Brian;

> But, aren't we just talking about the odds of a single prediction being
right?

Yes.

> OK. 10 coin tosses. 5 heads. 5 tails.

> How does the distribution of the results affect the odds of any one toss?

It doesn't but that isn't relevant to this case.

Read my explanation again. The odds are not a function of the number of quakes but are of the number of predictions that are correct.

Take one day predictions. Suppose there are 365 days containing 600 quakes. What are the odds that a particular day will have a quake?

The answer is that you don't know. Due to clustering, the odds are anywhere from 1/365 to 365/365.

Think about it.

Roger

Roger