I have a set of hypothesis. Each hypothesis claims of probability of an event. For example:
- p(A) = 0;
- p(A) = 1/2;
- p(A) = 1;
- p(A) = 0.
Does the whole probability of A equal (0 + 1/2 + 1 + 0) / 4 = 0,375? All hypothesis are independent.
We can disregard the "physical nature" of these hypothesis and treat p(A) as a discreet random variable. This way you have a probability distribution specified and you can, for example, calculate expected value as (0*2+1/2+1)/4=0.375. But that does not mean that value of p(A) (probability of event A) is always equal to 0.375, it's just an expected value and p(A) is still a random variable.