SMTO 2024 Power Round

Aitchison College - Team 326

Problem 2.1
Let us inspect the relation x ∼ y, which states that x ≽ y and y ≽ x.
The relation is symmetric, as x ∼ y would refer to a case where y ≽ x and x ≽ y. As these are identical
cases, we conclude that this is symmetric.
Problem 2.2
Let us assume a set G exists for the binary relation defined, assuming that its elements follow the below
given transitive property:
x≽y (1)

For a baseline, let us assume |G| = 1, then the statement for the definition holds true for set’s maximum
capacity of holding one element only. This one element only will then prove true as the maximal element for
the set G. This is the fundamental case thats sets the base for the forward mathematical induction we are
going to perform.

Let us assume that for any set G with n elements, where n ⩾ 1, the statement holds true since there
exists some x ∈ G such that for any y ∈ G, we have x ≽ y. Now let’s consider a set G with n + 1 elements.
We’ll mark the elements of G as g1 , g2 .....gn+1 . Since ⩾ is complete for any pairs of elements gi , gj ∈ G,
either gi ≽ gj or gj ≽ gi holds true. Thus, we get the properties to hold for this part.
′
Let’s consider the set G = [g1 , g2 ....gn ], which contains the first n elements of G. By the inductive hy-
′
pothesis set, G′ has a maximal element, let’s call it x, such as that x ≽ g for all g ∈ G .

Now let’s consider the [n + 1]th element, gn+1 . Since any element ⩾ is complete, either x ≽ gn+1 or
gn+1 ≽ x. If x ≽ gn+1 , then x remains the maximal element of G. Otherwise, if gn+1 ≽ x, then gn+1
becomes the new maximal element of G. In either case, G has a maximal element. Thus, by mathematical
induction, we have shown true that for any finite set G with a complete and transitive binary relation ⩾,
there exists a maximal element in the set G.
Problem 2.3
Let’s consider an infinite set of hotel rooms numbered as H = {1, 2, 3, . . .}. Each room r in H is slightly
preferred to the previous room and is denoted by the relation ⪰:

r′ ⪰ r

Where r, r′ ∈ H, if:

If r′ = r, then r′ ⪰ r.
If r′ =
̸ r, then r′ is the next room, and it is slightly preferred to r, i.e., r′ > r.

This relation ⪰ is complete because for any pair of rooms r and r′ in H, either r′ = r or r′ > r holds
true, satisfying the definition of completeness. Additionally, this relation ⪰ is transitive because if room r′
is preferred to room r, and room r′′ is preferred to room r′ , then it follows that room r′′ is preferred to room
r, which satisfies the transitivity property.

However, there is no maximal element in H with respect to this relation. For any room r in H, there
exists a room r′ = r + 1, which is slightly preferred to r. Since H is infinite, this process can continue
indefinitely, and there is no last or maximal room.

Therefore, the relation ⪰ defined over the set H satisfies the properties of completeness and transitivity
but does not have a maximal element.
Problem 2.4
≻ is a transitive relationship. Suppose the elements x, y and z within the set G with the relation x ≻ y,
and y ≻ z . Since x is strictly preferred to y, and y is strictly preferred to z, it implies that x is also strictly
preferred to z. Therefore the relation x ≻ y, and y ≻ z implies that x ≻ z, which makes it ≻ a transitive
relation according to Definition 2.2.
Problem 2.5
2
The maximum number of binary relations over n elements without being complete or transitive is 2n possible
combinations.
Problem 2.6
Assumptions and Transitivity
If we assume that the preference relation ⪰ can be represented by a utility function u, then the preference
x ⪰ y can be expressed as u(x) ⪰ u(y). This means that comparing x ⪰ y is equivalent to comparing
u(x) ⪰ u(y).

Transitivity
Transitivity implies that if x ⪰ y and y ⪰ z, then x ⪰ z must also hold. Given our earlier assumption
about utility values, we have u(x) ⪰ u(y) and u(y) ⪰ u(z). By the properties of real numbers, this leads
us to conclude u(x) ⪰ u(z). This is because u(x) ⪰ u(y) ⪰ u(z), so we can ignore u(y). Thus, based on
our assumption that preferences are represented by utility values, we conclude that x ⪰ z, confirming the
transitivity of ⪰.

Completeness
To show that ⪰ is complete, we need to demonstrate that for any two elements x, y ∈ G, either x ⪰ y or
y ⪰ x holds. Since u : G → R assigns a real value to each element in G, it follows that either u(x) ⪰ u(y)
or u(y) ⪰ u(x) must be true. Recalling our earlier definition of the utility function, we can conclude that
either x ⪰ y or y ⪰ x. Thus, ⪰ is proved to be complete.
Problem 2.7
• First Scenario: x ⪰ y implies g(x) ⪰ g(y): Since f is monotonic and g(x) ⪰ g(y), then by the
definition of monotonicity, f (g(x)) ⪰ f (g(y)). Therefore, if x ⪰ y under g, it also holds under f ◦ g.
• Second Scenario: x ≻ y implies g(x) ≻ g(y): Since we know that g(x) ≻ g(y) and that f is
monotonic, f (g(x)) ≻ f (g(y)). Thus, x ≻ y is preserved in f ◦ g.

To ensure that f ◦ g fully represents ⪰, we also need the converse to be true: If f (g(x)) ⪰ f (g(y)), then by
the monotonic nature of f and the properties of g as a utility function representing ⪰, it must follow that
g(x) ⪰ g(y), and therefore, since we know x ⪰ y, our proof is complete
Problem 3.1
In the given example, the question has already shown that x ≻ y.
According to both Alice and Carl, y ≻ z, which makes this statement true for the whole society as it is true
for the majority.

According to the transitive property, if x ≻ y and y ≻ z then x ≻ z.

However, according to both Bob and Carl, z ≻ x. The majority makes this statement true, directly contra-
dicting what the transitive property implies.

Hence, the given aggregation rules fails transitively.

Problem 3.2
• Alice: y ≻ x ≻ z;
• Bob: z ≻ x ≻ y;

• Carl: y ≻ z ≻ x;

According to:

1. Alice, x ≻ z so z cannot be a Condorcet winner;

2. Bob, x ≻ y so y cannot be a Condorcet winner;
3. Carl, z ≻ x so x cannot be a Condorcet winner.
Problem 3.3
Let f (x) be the number of individuals who prefer candidate x. Therefore, among the set of outcomes (a1 , ...
, an ); for there to be no Condorcet winners, f (i) >f (i + 1), for all values of i <= n − 1, while f (1) > f (n);
as this would ensure that every candidate is less preferred than at least one other candidate, and therefore
cannot be Condorcet.
An alternative case would be where f (n) > f (1), thus effectively breaking the ”cycle”. Since f (n) >
f (n − 1) > ... > f (1), The majority of the people would prefer candidate n, therefore n would be the
Condorcet candidate, Effectively proving the problem statement.
Problem 3.4
To find how many problem writers vote for each proposal and which proposal wins the majority of votes, we
need to consider the preferences of each problem writer based on their ideal number of bobas.
Evaluating both proposals:

• Erick’s proposal (7 bobas):

– Problem writers with ideal bobas from 0 to 7 would prefer Erick’s proposal over Vicktor’s because
it is closer to their ideal number.
– There are 8 problem writers with ideal bobas from 0 to 7: those with ideal bobas 0 through 7.

• Vicktor’s proposal (18 bobas):

– Problem writers with ideal bobas from 18 to 20 would prefer Vicktor’s proposal over Erick’s
because it is closer to their ideal number.
– There are 4 problem writers with ideal bobas from 18 to 20: those with ideal bobas 18 through
20.

Erick’s proposal receives more votes, so it wins the majority of votes among the problem
writers.
Problem 3.10
The order would be y ≻ z ≻ x.
As y is the first term in Dave’s ordering, it starts off as standing best.
The next item in Dave’s ordering is z. As Alice and Carl both have y ≻ z , we have z discarded and y is
still the standing best.
The next item in Dave’s ordering is x. As Alice and Bob both have x ≻ y , x is now the new standing best,
thus x is the final standing best.
Problem 3.11
The order would be x ≻ z ≻ y.
As x is the first term in Dave’s ordering, it starts off as standing best.
The next item in Dave’s ordering is z. As Alice and Carl both have z ≻ x , we have z become the new
standing best.
The next item in Dave’s ordering is y. As Alice and Carl both have y ≻ z , y is now the new standing best,
thus y is the final standing best.
Problem 4.1
Showing that the social choice function does not always output complete aggregated prefer-
ences
To show the unanimity social choice function does not always output the complete aggregated preferences.
Let us consider a society set S with possible outcomes O and individuals in the society as I. Let us assume
there exist 3 elements in the set O as the possible outcomes: a, b, b ∈ O and individuals i1 , i2 , i3 ∈ I such
that:
1. i1 prefers a over b in the ideal choice function

2. i2 prefers b over c
3. i3 prefers c over a
From this, we can make the following conclusion: No Particular Outcome is unanimously preferred over
the others as by the cases, there is no no outcome that every individual prefers over another to all the other
outcomes from the various individuals. Therefore, the defined social choice function can’t be applied to
always output complete aggregated preferences.