Keep it bouncing

Assume a frog is hopping between two leafs each second, with probability p p to jump to the second leaf and 1 p 1-p to stay. Let p N p_N be the probability that after N seconds the frog is in the same starting position. Compute the probability distribution of p N p_N , and solve for p = 1 3 p=\frac{1}{3} , and N = 5 N=5 .

Assumption: The probability of jumping between leaves is independent of the past.


The answer is 0.502.

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

Chris Lewis
Jun 18, 2019

We need the probability that the frog makes an even number of jumps in order to return to its original leaf.

The probability that the frog makes exactly k k jumps is ( N k ) p k q N k {N\choose k} p^k q^{N-k} , where q = 1 p q=1-p . This is the k th k^{\text{th}} term of the binomial expansion of ( p + q ) N (p+q)^N :

( p + q ) N = k = 0 N ( N k ) p k q N k = ( N 0 ) p 0 q N + ( N 1 ) p 1 q N 1 + ( N 2 ) p 2 q N 2 + (p+q)^N=\sum_{k=0}^N {N\choose k} p^k q^{N-k} = {N\choose 0} p^0 q^{N} + {N\choose 1} p^1 q^{N-1} + {N\choose 2} p^2 q^{N-2} + \cdots

To get the terms for even k k , we compare this to the binomial expansion of ( p + q ) N (-p+q)^N :

( p + q ) N = k = 0 N ( N k ) ( 1 ) k p k q N k = ( N 0 ) p 0 q N ( N 1 ) p 1 q N 1 + ( N 2 ) p 2 q N 2 (-p+q)^N=\sum_{k=0}^N {N\choose k} (-1)^k p^k q^{N-k} = {N\choose 0} p^0 q^{N} - {N\choose 1} p^1 q^{N-1} + {N\choose 2} p^2 q^{N-2} - \cdots

We can cancel out the terms with odd k k simply by adding the two expansions; hence

2 p N = ( p + q ) N + ( p + q ) N 2p_N = (p+q)^N + (-p+q)^N

Of course, p + q = 1 p+q=1 , so this simplifies to

p N = 1 2 ( 1 + ( 1 2 p ) N ) p_N=\frac{1}{2} \left(1+(1-2p)^N \right)

For the particular question, we have p 5 = 1 2 ( 1 + ( 1 2 3 ) 5 ) = 0.502 p_5 = \frac{1}{2} \left(1+\left(1-\frac{2}{3} \right)^5 \right) = \boxed{0.502\ldots} .

Alternatively, we can set up a recursion: p 0 = 1 p_0=1 , and p N + 1 = ( 1 p ) p N + p ( 1 p N ) p_{N+1} = (1-p)p_N + p(1-p_N) (we just consider the two possible positions of the frog after N N jumps). This is somewhat simpler than the first approach!

As we'd expect, as N N \to \infty we have p N 1 2 p_N \to \frac12 .

Great answer man!

Manifold M - 1 year, 11 months ago

Log in to reply

Thanks! And thanks for posting the problem. How did you do it? Did you have another way?

Chris Lewis - 1 year, 11 months ago

Log in to reply

Yes, I'll surely post it later on :) It just another way to express the sum of even values of k using some properties of probability distribution! It's always fun to see other solutions to the same problem.

Manifold M - 1 year, 11 months ago

Log in to reply

@Manifold M Absolutely agree. I got greedy and posted two ways (well, one and a half), it would be great to see another.

Chris Lewis - 1 year, 11 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...