21

C2. A diagonal of a regular 2006-gon is called odd if its endpoints divide the
boundary into two parts, each composed of an odd number of sides. Sides
are also regarded as odd diagonals, Suppose the 2006-gon has been
dissected into triangles by 2003 nonintersecting diagonals, Find the
maximum possible number of isosceles triangles with two odd sides.

Solution 1.

Call an isosceles triangle odd if it has two odd sides. Suppose we are given
a dissection as in the problem statement. A triangle in the dissection which
is odd and isosceles will be called iso-odd for brevity.
Lemma. Let AB be one of dissecting diagonals and let L be the shorter part
of the boundary of the 2006-gon with endpoints A , B. Suppose that L
consists of n segments. Then the number of iso-odd triangles with vertices
on L does not exceed n /2.
Proof. This is obvious for n=2. Take n with 2<n ≤ 1003 and assume the claim
to be true for every L of length less than n . Let now L (endpoints A , B )
consist of n segments. Let PQ be the longest diagonal which is a side of an
iso-odd triangle PQS with all vertices on L (if there is no such triangle, there
is nothing to prove). Every triangle whose vertices lie on L is obtuse or right-
angled; thus S is the summit of PQS . We may assume that the five points
A , P , S , Q , B lie on L in this order and partition L into four pieces
L AP , L PS , LSQ , LQB (the outer ones possibly reducing to a point).

By the definition of PQ , an iso-odd triangle cannot have vertices on both L AP

and LQB. Therefore every iso-odd triangle within L has all its vertices on just
one of the four pieces. Applying to each of these pieces the induction
hypothesis and adding the four inequalities we get that the number of iso-
odd triangles within L other than PQS does not exceed n /2. And since each of
LPS , LSQ consists of an odd number of sides, the inequalities for these two
pieces are actually strict, leaving a 1/2+1/2 in excess. Hence the triangle
PSQ is also covered by the estimate n /2. This concludes the induction step
and proves the lemma.

The remaining part of the solution in fact repeats the argument from the
above proof. Consider the longest dissecting diagonal XY . Let L XY be the
shorter of the two parts of the boundary with endpoints X , Y and let XYZ be
the triangle in the dissection with vertex Z not on L XY . Notice that XYZ is
acute or right-angled, otherwise one of the segments XZ ,YZ would be longer
than XY . Denoting by L XZ , LYZ the two pieces defined by Z and applying the
lemma to each of L XY , LXZ , LYZ we infer that there are no more than 2006/2
iso-odd triangles in all, unless XYZ is one of them. But in that case XZ and YZ
are odd diagonals and the corresponding inequalities are strict. This shows
that also in this case the total number of iso-odd triangles in the dissection,
including XYZ , is not greater than 1003.
This bound can be achieved. For this to happen, it just suffices to select a
vertex of the 2006-gon and draw a broken line joining every second vertex,
starting from the selected one. Since 2006 is even, the line closes. This
already gives us the required 1003 iso-odd triangles. Then we can complete
the triangulation in an arbitrary fashion.

Solution 2.

Let the terms odd triangle and iso-odd triangle have the same meaning as in
the first solution.

Let ABC be an iso-odd triangle, with AB and BC odd sides. This means that
there are an odd number of sides of the 2006-gon between A and B and also
between B and C . We say that these sides belong to the iso-odd triangle ABC
.

At least one side in each of these groups does not belong to any other iso-
odd triangle. This is so because any odd triangle whose vertices are among
the points between A and B has two sides of equal length and therefore has
an even number of sides belonging to it in total. Eliminating all sides
belonging to any other iso-odd triangle in this area must therefore leave one
side that belongs to no other iso-odd triangle. Let us assign these two sides
(one in each group) to the triangle ABC .

To each iso-odd triangle we have thus assigned a pair of sides, with no two
triangles sharing an assigned side. It follows that at most 1003 iso-odd
triangles can appear in the dissection.
This value can be attained. as shows the example from the first solution.

N4. Find all triples of positive integers (a ,b , p) with p prime and

p
a =b !+ p .
Answer: (2 , 2 ,2) and (3 , 4 ,3).
Solution 1.

Clearly, a> 1. We consider three cases.

Case 1: We have a< p . Then we either have a ⩽b which implies a ∣ a p−b!= p
leading to a contradiction, or a> b which is also impossible since in this case
we have b ! ⩽ a! <a p −p , where the last inequality is true for any p>a >1.
Case 2: We have a> p . In this case b !=a p− p> p p− p ⩾ p is divisible by p and
b !=a − p is not divisible by a =b !+ p is divisible by p. Hence, a is divis a / p< p
p p

divides both a p and b ! and hence it also divides This means that b< 2 p. If
a< p then a / p< p on the other hand the case a ⩾ p is also impossible since
2 2

p=a −b ! which is impossible. On the other hand, the ⩾ b !+ p . then

p

2 p
a ⩾ ( p ) >(2 p−1)! + p ⩾ b !+ p.
p

Comment. The inequality p2 p >( 2 p−1) !+ p can be shown e.g. by using

(( ) )
2 p−1
2p 2 p −1
(2 p−1)!=[1 ⋅(2 p−1)]⋅[2 ⋅( 2 p−2)]⋯ ⋅[( p−1)(p +1)]⋅ p< ⋅ p= p
2

Case 3: We have a= p. In this case b !=5 does not lead to a solution. So we

now assume that p ⩾7. to the claimed solutions and p ! and so b ⩾ p+1 which
implies that

We have b != p p− p> p ! and v 2( p−1)+ v 2 ( p+1)−1=v 2 ( p−1

2
⋅(p−1)⋅( p+1)),

LTE
v 2(( p+1)!)⩽ v 2 (b! )=v 2 ( p −1 ) = 2 v 2 ( p−1)+v 2 ( p+1). On the RHS we have three factors of
p−1

where in the middle we used lifting-the-exponent lemma. On thers among

1 , 2, … , p+1, so this ( p+1)!. But, due to p+1 ⩾ 8, there are at least 4 er case is
not possible.
Solution 2. The cases a ≠ p are covered as in solution 1 , as are p=2, 3.
b != p ( p p−1−1 ). By Zsigmondy's Theorem there exists some ord q ⁡( p)= p−1, and
hence q ≡ 1 but does not divide pk −1 for k < p−1. It must have q ⩾2 p−1 ,
giving mod ( p−1). Note that p ≠ q. But then we ( p+1)¿ ⋅ p> ¿,
b ! ⩾(2 p−1)!=[1 ⋅(2 p−1)]⋅[2 ⋅(2 p−2)]⋯⋯ [( p−1)⋅( p+1)]⋅ p>(2 p−1) p−1 p> p > p− p ¿
a contradiction.
Solution 3. The cases a ≠ p are covered as in solution 1 , as are p=2, 3. Also
b> p , as p p > p !+ p for p>2. The cases p=5 , 7 , 11 are also checked manually, so
assume p ⩾13. Let q ∣ p+ 1 be an odd prime. By LTE

( )
p −1

v q ( p − p )=v q ( p )
p 2 2
−1 =v q ( p −1 ) + v q
2
( p−1
2 )
=v ( p+ 1).
q

But b ⩾ p+1, so then v q (b !)> v q ( p+1), since q < p+1, a contradiction. This
means that p+1 has no odd prime divisor, i.e. p+1=2 k for some k .
C3. In the coordinate plane consider the set S of all points with integer
coordinates. For a positive integer k , two distinct points A , B ∈ S will be
called k -friends if there is a point C ∈ S such that the area of the triangle
ABC is equal to k . A set T ⊂ S will be called a k -clique if every two points in T
are k -friends. Find the least positive integer k for which there exists a k -
clique with more than 200 elements.

Solution.

To begin, let us describe those points B∈ S which are k -friends of the point
(0 , 0). By definition, B=(u , v ) satisfies this condition if and only if there is a
1
point C=( x , y )∈ S such that ∨uy−vx ∨¿ k . (This is a well-known formula
2
expressing the area of triangle ABC when A is the origin.)
To say that there exist integers x , y for which ¿ uy−vx∨¿ 2 k , is equivalent to
saying that the greatest common divisor of u and v is also a divisor of 2 k .
Summing up, a point B=(u , v )∈ S is a k -friend of (0 , 0) if and only if gcd ⁡(u , v )
divides 2 k .

Translation by a vector with integer coordinates does not affect k -friendship;

if two points are k -friends, so are their translates. It follows that two points
A , B ∈ S , A=(s ,t), B=(u , v), are k -friends if and only if the point (u−s , v−t) is a
k -friend of (0 , 0); i.e., if gcd ⁡(u−s , v−t)∣2 k .

Let n be a positive integer which does not divide 2 k . We claim that a k -clique
cannot have more than n2 elements.

Indeed, all points (x , y )∈ S can be divided into n2 classes determined by the

remainders that x and y leave in division by n . If a set T has more than n2
elements, some two points A , B ∈ T , A=(s , t), B=(u , v ), necessarily fall into the
same class. This means that n ∣u−s and n ∣ v−t . Hence n ∣ d where
d=gcd ⁡(u−s , v −t). And since n does not divide 2 k , also d does not divide 2 k .
Thus A and B are not k -friends and the set T is not a k -clique.

Now let M (k ) be the least positive integer which does not divide 2 k . Write
M (k )=m for the moment and consider the set T of all points (x , y ) with
0 ≤ x , y <m . There are m2 of them. If A=( s , t), B=(u , v) are two distinct points in
T then both differences ¿ u−s∨,∨v−t∨¿ are integers less than m and at least
one of them is positive. By the definition of m , every positive integer less
than m divides 2 k . Therefore u−s (if nonzero) divides 2 k , and the same is
true of v−t . So 2 k is divisible by gcd ⁡(u−s , v−t) , meaning that A , B are k -
friends. Thus T is a k -clique.

It follows that the maximum size of a k -clique is M ¿, with M (k ) defined as

above. We are looking for the minimum k such that M ¿.

By the definition of M (k ), 2 k is divisible by the numbers 1 , 2, … , M (k )−1 , but

not by M (k ) itself. If M ¿ then M (k )≥15 . Trying to hit M (k )=15 we get a
contradiction immediately ( 2 k would have to be divisible by 3 and 5 , but
not by 15 ).
their least common M (k )=16 . Then 2 k is divisible by the numbers 1 , 2, … , 15,
hence also by that k =L/2 is the least k . L. but not by 16 . And since L is not a
multiple of 16 , we infer Finally, observe that if M (k )≥17 w M (k )=16 .
of 1 , 2, … , 16, which is equal to 2 L. Then 2 k must be divisible by the least
common multiple In conclusion, the least k with the required 2 L, yielding
k > L/2 .
In conclusion, the least k with the required property is equal to L/2=180180 .
C4. Let n and k be fixed positive integers of the same parity, k ≥ n. We are
given 2 n lamps numbered 1 through 2 n; each of them can be on or off. At the
beginning all lamps are off. We consider sequences of k steps. At each step
one of the lamps is switched (from off to on or from on to off).

Let N be the number of k -step sequences ending in the state: lamps 1 , … , n

on, lamps n+1 , … , 2 n off.

Let M be the number of k -step sequences leading to the same state and not
touching lamps n+1 , … , 2 n at all.
Find the ratio N / M .

Solution.

A sequence of k switches ending in the state as described in the problem

statement (lamps 1 , … , n on, lamps n+1 , … , 2 n off) will be called an admissible
process. If, moreover, the process does not touch the lamps n+1 , … , 2 n, it
will be called restricted. So there are N admissible processes, among which
M are restricted.

In every admissible process, restricted or not, each one of the lamps 1 , … , n

goes from off to on, so it is switched an odd number of times; and each one
of the lamps n+1 , … , 2 n goes from off to off, so it is switched an even number
of times.

Notice that M >0; i.e., restricted admissible processes do exist (it suffices to
switch each one of the lamps 1 … , n just once and then choose one of them
and switch it k −n times, which by hypothesis is an even number).

Consider any restricted admissible process p. Take any lamp l ,1 ≤ l≤ n, and

suppose that it was switched k l times. As noticed, k l must be odd. Select
arbitrarily an even number in 2k −1 ways (because a k t-element set has 2k −1
i l

subsets of even cardinality). Notice that k 1+⋯+ k n=k .

These actions are independent, in the sense that the action involving lamp l
does not affect the action involving any other lamp. So there are
k −1 k −1 k −1 k−n
2 ⋅ 2 ⋯ 2 =2 ways of combining these actions. In any of these
1 2 n

combinations, each one of the lamps n+1 , … , 2 n gets number of times, so the
final state is the same as that resulting from the original process p. rise to
k−n k−n
2 distinct admissibtricted admissible process p can be modified in 2
ways, giving
Now we show that every ad processes (with all lamps allowed).
enough to replace every switch of a lamp with q can be achieved in that way.
Indeed, it is the corresponding lamp l−n ; in the resulting process p the that
occurs in q by the switch of

Switches of each lamp with a label l>n had occurrel imps n+1 , … , 2 n are not
involved. the performed replacements have affected each lamp with in q an
even number of times. So times; hence in the overall effect the final state of
each lamp abel l ≤n also an even number of that the resulting process p is
admissible - and clearly remp has remained the same. This means not
involved in it any more. These reversed replacements are nothinge all these
replacements, then we obtain process q . paragraphs.

Thus there is a one to ( 2k−n ) correspos and the total of N admissible

processes.

Shortlisted problems - solutions

85
Putting together (2) and (3), for n ⩾6 we get a contradiction, since

( n (n−1)
2 )
2

Ln <2n < !< m!=L .

n

Hence n ⩾6 is not possible.

Checking manually the cases n ⩽5 we find
 L1=1=1! , L2=6=3 ! , 5 !< L3=168< 6 ! ,
7 !< L 4=20160< 8! and 10 !< L5=9999360<11! .

So, there are two solutions:

(m , n)∈ {(1 , 1),(3 , 2)}

Solution 2. Like in the previous solution, the cases n=1 ,2 , 3 , 4 are checked
manually. We will exclude n ⩾5 by considering the exponents of 3 and 31 in
(1).

For odd primes p and distinct integers a , b , coprime to p, with p ∣a−b , the
Lifting The Exponent lemma asserts that

v p ( ak −b k )=v p (a−b)+ v p (k ).

Notice that 3 divides 2k −1 if only if k is even; moreover, by the Lifting The

Exponent lemma we have

v 3 ( 22 k −1 )=v3 ( 4k −1 )=1+ v 3 (k )=v 3 (3 k ).

Hence,

v 3 ( Ln ) = ∑ ❑ v 3 ( 4 k −1 ) = ∑ ❑ v 3 (3 k)
2 k≤ n
k≤ ∖
π
2 |
Notice that the last expression is precisely the exponent of 3 in the prime

(
factorisation of 3 ⌊
n
2 )
⌋ !. Therefore

((
n
v 3 (m!)=v 3 ( Ln )=v 3 3 ⌊ ⌋ !
2 ))
n n
3 ⌊ ⌋ ⩽ m⩽ 3 ⌊ ⌋ +2.
2 2

Suppose that n ⩾5. Note that every fifth factor in Ln is divisible by 31=25 −1,
n
and hence we have v 31 ( Ln ) ⩾ ⌊ ⌋ . Then
5
∞ ∞
n n m m m
⩽ ⌊ ⌋ ⩽ v 31 ( Ln ) =v 31 (m!)=∑ ❑ ⌊ k ⌋ < ∑ ❑ k = .
10 5 k=1 31 k =1 31 30

By combining (4) and (5),

3n
3 n<m ⩽ +2
2
 4
so n< which is inconsistent with the inequality n ⩾5.
3
Comment 1. There are many combinations of the ideas above; for example
combining (2) and (4) also provides n<5 . Obviously, considering the
exponents of any two primes in (1), or considering one prime and the
magnitude orders lead to an upper bound on n and m .

N1. sum of elements of

Find all pairs (m , n) of positive integers satisfying the equation

( 2n −1 )( 2n−2 ) ( 2n−4 ) ⋯ ( 2n−2n−1) =m!

Answer: The only such pairs are (1 , 1) and (3 , 2).

Common remarks. In all solutions, for any prime p and positive integer N ,
we will denote by v p (N) the exponent of the largest power of p that divides
N . The left-hand side of (1) will be denoted by Ln; that is,
Ln=( 2n−1 ) ( 2n−2 ) ( 2n −4 ) ⋯ ( 2 n−2n−1 ).

Solution 1. We will get an upper bound on n from the speed at which v 2 ( Ln )

grows.
From

Ln=( 2n−1 ) ( 2n−2 ) ⋯ ( 2n−2 n−1 ) =21+2 +⋯+(n−1) ( 2n−1 ) ( 2n −1−1 ) ⋯ ( 21−1 )

we read

n(n−1)
v 2 ( Ln ) =1+ 2+ ⋯+(n−1)= .
2

On the other hand, v 2( m!) is expressed by the Legendre formula as

∞
m
v 2( m!)=∑ ❑ ⌊ ⌋.
i =1 2i

As usual, by omitting the floor functions,

∞
m
v 2( m!)< ∑ ❑ i =m.
i=1 2

Thus, Ln=m ! implies the inequality

 n(n−1)
<m .
2

In order to obtain an opposite estimate, observe that

n
Ln=( 2 −1 ) ( 2 −2 ) ⋯ ( 2 −2 ) < ( 2n ) =2n .
2
n n n n−1

We claim that

( n (n−1)
2 )
2

2n < ! for n ⩾ 6.

For n ⩾7 we prove ( 3 ) by the following inequalities:

( )
n(n−1)
n(n−1) n(n−1) 36 2
−15
! ¿ 15 ! ⋅16 ⋅17 ⋯ >2 ⋅16
2 2
¿ ¿

6. We are given an infinite deck of cards. each with a real number on it. For
every real humber x , there is exactly one card in the deck that has x written
on it. Now two players draw disjoint sets A and B of 100 cards each from
this deck. We would like to define a rule that declares one of them a winner.
This rule should satisfy the following conditions:

1. The winner only depends on the relative order of the 200 cards: if the
cards are laid down in increasing order face down and we are told
which card belongs to which player, but not what numbers are written
on them, we can still decide the winner.
2. If we write the elements of both sets in increasing order as
A={ a1 , a2 , … , a100 } and B= { b1 ,b 2 , … , b100 }, and a i> bi for all i, then A beats
B.
3. If three players draw three disjoint sets A , B ,C from the deck, A beats
B and B beats C , then A also beats C .
How many ways are there to define such a rule? Here, we consider two rules
as different if there exist two sets A and B such that A beats B according to
one rule, but B beats A according to the other.
(Russia)

Answer. 100.
Solution 1. We prove a more general statement for sets of cardinality n (the
problem being the special case n=100, then the answer is n ). In the
following, we write A> B or B< A for " A beats B ".

Part I. Let us first define n different rules that satisfy the conditions. To this
end, fix an index k ∈ {1 ,2 , … , n }. We write both A and B in increasing order as
A={ a1 , a2 , … , an } and B= { b1 ,b 2 , … , b n } and say that A beats B if and only if a k > bk
. This rule clearly satisfies all three conditions, and the rules corresponding
to different k are all different. Thus there are at least n different rules.

Part II. Now we have to prove that there is no other way to define such a
rule. Suppose that our rule satisfies the conditions, and let k ∈ {1 ,2 , … , n } be
minimal with the property that

A k ={1 , 2 ,… , k , n+ k +1 ,n+ k +2 , … , 2n }< Bk ={k +1 , k +2 , … , n+k }.

Clearly, such a k exists, since this holds for k =n by assumption. Now

consider two disjoint sets X ={ x 1 , x 2 ,… , x n } and Y = { y 1 , y 2 , … , y n }, both in
increasing order (i.e., x 1< x2 <⋯< x n and y 1 < y 2 <⋯< y n ). We claim that X <Y if
(and only if this follows automatically) x k < y k .
To prove this statement, pick arbitrary real numbers ui , v i , wi ∉ X ∪ Y such
that

u1 <u2 <⋯<u k−1< min ( x1 , y 1 ) ,max ( xn , y n ) < v k+1 < v k+2 <⋯< v n ,

and

x k <v 1 < v 2< ⋯<v k < w1 <w 2< ⋯<w n< uk <uk +1< ⋯<un < y k ,

and set

U ={ u1 ,u 2 , … , un } , V ={ v 1 , v 2 , … , v n } , W ={ w 1 , w 2 , … , w n } .

Then

 ui < y i and x i <v i for all i, so U <Y and X <V by the second condition.

Number Theory

N1. Find all pairs (k , n) of positive integers for which 7 k −3n divides k 4 +n2.
Answer. (2 , 4) ,

Solution.
 Suppose that a pair (k , n) satisfies the condition of the problem. Since 7 k −3n
is even, k 4 +n2 is also even, hence k and n have the same parity. If k and n are
odd, then k 4 +n2 ≡1+1=2 (mod 4 ), while 7 k −3n ≡7−3 ≡0(mod 4) , so k 4 +n2 cannot
be divisible by 7 k −3n. Hence, both k and n must be even.

a b
7 −3
Write k =2 a , n=2 b . Then 7 k −3n=7 2 a−32 b= ⋅ 2 ( 7 +3 ), and both factors are
a b
2
integers. So 2 ( 7a +3 b ) ∣ 7k −3 n and 7 k −3n ∣k 4 +n2=2 ( 8 a 4 +2 b2 ), hence

a b 4 2
7 +3 ≤ 8 a +2 b .

We prove by induction that 8 a 4 <7 a for a ≥ 4 ,2 b 2< 3b for b ≥ 1 and 2 b2 +9 ≤ 3b for

b ≥ 3. In the initial cases a=4 , b=1 , b=2 and b=3 we have
4 4 2 2 2 3
8 ⋅ 4 =2048<7 =2401 , 2<3, 2 ⋅2 =8<3 =9 and 2 ⋅3 +9=3 =27 , respectively.
If 8 a 4 <7 a (a ≥ 4) and 2 b2 +9 ≤ 3b (b ≥ 3), then

as desired.
For a ≥ 4 we obtain 7 a+3 b >8 a 4 +2 b2 and inequality (1) cannot hold. Hence a ≤ 3,
and three cases are possible.

Case 1: a=1. Then k =2 and 8+2 b 2 ≥ 7+3b , thus 2 b2 +1 ≥3 b. This is possible only
4 2 4 2
k + n 2 +2 1
if b ≤ 2. If b=1 then n=2 and k n
= 2 2 = , which is not an integer. If b=2
7 −3 7 −3 2
4 2 4 2
k +n 2 +4
then n=4 and k n
= 2 4 =−1, so (k , n)=(2 , 4) is a solution.
7 −3 7 −3

Case 2: a=2. Then k =4 and k 4 +n2=256+4 b2 ≥|7 4−3n|=|49−3b|⋅ ( 49+ 3b ). The

smallest value of the first factor is 22 , attained at b=3, so 128+2 b2 ≥ 11 ( 49+3 b )
, which is impossible since 3b >2 b 2.

Case 3: a=3. Then k =6 and k 4 +n2=1296 +4 b2 ≥|7 6−3n|=|343−3b|⋅ ( 343+3 b ).

Analogously, |343−3b|≥ 100 and we have 324+ b2 ≥25 ( 343+ 3b ), which is
impossible again.
We find that there exists a unique solution (k , n)=(2 , 4).
N1. Determine all pairs (x , y ) of integers satisfying the equation

x 2 x+1 2
1+2 +2 =y

we get the two solutions (0 , 2) and (0 ,−2) . Now let (x , y ) be a solution with
x >0 ).
x >0; without loss of generality confine attention to y >0 .

2 x ( 1+2 x+1 )=( y−1)( y +1)

shows that the factors y−1 and y +1 are even, exactly one of them divisible
by 4 . Hence x ≥ 3 and one of these factors is divisible by 2 x−1 but not by 2 x. So

x−1
y=2 m+ϵ , m odd , ϵ=±1.

Plugging this into the original equation we obtain

2
2 ( 1+2 )=( 2x−1 m+ϵ ) −1=22 x−2 m2 +2 x mϵ ,
x x+1

or, equivalently
x+1 x−2 2
1+2 =2 m + mϵ

1−cm=2x−2 ( m2−8 )

For ϵ=1 this yields m 2−8 ≤ 0, i.e., m=1, which fails to satisfy (2).
For ϵ=−1 equation (2) gives us

1+m=2 x−2 ( m2−8 ) ≥2 ( m2−8 ) ,

implying 2 m2−m−17 ≤ 0. Hence m ≤3 ; on the other hand m cannot be 1 by (2).

Because m is odd, we obtain m=3 , leading to x=4. From (1) we get y=23.
These values indeed satisfy the given equation. Recall that then y=−23 is
also good. Thus we have the complete list of solutions (x , y ):(0 , 2),(0 ,−2).