Solutions 34 2

USA Mathematical Talent Search
Round 2 Solutions
Year 34 — Academic Year 2022–2023
www.usamts.org
1/2/34. Fill in the grid below with the numbers 1 through 25, with each number used exactly
once, subject to the following constraints:
1. Each shaded square contains an even number, and each unshaded square contains an
odd number.
2. For any pair of squares that share a side, if x and y are the two numbers in those
squares, then either x ≥ 2y or y ≥ 2x.
Four numbers have been filled in already.
Create PDF with GO2PDF for free, if you wish to remove this line, click here to buy Virtual PDF Printer
16
25
13
21
There is a unique solution, but you do not need to prove that your answer is the only
one possible. You merely need to find an answer that satisfies the conditions of the problem.
(Note: In any other USAMTS problem, you need to provide a full proof. Only in this
problem is an answer without justification acceptable.)
Solution
16
24 6 17
25 12 3 8 18
22 11 5 13 4 9 19
23 2 1 10 21
15 7 20
14
Round 2 Solutions
www.usamts.org
2/2/34. Grogg’s favorite positive integer is n ≥ 2, and Grogg has a lucky coin that comes up
heads with some fixed probability p, where 0 < p < 1. Once each day, Grogg flips his coin,
and if it comes up heads, he does two things:
1. He eats a cookie.
2. He then flips the coin n more times. If the result of these n flips is n − 1 heads and 1
tail (in any order), he eats another cookie.
He never eats a cookie except as a result of his coin flips. Find all possible values of n and
p such that the expected value of the number of cookies that Grogg eats each day is exactly
1.
Solution
For (1), the probability that the coin comes up heads is p, so this event contributes an
expected value of p · 1 cookies.
For (2), binomial probability tells us that the probability that the n flips result in n − 1
heads and 1 tail is npn−1 (1 − p). Since Grogg only does the n coin flips if the initial flip
was heads, we multiply by p. The second event contributes an additional expected value of
pnpn−1 (1 − p) cookies.
Thus, the expected number of cookies that Grogg eats each day is p(1 + npn−1 (1 − p)).
Since we want to find all possible values of n and p that give us an expected value of 1,
we solve the equation p(1 + npn−1 (1 − p)) = 1. The equation rearranges and factors as
1
(p − 1)(npn − 1) = 0. Since p < 1, we must have npn = 1, so p = √ is the solution for
n
n
any n.
Round 2 Solutions
www.usamts.org
3/2/34. Let n ≥ 3 be a positive integer. Alex and Lizzie play a game. Alex chooses n positive
integers (not necessarily distinct), writes them on a blackboard, and does nothing further.
Then, Lizzie is allowed to pick some of the numbers—but not all of them—and replace
them each by their average. For example, if n = 7 and the numbers Alex writes on the
blackboard to start are 1, 2, 4, 5, 9, 4, 11, then on her first turn Lizzie could pick 1, 4, 9, erase
them, and replace them each with the number 1+4+9 3
, leaving on the blackboard the numbers
14 14 14
3
, 2, 3 , 5, 3 , 4, 11. Lizzie can repeat this process of selecting and averaging some numbers
as often as she wants. Lizzie wins the game if eventually all of the numbers written on the
blackboard are equal. Find all positive integers n ≥ 3 such that no matter what numbers
Alex picks, Lizzie can win the game.
Solution
Lizzie wins if and only if n is composite.
Suppose n is composite, so that n = rs for some 1 < r, s < n. Lizzie can win in r + s
turns, as follows. First, Lizzie arranges the numbers into an r × s matrix. On her first r
turns, Lizzie takes the s numbers in one of the r rows, in succession, and replaces them with
their average. Note that after this, each column of the matrix is identical. Then on her
next s terms, Lizzie takes the r numbers in one of the s columns, in succession, and replaces
them with their average. Since each column was identical, the result will be that all of the
numbers in the matrix will be equal.
Conversely, suppose that n is prime, and Alex writes n − 1 1’s and one 2 on the board.
Then at the end of the game, each number must be n+1 n
. However, since Lizzie must always
choose k < n numbers, she will always multiply any existing denominator of a number on
the board by a factor of k when averaged. In particular, since n is prime there is no way to
get a n in any denominator at any time during the game.
Round 2 Solutions
www.usamts.org
4/2/34. A lattice point of the coordinate plane is a point (x, y) in which both x and y are
integers. Let k ≥ 2 be a positive integer. Find the smallest positive integer ck (which may
depend on k) such that every lattice point can be colored with one of ck colors, subject to
the following two conditions:
1. If (x, y) and (a, b) are two distinct neighboring points; that is, |x−a| ≤ 1 and |y−b| ≤ 1,
then (x, y) and (a, b) must be different colors.
2. If (x, y) and (a, b) are two lattice points such that x ≡ a (mod k) and y ≡ b (mod k),
then (x, y) and (a, b) must be the same color.
Solution

 4 if k is even,
The answer is ck = 9 if k = 3,
5 if k > 3 is odd.

Call two distinct lattice points adjacent if they satisfy the condition in (1). Note that
(0, 0), (0, 1), (1, 0), and (1, 1) are all mutually adjacent, so regardless of k, we need ck ≥ 4.
Also note that any two adjacent points will differ in parity in at least one coordinate.
If k is even, then 4 colors suffice: color all (odd, odd) points with one color, (odd, even)
points with a second color, (even, odd) points with a third color, and (even, even) points with
a fourth color. Then since adjacent points differ in parity in at least one coordinate, they
will always have different colors, and furthermore since k is even, two points with each pair
of coordinates equivalent modulo k will have each pair of coordinates that are respectively
equal in parity, and thus be the same color.
If k = 3, then the color pattern must repeat in a 3 × 3 grid. This implies that each point
in any 3 × 3 grid within the coordinate plane must have a different color. So c3 = 9.
If k > 3 is odd, then 4 colors is insufficient: if we attempt to use only 4 colors, then as
noted earlier (0, 0), (0, 1), (1, 0), and (1, 1) must use all four colors, and an easy induction
shows that the points (2n + 1, 0) and (2n + 1, 1) must use the same two colors as (1, 0) and
(1, 1) (in either order). But then (k, 0) will not match (0, 0), a contradiction.
However, 5 colors is sufficient, as follows. Color the points (i, j) with 0 ≤ i, j ≤ 4 using
five colors A,B,C,D,E as below:
D E A B C
B C D E A
E A B C D
C D E A B
A B C D E
Round 2 Solutions
www.usamts.org
Then extend this coloring to all (i, j) with 0 ≤ i < k and 0 ≤ j ≤ 4 inductively by
coloring (i + 2, j) the same as (i, j). Finally, extend this coloring to all (i, j) with 0 ≤ i, j < k
inductively by coloring (i, j + 2) the same as (i, j), and copy this k × k grid across the entire
coordinate plane.
Round 2 Solutions
www.usamts.org
5/2/34. Let r and s be positive real numbers, and let A = (0, 0), B = (1, 0), C = (r, s), and
D = (r + 1, s) be points on the coordinate plane. Find a point P , whose coordinates are
expressed in terms of r and s, with the following property: if E is any point on the interior
of line segment AB, and F is the unique point other than E that lies on the circumcircles
←→
of triangles BCE and ADE, then P lies on line EF .
Solution
C D
F
P
G
A E B

3r + 1 s
The answer is P = , .
2 2
Note
that ABDC
is a parallelogram (since AB = CD = 1 and AB k CD), and let
r+1 s
G= , be the intersection of its diagonals AD and BC.
2 2
Claim: C, D, F, G are the vertices of a cyclic quadrilateral.
Proof: Assume F is on the same side of CD as E, so that the quadrilateral is CDF G.

Round 2 Solutions
www.usamts.org
It suffices to show that ∠DF C = ∠DGC. We have
∠DF C = ∠DF E − ∠CF E = 180◦ − ∠DAE − ∠CF E (because AEF D is cyclic)

◦
= 180 − ∠DAE − ∠CBE (because CF BE is cyclic)
= ∠AGB (using 4ABG)
= ∠DGC (using vertical angles at G)
(The case where F is on the opposite side of CD as E is a similar angle chase.) This proves
the claim. 2
←→
Let P be the point of intersection of this circumcircle with EF (other that F itself).
Claim: CDP G is an isosceles trapezoid.
Proof: It suffices to show that ∠DGP = ∠CP G. But this is true, again assuming that
F is on the same side of CD as E, because
∠DGP = 180◦ − ∠DF P (because DF P G is cyclic)

= ∠DAE (because DF EA is cyclic)
= ∠CDG (because ABDC is a parallelogram)
= ∠CP G (because CDP G is cyclic)
(The case where F is on the opposite side of CD as E is a similar angle chase.) 2

←→
Hence regardless of the choice of E, point P is determined on EF as the unique point
3r + 1 s
such that CDP G is a cyclic isosceles trapezoid. Chasing coordinates gives P = , .
2 2
Problems by Agustin Marchionna, Andrew Wu, and USAMTS Staff.

c 2022 Art of Problem Solving Initiative, Inc.

Solutions 34 2

Uploaded by

Copyright:

Available Formats

Solutions 34 2

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Solutions 34 2

Uploaded by

Copyright:

Available Formats

USA Mathematical Talent Search

Proof: Assume F is on the same side of CD as E, so that the quadrilateral is CDF G.

It suffices to show that ∠DF C = ∠DGC. We have

∠DF C = ∠DF E − ∠CF E = 180◦ − ∠DAE − ∠CF E (because AEF D is cyclic)

Claim: CDP G is an isosceles trapezoid.

∠DGP = 180◦ − ∠DF P (because DF P G is cyclic)

(The case where F is on the opposite side of CD as E is a similar angle chase.) 2

Problems by Agustin Marchionna, Andrew Wu, and USAMTS Staff.

You might also like