 19[7-13-(18!4 189] _ 91
20! 20! ~ 10
Second Solution: In general, suppose there are k boys and
n—k girls, For i= 1,2,...,n—1 let Aj be the probability that
there is a boy-girl pair in positions (i, i+1) in the line. Since there
is either 0 or 1 pair in (i,i + 1), Aj is also the expected number
of pairs in these positions. By symmetry, all A;’s are the same
(or note that the argument below is independent of i). Thus, the
answer is (n — 1) Aj.
We may consider the boys indistinguishable and likewise the
girls. (Why?) Then an order is just a sequence of k Bs and n—k
Gs. To have a pair at (i,i-+1) we must have BG or GB in those
positions, and the remaining n—2 positions must have k— 1 boys
and n —k —1 girls. Thus there are 2(?77) seque with a pair
at (i,i +1). Since there are (/}) sequences, the answer is
(n= 1)2(-7) _ 2k(n =k)
a 7
Thus, when n = 20 and k = 7, the answer is (2:7-13)/20 = 91/10
[AIME 1996] A bored student walks down a hall that contains a
row of closed lockers, numbered 1 to 1024. He opens the locker
numbered 1, and then alternates between skipping and opening
each closed locker thereafter. When he reaches the end of the
hall, the student turns around and starts back. He opens the
first closed locker he encounters, and then alternates between
skipping and opening each closed locker thereafter. The student
continues wandering back and forth in this manner until every
locker is open. What is the number of the last locker he opens?
(n-1)A; =
First Solution: Suppose that there are 2* lockers in the row,
and let L, be the number of the last locker opened. After the
student makes his first pass along the row, there are 2*~! closed
lockers left. These closed lockers all have even numbers and are
in descending order from where the student is standing. Now,
renumber the closed lockers from 1 to 2*~1, starting from the end40
30.
Solutions to Introductory Problems
where the student is standing. Notice that the locker originally
numbered n (where n is even) is now numbered 2*—! + 1 — 3
Thus, because Ly—1 is the number of the last locker opened with
this new numbering, we have
Le
+1
Solving for Ly we find
Ly = 42-2Ly-1
Iterate this recursion once to obtain
Tp = 2% 42 —2(2%-1 42-2049) = ALp_2 — 2 (1)
When there are 1024 = 2" lockers to start with, the last locker
to be opened is numbered Ly. Apply (1) repeatedly to Lo = 1
to find that Ly = 4Ly) —-2 = 2, Ly = 6, Le = 22, Lg = 86, and
Tio = 342.
Second Solution: Follow the given solution to the recursion
(1), which can be written in the form
Because Lo = 1 and L; = 2, it follows that
9 1-=)4% — ifkiseven,
Ly =
3 2-2 4S ifkisodd
These formulas may be combined to yield
= fg!) 49
k= 3 (4 $ +2)
for all nonnegative k. In particular, L4) = 342.
Note: How would the solution change if there were 1000 lockers
in the hall?
[AIME 1995] Let n = 2313!°, How many positive integer divisors
of n? are less than n but do not divide n?
First Solution: Let n = pg’, where p and q are distinctSolutions to Introductory Problems Al
BL
primes. Then n? = p?"q25, so n? has
(2r +1)(2s + 1)
factors. For each factor less than n, there is a corresponding
factor greater than n. By excluding the factor n, we see that
there must be
(Qr+1(2s+1) -1
2
factors of n? that are less than n. Because n has (r + 1)(s + 1)
factors (including n itself), and because every factor of n is also
a factor of n?, there are
2Irst+r+s
@strts—[(r+1(s+1)-=rs
factors of n? that are less than n but not factors of n. When
r=31 and s=19, there are rs = 589 such factors.
Second Solution: (By Chengde Feng) A positive integer
divisor d of n? is less than n but does not divide n if and only if
Biragh—o if ae < 3°,
d= { g3l-agl9+h if gas, 3b
where a and b are integers such that |< a< 31 and 1 1
do there exist a number m that can be written in the form
a) +--+ ay (with a € {]},a2 € {1,2},...,an € {1,-..,n})
in (n — 1)! or more ways?
First Solution: Note that for n = 1,2,3,4, we may choose
m = 1,3,5,7, respectively.
Note that each of the ways to write the number m in the formSolutions to Introductory Problems 43
a) +++: + ay (with a, € {1},a2 € {1,2},...,an € {1,....n})
requires a different ordered (n — 1)-tuple (a1,q2,..-,@n—1)- Fur-
thermore, there are only (n — 1)! such (n — 1)-tuples, so each of
those must work for m; i.e., we must have
2n—
Sl+l+-:+lin>m
Sees
n=
or else there would be no valid expression for m with a, = a2 =
+= 1, and also
m>14+24$--4(n—-1 415"
(2-1)
~_ th
or else there would be no valid expression for m with a, = 1,42 =
2,...,dn—1 =n — 1. Combining the two inequalities above, we
have
-1
An —1) > R@=)
a)
orn <4.
Hence n = 1,2,3,4 are the only n satisfying the conditions of
the problem.
Second Solution: (David Vincent) For each n, define the
polynomial
Sale) = ala 2) (ep er peta”).
It is clear that f,.(a) isa 1+24---+n = 240 degree polynomial
We can write
(ut)
Fale) = fart + fae? otf, ng 2G
Then the coefficient of the term x” in f, (x), [2"](fn(x)) = frm,
is equal to the number of ways that m can be written in the form
ay +--+ ay with ay € {1}, a2 € {1,2},...,4n € {1,...,n}. For
convenience, we may extend this definition to the other powers
of x by letting frm = 0 for all m not yet mentioned.44 Solutions to Introductory Problems
We have
file) =2,
folx) = 2? 42°,
f(a) = 29 + Qa? + 20° 4 2
fa(v) = a4 4 30° + 50° + 6x7 + 508 4 3e° 4 2
It follows that for n = 1,2,3,4,m = 1; m=2orm=3;m=4
or m= 5; m= 7, work, respectively.
It is not difficult to see that there are (1+2---+n)-—n+1=
a@=)) 41 terms in f,(x). For n > 5, it is also not difficult to see
that fn-1(x) has
(n=H(n-2)
Hh
terms. Since fn() = fn—i(w)(w+2?+--- a”), for positive integers
m,
[e"|(fa(@))
= fa-iym=1 t+ fn-tym-2 +++ + fr-iym—n
nin)
< > fa-aa = fna() = (n=
34. [AIME 1986] Let the sum of a set of numbers be the sum of its
elements. Let S be a set of positive integers, none greater than
15. Suppose no two disjoint subsets of S have the same sum.
What is the largest sum a set S with these properties can have?
Solution: First we show that S contains at most 5 elements.
Suppose otherwise. Then $ has at least (?) + (9) + (9) +())
or 56 subsets of 4 or fewer members. ‘The sum of each of these
subsets is at most 54 (since 15 + 14 + 13 + 12 = 54); hence, by
the Pigeonhole Principle, at least two of these sums are equal. If
the subsets are disjoint, we are done; if not, then the removal of
the common element(s) yields the desired contradiction.
On the other hand, it is not difficult to show that the set
S’ = {15, 14,13, 11,8} satisfies the conditions of the problem
The sum of S’ is 61. Hence the set S we seek is a 5-element set
with a sum of at least 61. Let S = {a,b,c,d,e} with a 4) boxes. Each time, Mr. Fat is allowed to pick two boxes,
take one candy bar from each of the two boxes, and put those
candy bars into a third box. Determine if it is always possible to
put all the candy bars into one box
Solution: It is always possible to put all the candy bars into
one box. We will prove our statement by induction on m, the
number of candy bars.
For the base case m = 4, there are at most 4 nonempty boxes.
We disregard all the other empty boxes and consider all the
possible initial distributions:
(1) (,1,1,1) (2) (1,2,1,0) (8) (2,2,0,0) (4) (1,3,0,0)
For distribution (1), we proceed as follows:
(1,1, 1,1) + (3, 1,0,0) + (2,0,2,0) > (1,0, 1,2) > (0,0, 0,4)
It is easy to see that all the other initial distributions are covered
in the above sequence of operations. Thus the base case is proved
Now we assume that the statement is true for some positive
integer m > 4. If we are given m+ 1 candy bars, we mark one
of them and called it special. We first ignore the special candy
bar and consider only the other m candy bars. By the induction
hypothesis, we can put all m candy bars into one box. If this box
also contains the special piece, we are done. If not, we pick two
empty boxes and proceed as follows:
(1, m, 0,0) — (0, m= 1, 2,0) + (0,m = 2, 1,2)
— (2,m = 3,0,2) > (1,m = 1,0,1) + (0,m+1,0,0)
Now all the candy bars are in one box and our induction is
complete46 Solutions to Introductory Problems
36. Determine, with proof, if it is possible to arrange 1,2,...,1000
in a row such that the average of any pair of distinct numbers is
not located in between the two numbers.
Solution: We claim that that it is possible to arrange 1,2,...,n
in a row such that the average of any pair of distinct numbers is
not located in between the two numbers.
We first prove that this is true for n = 2” for all positive
integers m. We induct on m. The base m= 1 is trivial
Now we assume that we can arrange 1,
positive integer m, ina row (a1, @,...,d2m) such that the average
of any pair of distinct numbers is not located in between the two
numbers. It is not difficult to
-,2™, for some
e that
(bi, bo, ome)
= (2a, = 1, 2a —1,...,2a9m — 1, 2ay,2ay,..., 2am)
is an arrangement of the numbers 1,2,...,2'+1 satisfying the
conditions of the problem. Indeed, the average of a pair of
numbers bj and 6; with either 1 3) be a regular n-sided polygon with O as its center
Triangular regions OA;Aiy1, 1 3) colors such that adjacentSolutions to Introductory Problems AT
38
regions are colored in different colors. Let pp, denote the number
of such colorings. We want to find pi2,4.
‘There are k ways to color the region OA; As, and then k — 1
ways to color regions OA»A3, OA3Aq and so on. We have to be
careful about the coloring of the region OA, Aj. It is possible
that it has the same color as that of region OA, Ay. But then, we
simply end up with a legal coloring for n — 1 regions by viewing
region OA, Az as one region. This is a clear bijection between
this special kind of illegal colorings of n regions to legal colorings
of n—1 regions. Hence pa, = k(k —1)""! — pn, Note that
Ps,k = k(k — 1)(k — 2). It follows that:
Pn = k(k = 17) = kk = 1"? + (k= 18 =
+ (=1)~4k(k = 18 + (-1)"Sk(k = 1)(k — 2)
(k= 1)" + (-1)"4(k = D8
=k- 1+(-1) + (-1)“Sk(k = 1)(k = 2)
= (k-1)" + (-1)"(k — 1)? + (- I)" k(k — (hk — 2)
= (k= 1" + (-1)"(k = U[(k = 1)? = k(k - 2)]
= (k= )" + (-1)"(k- 1)
Hence pi2,4 = 3!* +3 = 531, 444 legal ways to color this regular
dodecagon
‘There are 2n people at a party. Each person has an even number
of friends at the party. (Here friendship is a mutual relationship.)
Prove that there are two people who have an even number of
common friends at the party
Solution: Assume for a contradiction that every two of the
people at the party share an odd number of friends. Consider
any person P. Let A be the set of P’s friends, and let B be the
set containing everyone else. Observe that since |A| is even and
the total number of people at the party is 2n, |B| is odd. Consider
any person Q in B. By definition of B, Q is not a friend of P. By
assumption, Q shares an odd number of friends with P, so Q has
an odd number of friends in A. Since the total number of friends
of Q is even, Q must also have an odd number of friends in B
Now, summing the number of friends in B over all the Q’s in B,
we should obtain twice the number of friendships among people48
40.
Solutions to Introductory Problems
in B. But the sum is odd, since as noted previously, |B] is odd
This is a contradiction, and hence two of the people at the party
must share an even number of common friends.
Note: One can show that for every person P at the party there
exists a person Q who has an even number of common friends
with P at the party, Indeed, let sets A and B be as in the
solution. The set B is non-empty, since |B| is odd. ‘There must
be a person Q who has an even number of friends in B. Then Q
must also have an even number of friends in A, In order to justify
this stronger statement we did not use a proof by contradiction.
[AIME 1997] How many different 4 x 4 arrays whose entries are
all 1's and —1's have the property that the sum of the entries in
each row is 0 and the sum of the entries in each column is 0?
Solution: Each row and each column must contain two 1/s and
two —1's, so there are (3) = 6 ways to fill the first row. There are
also six ways to fill the second row. Of these, one way has four
matches with the first row, four ways have two matches with the
first row, and one way has no matches with the first row. The
first case allows one way to fill the third row, the second case
allows two ways to fill the third row, and the third case allows
six ways to fill the third row. Once the first three rows are filled,
the fourth row can be filled in only one way. Thus there are
6(1-1+4-2+1-6) = 90 ways to fill the array to satisfy the
conditions.
(IMO Shortlist 1996] A square of dimensions (n — 1) x (n — 1)
is divided into (n — 1)? unit squares in the usual manner. Each
of the n? vertices of these squares is to be colored red or blue
Find the number of different colorings such that each unit square
has exactly two red vertices. (Two coloring schemes are regarded
as different if at least one vertex is colored differently in the two
schemes.)
Solution: Let the vertices in the bottom row be assigned an
arbitrary coloring, and suppose that some two adjacent vertices
have the same color. Then it is not difficult to see that the
coloring of the remaining vertices are fixed. There are 2” — 2Solutions to Introductory Problems 49
Al.
42.
colorings of the bottom row with the property that some two
adjacent vert: have the same color (as there are a total of 2”
colorings and 2 ways alternates the coloring of adjacent: vertices.
If the vertices of the bottom row are colored alternately, this
property must be true for each of the other rows, as well. Hence
each row can be colored in 2 ways for a total of 2” ways.
Therefore the answer is 2" — 2+ 2" = 2"+! — 2 ways satisfying
the conditions of the problem.
Sixty-four balls are separated into several piles. At each step we
are allowed to apply the following operation. Pick two piles, say
pile A with p balls and pile B with q balls and p > q, and then
remove g balls from pile A and put them in pile B. Prove that it
is possible to put all the balls into one pile.
Solution: We use induction to prove that it is possible to put
all the n balls into one pile ifn = 2” for some nonnegative integer
m. The base cases m = 0 and m = 1 are trivial.
Now we assume that it is possible to put all the 2™ balls into
one pile for some positive positive integer m. We will show that it
is possible to put 2”+! balls into one pile. We first note that there
are an even number of piles each containing an odd number of
balls. We match those piles and apply the operation to each pair.
Hence after finitely many operations, each of the piles contains
an even number of balls. We then bind each pair of balls in each
pile to form a super ball. Hence we obtain a certain number of
piles of 2” super balls. By our induction hypothesis, we can put
these super balls into one pile. Hence all the 2"*? balls are now
in one pile and our induction is complete.
[USAMO 1999 submission, Richard Stong] A game of solitaire
is played with a finite number of nonnegative integers. On the
first. move the player designates one integer as large, and replaces
another integer by any nonnegative integer strictly smaller than
the designated large integer. On subsequent steps play is similar,
except that integer replaced must be the one designated as large
on the previous play. Prove that in some finite number of steps
play must end50
43.
44.
Solutions to Introductory Problems
Solution: Let the integers at any time be a1, a2,...,@,, and let
¢ be the index of the integer chosen as large in the previous step.
Define the score of the position to be S = Dj, ai. At any step
we will choose a new large integer a¢: (which currently contributes
to S but will not after the move), and we will replace ag (which
currently does not contribute to $) with something smaller than
av (which will contribute to the new S). Thus S is decreased by
at least 1 on every move. Since S starts with a finite value and
S > 0, play must stop in a finite number of moves
[USAMO 2000 submission, Cecil Rousseau] Given S C {1,2,
,n}, we are allowed to modify it in any one of the following
ways
(a) if 1 ZS, add the element 1;
(b) ifn €S, delete the element n;
(c) for l {n},
starting with @ and ending with {n}, in which each of the 2”
subsets of {1,2,...,n} appears exactly once. Prove that n =
2” — | for some m.
Solution: Let m be the sum of the set elements. Whenever
operation (a) or (c) is performed, m increases by 1, and whenever
(b) is performed, m decreases by n. If (b) is performed d times
in a sequence of k set modifications that starts and ends with the
same set S, then (k —d) — dn = 0, that is, k = d(n +1). Since
adding {n} + 9 to the presumed sequence gives
0 {1} 4 {2} 4--- > {n} 9,
a cycle of length k = 2", we have (n + 1) | 2”. Thus n must be
of the form 2” — 1 for some m a; > a2 > +++ > aj, the sum of the
alternating sums for each such pair of subsets is given by
(a) — a2 +--+ taj) +(n — a, a2 —-.-Faj;)=n.
And since there are 2" subsets of {1,2,...,n} and, consequently,
2"-1 such pairs of subsets, the required sum is n2”-!. Finally,
taking n = 7, we obtain 448
[AIME 1992] In a game of Chomp, two players alternately take
“bites” from a 5-by-7 grid of unit squares. ‘To take a bite,
the player chooses one of the remaining squares, then removes
(“eats”) all squares found in the quadrant defined by the left
edge (extended upward) and the lower edge (extended rightward)
of the chosen square. For example, the bite determined by the
shaded square in the diagram would remove the shaded square
and the four squares marked by xSolutions to Introductory Problems 53
(The squares with two or more dotted edges have been removed
from the original board in previous moves.) The object of the
game is to make one’s opponent take the last bite. The diagram
shows one of the many subsets of the set of 35 unit squares that
can occur during the game of Chomp. How many different subsets
are there in all? Include the full board and the empty board in
your count.
Solution: At any stage of the game, the uneaten squares will
form columns of non-increasing heights as we read from left to
right
It is not hard to show that this condition is not only necessary,
but is also sufficient for a given configuration of squares to occur
in a game. (The reader should prove this fact.) Moreover, any
such configuration can be completely described by the twelve-
step polygonal path that runs from the upper left to the lower
right of the original board, forming the boundary between the
eaten and uneaten squares. This polygonal boundary can be
described by a twelve-letter sequence of V's and H's. Such
a sequence contains seven H's, where each H represents the
top of an uneaten column (or bottom of a completely eaten
one) and five V/s, where each V represents a one-unit drop
in vertical height in moving from the top of an uneaten col-
umn to the top of an adjacent, but shorter column. For ex-
ample, the state that appears in the diagram accompanying
the problem is described by HHHVHVVHHHVY, while the
sequences HHHH HHHV / and VVVVVHHHHHHH54
47.
Solutions to Introductory Problems
describe the full board and the empty board, respectively. ‘Thus
the number of possible subsets is (1?) = 792
Note: The game of Chomp is due to David Gale, and was
introduced (and named) by Martin Gardner in his Scientific
American column “Mathematical Games”. The column reap-
peared in Gardner’s collection Knotted Doughnuts.
Each square of a 1998 x 2002 chess board contains either 0 or
1 such that the total number of squares containing | is odd in
each row and each column. Prove that the number of white unit
squares containing 1 is even
Solution: Let (i,j), 1 y5-1,5
f=1 j=l
is the sum of all the numbers in the 999 odd rows, i.e., Roda is
odd as it is the sum of 999 odd numbers. Likewise, sum of all the
numbers in even columns
1001 1998
Coven = > YP a25,i
j= i=l
is also odd as it is the sum of 1001 odd numbers. Let B denote
the set. of all the black squares in the even columns, and let $(B)
denote the sum of the numbers in the squares in set B. Note that
the numbers in each of the squares in B appears exactly once in
the sum Roaa. Note also that the numbers in each of the squares
in B appear exactly once in the sum Ceyen. Finally, note that
each of the numbers in the white square appears exactly once in
combined sum Roaa + Ceven. Thus the total of the numbers of
the white unit squares is Roaa + Ceven — 25(B), which is even.
Therefore the number of white unit squares containing 1 is even.
[AIME 1989] Let S' be a subset of {1,2,3,... , 1989} such that no
two members of $ differ by 4 or 7. What is the largest number
of elements S can have?Solutions to Introductory Problems 55
Solution: We first show that, given any set of 11 consecutive
integers from {1,2,3,..., 1989}, at most five of these 1] can be
elements of 8. We prove this fact for the set T= {1,2,3,..., 11},
but the same proof works for any set of 11 consecutive integers.
Consider the following partition of T, where each subset was
formed so that it can contribute at most one element to S:
{1,5} {2,9} {3,7} {4,11} {6,10} {8} ())
If it were possible to have 6 elements of 7’ in S, then each of the
sets in (1) would have to contribute exactly one element. That
this is impossible is shown by the following chain of implications:
8ESS1ESS5ESSIESS2ESS6ESSWESS
B8ESSTESSIUESSAESS8ES.
With the aid of (1), or otherwise, it is easy to find a 5-element
subset of 7 that satisfies the key property of S (i.e, no two
numbers differ by 4 or 7). One such set is
T’ = {1,3,4,6, 9}
We also find (perhaps to our surprise) that 7” has the remarkable
property of allowing for a periodic continuation. That is, if I
denotes the set of integers, then
S' = {k+1In|k ET! and n€ I}
also has the property that no two elements in the set differ by
4 or 7. Moreover, since 1989 = 180-11 +9, it is clear that
S cannot have more than 181-5 = 905 elements. Because the
largest element in T’ is 9, it follows that the set
S$ =S'N{1,2,3,...,1989}
has 905 elements and hence shows that the upper bound of 905
on the size of the desired set can be attained. This completes the
argument.
Note: The reader may wish to find other 5-clement subsets
of {1,2,3,...,11} that exhibit the key property of 8. Which of
these subsets can be used, as above, to generate a maximal S?56
49.
Solutions to Introductory Problems
The reader is also encouraged to explore similar problems with
other pairs (triples, etc.) of integers in place of 4 and 7, and to
find the appropriate motivations for the choice of 11 as the size
of the blocks of integers considered in the above solution
[USAMO 2002 submission, Zuming Feng] A class of fifteen boys
and fifteen girls is seated around a round table. Their teacher
wishes to pair up the students and hand out fifteen tests—one
test to each pair.
As the teacher is preparing to select the pairs and hand out the
tests, he wonders to himself, “How many seating arrangements
would allow me to match up boy/girl pairs sitting next to each
other without having to ask any student to change his or her
seat?” Answer the teacher's question. (Two seating arrangements
are regarded as being the same if one can be obtained from the
other by a rotation.)
Solution: We call a pairing good if each contains a boy and a
girl. It is clear that there are 15! good pairings. For each good
pairing, there are 14! x 2!° ways to arrange the students around
the table. Such a seating arrangement is called a good working
relation. Hence there are a total of 14! x 15! x 2!° good working
relations.
We call a seating arrangement good if it allows the teacher to
match up boy/girl pairs sitting next to each other without having
to ask any student to change his or her seat. We want to evaluate
x, the number of good arrangements. ‘There are two types of good
seating arrangements:
(a) A good arrangement that generates exactly one good working
relation. This means that there are at least two boys sitting
next to each other in the arrangement. These arrangements
are called good arrangements of the first type. Let #1 denote
the total number of good arrangements of the first type.
Es
A good arrangement that generates exactly two good working
relations. ‘This means that boys and girls are sitting alter-
nately. These arrangements are called good arrangements of
the second type. Let #2 denote the total number of good
arrangements of the second type. Then x) = 14! x 15! as
there are 14! ways to arrange all the boys around the deskSolutions to Introductory Problems 57
and there are 15! ways to arrange all the girls each in a gap
between two neighboring boys
We have « = 1 + 2, where 2 = 14! x 15! and
xy + Qe = 14! x 15) x 2!
Therefore « = 14!. 15!(2!5 — 1).
[Baltic Way 1999] Two squares on an 8 x 8 chessboard are called
touching if they have at least one common vertex. Determine
if it is possible for a king to begin in some square and visit all
the squares exactly once in such a way that all moves except the
first are made into squares touching an even number of squares
already visited.
Solution: It is not possible for the king to visit all the squares.
Assume for a contradiction that there exists a path such that
all moves except the first are made into squares touching an even
number of squares already visited. Clearly, the first move is made
into a square touching exactly one square already visited, namely
the starting square. Summing the number of touching squares
previously visited over all the moves, we therefore obtain an odd
number. On the other hand, every pair of touching squares is
counted exactly once in this sum, by the member of the pair that
was visited second. Thus, the sum is equal to the total number
of touching pairs. But this number is even, since the numbers of
touching pairs oriented north-south and east-west are equal, as
are the numbers of touching pairs oriented northeast-southwest
and northwest-southeast. Thus we have a contradiction, and no
path exists
[St. Petersburg 1988] A total of 119 residents live in a building
with 120 apartments. We call an apartment overpopulated if there
are at least 15 people living there. Every day the inhabitants of
an overpopulated apartment have a quarrel and each goes off
to a different apartment in the building (so they can avoid each
other). Is it true that this process will necessarily be completed
someday?
Solution: Let pi, p:
let a; denote the number of residents in apartment pj. We
.,pi20 denote the 120 apartments, andSolutions to Introductory Problems
consider the quantity
= 4i(a1—1) , a2(a2—1) @)20(a120 — 1)
i a tt
(Assume that all the residents in an apartment shake hand with
each other at the beginning of the day, then quantity S denotes
the number of the handshakes in that day.) If all a; < 15, then
the process is completed and we are done. If not, without. loss of
generality, we assume that a; > 15 and that the inhabitants in p;
go off to different apartments in the building. Assume that they
go to apartments pj,,Piz,-.-Pi,, On the next day, the quantity
is changed by an amount of
a(a—1
Gj, + jg to + dig, ee
which is positive as
a3, taj, to +a;,, < 9-a, < 119-15 = 104
and
ax(@ = 1) 15 x M4
2 -— 2
Hence the quantity is decreasing during this process. On the other
hand, S starts as a certain finite number and S is nonnegative.
Therefore this process has to be completed someday.
= 1054
Solutions to Advanced
ProblemsSolutions to Advanced Problems 61
1. [AIME 1985] In a tournament each player played exactly one
game against each of the other players. In each game the winner
was awarded 1 point, the loser got 0 points, and each of the
two players earned 1/2 point if the game was a tie. After the
completion of the tournament, it was found that exactly half of
the points earned by each player were earned in games against
the ten players with the least number of points. (In particular,
of the ten lowest scoring players earned half of her/his points
against the other nine of the ten). What was the total number of
players in the tournament?
Solution: Assume that a total of n players participated in
the tournament. We will obtain two expressions in n: one by
considering the total number of points gathered by all of the
players, and one by considering the number of points gathered
by the losers (10 lowest scoring contestants) and those gathered
by the winners (other n — 10 contestants) separately. To obtain
the desired expressions, we will use that fact that if k players
played against one another, then they played a total of k(& —1)/2
games, resulting in a total of k(&—1)/2 points to be shared among
them. In view of the last observation, the n players gathered a
total of n(n —1)/2 points in the tournament. Similarly, the losers
had 10 - 9/2 or 45 points in games among themselves; since this
accounts for half of their points, they must have had a total of 90
points. In games among themselves the n — 10 winners similarly
gathered (n — 10)(n — 11)/2 points; this also accounts for only
half of their total number of points (the other half coming from
games against the losers), so their total was (n — 10)(n — 11)
points. Thus we have the equation
n(n —1)/2 = 90+ (n — 10)(n - 11),
which is equivalent to
n® — 41n+ 400 = 0.
Since the left member of this equation may be factored as (n —
16)(n — 25), it follows that n = 16 or 25. We discard the first
of these in view of the following observation: if there were only
16 players in the tournament, then there would have been only 6
winners, and the total of their points would have been 30 points,62
Solutions to Advanced Problems
resulting in an average of 5 points for each of them. This is less
than the 90/10 or 9 points gathered, on the average, by each of
the losers! Therefore, n = 25; i.e., there were 25 players in the
tournament
Finally we show that such a tournament exists. Since n =
25, we have 15 winners and 10 losers. Every game that the
winners play among themselves results in a tie, giving each winner
(15 —1)/2 = 7 points from games played with other winners
Likewise, all the games played among the losers result in ties,
giving each of the 10 losers 4.5 points. For the ten games played
by each winner against the losers, six are wins, two are losses,
and two are ties, giving the winners another 7 points from games
played with losers. This gives each loser three wins, nine losses,
and three ties in games against winners, adding up to 4.5 more
points. Thus each of the 25 players receives exactly half of his/her
points in games against the losers, which is what we want.
. [USAMO 1999 submission, Titu Andreescu] Let n be an odd
integer greater than 1. Find the number of permutations p of
the set {1,2,...,n} for which
ne
Ip(L) ~ UI + [p(2) = 2] +--+ [p(n) = nl =
Solution: We have
Ip(L) — 1] + |p(2) = 2] +--+ [p(n) = n|
=+lt142424---4n4£n.
The maximum of |p(1)—1| + |p(2)—2| + «+++ |p(n)=n| is
a(-1-2-- -25t) see
2 2 2
3
(4 + +n)
n-1\n-1 n+3
=-(1+—— |] — ——
(et)
Let p(2#+) =k. We must have
{p0),00) oo (23) be . of