As the title suggests, this is a discrete random variable with a countably infinite number of possibilities. I am trying to find the number of coins flips it takes to get a least 1 heads and 1 tails. So my first flip is either heads or tails, then my second flip is heads or tails, and so on. So my sample space would be {(H,T),(T,H),(T,T,H),(H,H,T)...} So it looks there is 1/2 chance that my experiment ends in 2 flips, 1/4 it ends in 3 flips, 1/8 it ends in 4 flips, and so on. So my PMF would seem to take the form of $p(x)={1/2^{n-1} , n=2,3,4,5,...}$ and 0 otherwise.

So the expected value would be $E[X]=\sum_{n=2}^{\infty}n{\frac1{2^{n-1}}}$

So to solve this I let m=n-1 to get $E[X]=\sum_{m=1}^{\infty}(m+1){\frac1{2^{m}}}$

I feel like this should equal 3 but I am not sure exactly how to express that.

  • $\begingroup$ Your question is not clearly posed. You are asking for a PMF, expectation, and variance--which apply to a random variable. But "one heads and one tail" is not a random variable; it is an event. Moreover, this event is not clearly defined because you've not stated how many times the coin is flipped. $\endgroup$ – heropup Sep 29 '19 at 23:59
  • $\begingroup$ Hero, you are flipping the coin up to an infinite number of times. You stop when you have flipped a heads and a tails. $\endgroup$ – mattsprestige Sep 30 '19 at 4:58
  • $\begingroup$ And what is the random variable of interest? The total number of coin flips? $\endgroup$ – heropup Sep 30 '19 at 5:18
  • $\begingroup$ @heropup I worked on the problem last night and I think tried to clarify the experiment. Hopefully, it is more clear what I am trying to do $\endgroup$ – mattsprestige Sep 30 '19 at 17:38

Continuing from your working:

\begin{align} \sum_{m=1}^\infty \frac{m+1}{2^m} &= \sum_{m=1}^\infty \frac{m}{2^m} + \sum_{m=1}^\infty\frac{1}{2^m} \\ &=\sum_{m=1}^\infty m (0.5)^{m-1}(0.5) + \frac12 \cdot \frac{1}{1-0.5}\\ &= \frac1{0.5}+\frac12 \cdot \frac{1}{0.5} \\ &= 3 \end{align}

The first term is due to that is the expression of the expected value of geometric distribution with success probability $0.5$ and the second term is just a geometric sum.

A faster working is actually identifying it as $$1+X$$

where $X$ is a geometric distribution with success probability $0.5$. $1$ comes from the first toss and the success event is to get a different outcome from the first toss. From there you can obtain the expected value, pmf, and the variance.

We have $E[1+X]=E[X]+1$ and $Var(X+1)=Var(X)$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.