Thread: Simple but annoying probability question

1. Simple but annoying probability question

A, B and C are in three way race - If the probability of A beating B is p, and the probability of A beating C is q, what is the probability of A beating B and C?

It sound like it should be pq. But if if the three are equally good, then p = 0.5, q = 0.5 But the probability of A winning is 1/3, not 1/4.

What is the general answer for p and q?

2. Re: Simple but annoying probability question

P[A beating B and C] < min(p,q)

3. Re: Simple but annoying probability question

Thanks Mean Joe. But I was hoping for some way of actually calculating the answer. Or are you saying that there is no general answer? kat

4. Re: Simple but annoying probability question

I don't think there is a general answer, as I'd also have to consider the probability of B beating C.
Even if I knew that, these kinds of questions are tough.

5. Re: Simple but annoying probability question

Consider your original problem with the the racers that are equally good. I put a magic spell on b and c such that they always finish in a tie with each other. In this situation you can see that the probability that A bears both of them is just 0.5. We already saw a case where the probability is 1/3. So from this we can see that at the very least we don't have enough info to get a definitive conclusion.

6. Re: Simple but annoying probability question

I like that concrete example, Dason, and having seen that it's not hard to arrange for any value between 1/2 and 1/3. Cheers, kat

7. Re: Simple but annoying probability question

Mathematically, on average the probability is 1/3. Note that X1 and X2 are iid U~[0,1].

and

.

As such:

8. Re: Simple but annoying probability question

Or visually, for the equal likelihood case, you can imagine a 1x1x1 cube for AxBxC with a random point inside. The volume where A beats both B and C is a tetrahedron of volume 1/3.

9. Re: Simple but annoying probability question

Yes, in short, the right-hand side of the three expressions associated with the last equation I provided will yield probability density functions that provide expectations of: 1/2, 1/2, and 2/3, respectively (from left to right).

Hence, given these expectations, we have: 1/2 + 1/2 - 2/3 = 1/3 as the expected value; which is what you should be looking for, katxt.

 Tweet

Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts