Major Probability Question 2.

Probability Level pending

Three persons A , B , C A,B,C play the following game: a subset with k k elements of the set { 1 , 2 , . . . , 1986 } \{1,2, ... , 1986\} is selected randomly, all selections having the same probability. The winner is A , B , A, B, or C C according to whether the sum of the elements of the selected subset is congruent to 0 , 1 , 0, 1, or 2 2 modulo 3 3 . Which of the values listed for k k gives A , B , C A,B,C an equal probability of winning?

7893 9000 6002 8991

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

1 solution

Alan Yan
Oct 24, 2015

Lemma: \textbf{Lemma:} If p is a prime number and \textit{If p is a prime number and} a 0 , a 1 , . . . , a p 1 a_0, a_1, ..., a_{p-1} are rational numbers satisfying \textit{are rational numbers satisfying} a 0 + a 1 ϵ + a 2 ϵ 2 + . . . + a p 1 ϵ p 1 = 0 , a_0 + a_1\epsilon + a_2\epsilon^2 + ... + a_{p-1}\epsilon^{p-1} = 0, where \textit{where} ϵ = e 2 π p i , \epsilon = e^{\frac{2\pi}{p}i}, that is the primitive root of unity, then \textit{that is the primitive root of unity, then } a 0 = a 1 = a 2 = . . . = a p 1 . a_0 = a_1 = a_2 = ... = a_{p-1}.

Proof: \textbf{Proof:}

Observe that a 0 + a 1 x + a 2 x 2 + . . . + a p 1 x p 1 a_0 + a_1x + a_2x^2 + ... + a_{p-1}x^{p-1} and 1 + x + x 2 + . . . + x p 1 1+x + x^2 + ... + x^{p-1} are not relatively prime since they have a common root. Since the latter is irreducible over Q \mathbb{Q} , 1 + x + x 2 + . . . + x p 1 1 + x + x^2 + ... + x^{p-1} must divide a 0 + a 1 x + . . . + a p 1 x p 1 a_0 + a_1x + ... + a_{p-1}x^{p-1} . which can only happen if a 0 = a 1 = . . . = a p 1 a_0 = a_1 = ... = a_{p-1} . We have proven the lemma.


Denote a k ( j ) a_k^{(j)} to be the number of k k element subsets such that the sum of the elements is congruent to j j modulo 3 3 . We will also define ϵ = e 2 π 3 i . \epsilon = e^{\frac{2\pi}{3}i}.

It is pretty clear that j = 0 2 a k j ϵ j = 1 c 1 < c 2 < . . . < c k 1986 ϵ c 1 + c 2 + . . . + c k \sum_{j = 0}^{2}a_k^{j}\epsilon^{j} = \sum_{1 \le c_1 < c_2 < ... < c_k \le 1986}\epsilon^{c_1 + c_2 + ... + c_k} since they are both counting the same thing. However, notice that the right hand side is the x 1986 k x^{1986 - k} coefficient of A = ( x + ϵ ) ( x + ϵ 2 ) . . . ( x + ϵ 1986 ) . A = (x + \epsilon)(x + \epsilon^2)...(x + \epsilon^{1986}). We know that x 3 + 1 = ( x + ϵ ) ( x + ϵ 2 ) ( x + ϵ 3 ) x^3 + 1 = (x + \epsilon)(x + \epsilon^2)(x + \epsilon^3) , so we can greatly simplify the product A = ( x 3 + 1 ) 662 A = (x^3 + 1)^{662} Thus, we can find the x 1986 k x^{1986 - k} coefficient. Notice that if the coefficient of the x 1986 k x^{1986 - k} term is t t , then we know that by the lemma, one of the below cases must be fulfilled a k ( 0 ) t = a k ( 1 ) = a k ( 2 ) a k ( 0 ) = a k ( 1 ) t = a k ( 2 ) a k ( 0 ) = a k ( 1 ) = a k ( 2 ) t \begin{aligned} a_k^{(0)} - t & = a_k^{(1)} = a_k^{(2)} \\ a_k^{(0)} & = a_k^{(1)} - t= a_k^{(2)} \\ a_k^{(0)} & = a_k^{(1)} = a_k^{(2)} - t \end{aligned} However, we want to find the k k where a k ( 0 ) = a k ( 1 ) = a k ( 2 ) . a_k^{(0)} = a_k^{(1)} = a_k^{(2)}. Clearly, this occurs when the x 1986 k x^{1986 - k} coefficient is zero. Now notice that in A A , there only exist terms with exponents divisible by three. That means that the coefficient is zero whenever 3 k 3 \nmid k and the only value that follows that condition among the answer choices is 6002 \boxed{6002} .

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...