Combinatorics course week 7

Hoang Nguyen
July 2021

1 In-class problems
1. In a scientific conference, all participants can speak in total 2 · n languages
(n ≥ 2). Each participant can speak exactly two languages and each pair
of two participants can have at most one common language. It is known
that for every integer k, 1 ≤ k ≤ n−1 there are at most k−1 languages such
that each of these languages is spoken by at most k participants. Show
that we can choose a group from 2n participants which in total can speak
2n languages and each language is spoken by exactly two participants from
this group.
2. In a conference there are n ≥ 10  people. It is known that:
a. Each person knows at least n+2

3 other people.
b. For each pair of person A and B who don’t know each other, there exist
some people A(1), A(2), . . . , A(k) such that A knows A(1), A(i) knows
A(i + 1) and A(k) knows B.
c. There doesn’t exist a Hamilton path.
Prove that: We can divide those people into 2 groups: A group has a
Hamilton cycle, and the other contains of people who don’t know each
other.
3. There are 42 students taking part in the Team Selection Test. It is known
that every student knows exactly 20 other students. Show that we can
divide the students into 2 groups or 21 groups such that the number of
students in each group is equal and every two students in the same group
know each other.

2 Homework
1. There are at least 3 subway stations in a city. In this city, there exists a
route that passes through more than L subway stations, without revisiting.
Subways run both ways, which means that if you can go from subway
station A to B, you can also go from B to A. Prove that at least one of
the two holds.
(i). There exists three subway stations A, B, C such that there does not

1
 exist a route from A to B which doesn’t pass through
√ C.
(ii). There is a cycle passing through at least b 2Lc stations, without
revisiting a same station more than once.
2. There are 100 cities in a country, some of them being joined by roads.
Any four cities are connected to each other by at least two roads. Assume
that there is no path passing through every city exactly once. Prove that
there are two cities such that every other city is connected to at least one
of them.
3. In a city’s bus route system, any two routes share exactly one stop, and
every route includes at least four stops. Prove that the stops can be
classified into two groups such that each route includes stops from each
group.
4. Let n be a positive integer. Denote by Sn the set of points (x, y) with
integer coordinates such that

1
|x| + y + < n.
2

A path is a sequence of distinct points (x1 , y1 ), (x2 , y2 ), . . . , (x` , y` ) in Sn

such that, for i = 2, . . . , `, the distance between (xi , yi ) and (xi−1 , yi−1 ) is
1 (in other words, the points (xi , yi ) and (xi−1 , yi−1 ) are neighbors in the
lattice of points with integer coordinates). Prove that the points in Sn
cannot be partitioned into fewer than n paths (a partition of Sn into m
paths is a set P of m nonempty paths such that each point in Sn appears
in exactly one of the m paths in P).
5. Let N > M > 1 be fixed integers. There are N people playing in a chess
tournament; each pair of players plays each other once, with no draws. It
turns out that for each sequence of M + 1 distinct players P0 , P1 , . . . PM
such that Pi−1 beat Pi for each i = 1, . . . , M , player P0 also beat PM .
Prove that the players can be numbered 1, 2, . . . , N in such a way that,
whenever a ≥ b + M − 1, player a beat player b.
6. There is an integer n > 1. There are n2 stations on a slope of a mountain,
all at different altitudes. Each of two cable car companies, A and B,
operates k cable cars; each cable car provides a transfer from one of the
stations to a higher one (with no intermediate stops). The k cable cars
of A have k different starting points and k different finishing points, and
a cable car which starts higher also finishes higher. The same conditions
hold for B. We say that two stations are linked by a company if one can
start from the lower station and reach the higher one by using one or more
cars of that company (no other movements between stations are allowed).
Determine the smallest positive integer k for which one can guarantee that
there are two stations that are linked by both companies.

2
3 Previous exercises solution
3.1 ACM-ICPC Pacific Northwest 2016
For any cloud touching existing land, it can’t hurt to make that cloud
water, so we can greedily assign those. So, let’s assume there is no clouds
adjacent to land.

Now, in a optimal solution, all islands from clouds will consist of a single
square. Consider any optimal solution. If it has an island with more than
one square, we can just convert one of those to water and still have the
same number of islands.

So, we want to choose the maximum number of squares such that no

two are adjacent. This sounds a lot like the maximum independent set
problem. This problem is normally NP-hard, but we can notice that the
grid is bipartite. We can also notice that the complement of an indepen-
dent set is a vertex, so maximizing the size of an independent set is the
same as finding the size of a minimum vertex cover (follow from Konig’s
theorem). Using maximal matching algorithms and we are done.

3.2 Brazil MO
Take the obvious graph interpretation. If G has a perfect matching X ∪Y ,
then the second player plays symmetrical with respect the first player, i.e.,
if A plays in X, then B plays in its neighbor in Y . In this way, B makes
the last move, so B wins.
Now, if such matching does not exist, consider the maximal matching H of
G. In its first movement, A picks a vertex v0 not in H. By the maximality
of H, all the neighbors of v0 are in H, so B chooses a vertex u0 ∈ H. Then,
A chooses a neighbor v1 ∈ H of u0 . Now, observe that v1 cannot have
another neighbor v not in H, otherwise the graph would have the path
v0 − u0 − v1 − v, so we could take (v0 , u0 ) and (u1 , v) and increase the size
of H, a contradiction. Thus, since H has a matching with B playing first,
and the players keep playing in H after A making its first movement, we
have that A wins, so we are done!

3
4 References and hints
4.1 Problem sources
4.1.1 In-class exercises
(a) Vietnam TST 1992
(b) Vietnam TST 1998
(c) Vietnam TST 2012

4.1.2 Homework
(a) Korea Final 2015
(b) ARMO 2003
(c) ARMO 2012
(d) USAMO 2008
(e) USA TST 2009
(f) IMO 2020

4.2 Ideas
• #algorithm
• #maximalpath

5 Theory
This lesson covers the maximal path technique. By considering the max-
imal path, the starting and end points must be connected to vertices in
the path.

The path configuration can be connected to the chain and anti-chain or

the cycle configuration.

5.1 An idea for finding large cliques

P
Theorem 1. Prove that there exists a clique size ≥ v∈V 1/(n − deg V )

To find a clique of G: Suppose that G has n vertices. Find a vertex v of

the smallest possible degree in G. If the degree of v is n − 1, stop; G is a
clique, so the largest clique in G has size n.

Otherwise, remove v and all of its edges from G. Find the largest clique
in the smaller graph. Report that as the largest clique in G.

4
5.2 Theorems
Theorem 2. (Ore) A simple graph with n vertices (n ≥ 3n ≥ 3) is Hamil-
tonian if, for every pair of non-adjacent vertices, the sum of their degrees
is n or greater.

Theorem 3. (Dirac) If a simple undirected graph G has n vertices where

deg v ≥ δ∀v ∈ V then G has a path length min{2δ, n − 1}
Theorem 4. (Bondy-Chvatal) Adding a new edge uv connecting a non-
adjacent pair of vertices u and v with deg u + deg v ≥ n until no more
pairs with this property can be found. We have G is Hamiltonian if and
only if G + uv is Hamiltonian.
Theorem 5. (Redei) Any tournament on a finite number nn of vertices
contains a Hamiltonian path, i.e., directed path on all nn vertices