Basketball Practice

A group of 15 15 people want to play basketball. Unfortunately, there is only one ball available, so they decided to split into three teams, named A, B, and C, of 5 5 players each, and agreed on the following rules:

  1. They will play a total of 10 10 games. Each game will be played between two teams.

  2. Teams A and B will play the first game, and team C sits out.

  3. Whenever a game is finished, the losing team sits out, and the team that sat out during the game plays the following game with the winning team. Assume that there are no draws.

Suppose that the teams are equally good at playing basketball—i.e. the probability of any team winning against any other team is 1 2 \frac{1}{2} .

The expected number of games that team C gets to play (does not sit out) can be expressed as a fraction m n \frac{m}{n} , where m , n m,n are co-prime positive integers. Find m + n m+n .


The answer is 1849.

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

Sam Zhou
Sep 21, 2019

Let's consider the number of games for which each team is expected to sit out.

Let E A ( x ) E_{A}(x) denote the expected number of games for which team A sits out in the first x x games. Let E B ( x ) E_{B}(x) and E C ( x ) E_{C}(x) denote the same value for teams B and C respectively. In essence, E A ( x ) E_{A}(x) denotes the value for a team that plays the first game, and E C ( x ) E_{C}(x) denotes that for a team that sits out the first game.

Notice that E A ( x ) = E B ( x ) E_{A}(x)=E_{B}(x) and that E A ( x ) + E B ( x ) + E C ( x ) = x E_{A}(x)+E_{B}(x)+E_{C}(x)=x . This means that E C ( x ) = x 2 E A ( x ) E_{C}(x)=x-2E_{A}(x) .

It is clear that E C ( 1 ) = 1 E_{C}(1)=1 . After the first game, either team A or team B sits out, each with a probability of 1 2 \frac{1}{2} . Therefore, E C ( x + 1 ) = 1 + 1 2 E A ( x ) + 1 2 E B ( x ) = 1 + E A ( x ) E_{C}(x+1)=1+\frac{1}{2}E_{A}(x)+\frac{1}{2}E_{B}(x)=1+E_{A}(x) .

Since E A ( x ) = x E C ( x ) 2 E_{A}(x)=\frac{x-E_{C}(x)}{2} , E C ( x + 1 ) = 1 + x E C ( x ) 2 = 1 + x 2 E C ( x ) 2 E_{C}(x+1)=1+\frac{x-E_{C}(x)}{2}=1+\frac{x}{2}-\frac{E_{C}(x)}{2} .

We can then solve this recurrence relation.

E C ( x ) E_{C}(x) can be expressed in the form of q 1 k x + q 2 x + q 3 q_{1}k^{x}+q_{2}x+q_{3} .

From the recurrence relation, we can see that E C ( x ) 2 + E C ( x + 1 ) = 1 + x 2 \frac{E_{C}(x)}{2}+E_{C}(x+1)=1+\frac{x}{2}

We can get q 1 ( k x 2 + k x + 1 ) + q 2 ( 3 x 2 + 1 ) + 3 q 3 2 = 1 + x 2 q_{1}(\frac{k^{x}}{2}+k^{x+1})+q_{2}(\frac{3x}{2}+1)+\frac{3q_{3}}{2}=1+\frac{x}{2} after substitution.

It is clear that q 2 = 1 3 q_{2}=\frac{1}{3} from comparing the coefficients for x x .

The equation then becomes q 1 ( k x 2 + k x + 1 ) + 3 q 3 2 = 2 3 q_{1}(\frac{k^{x}}{2}+k^{x+1})+\frac{3q_{3}}{2}=\frac{2}{3} .

The first term of the left hand side has to be constant no matter what x x is. This means that k = 1 2 k=-\frac{1}{2} (The term is constant when k = 0 k=0 or 1 1 as well. However, both of these values lead to the result E C ( x ) = 1 3 x + 2 3 E_{C}(x)=\frac{1}{3}x+\frac{2}{3} , which does not satisfy the recurrence condition).

The equation now becomes 3 q 3 2 = 2 3 \frac{3q_{3}}{2}=\frac{2}{3} , giving q 3 = 4 9 q_{3}=\frac{4}{9} .

We know that E C ( 1 ) = 1 E_{C}(1)=1 , which means that q 1 k + q 2 + q 3 = 1 q_{1}k+q_{2}+q_{3}=1 . Substituting the values, we can get q 1 = 4 9 q_{1}=-\frac{4}{9} .

Therefore, E C ( 10 ) = 4 9 ( 1 2 ) 10 + 1 3 ( 10 ) + 4 9 = 967 256 E_{C}(10)=-\frac{4}{9}(-\frac{1}{2})^{10}+\frac{1}{3}(10)+\frac{4}{9}=\frac{967}{256} .

So the expected number of games that team C does not sit out is 10 967 256 = 1593 256 10-\frac{967}{256}=\frac{1593}{256} , giving the answer of 1593 + 256 = 1849 1593+256=\boxed{1849} .

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...