Brian;

Only if they have different odds.

If you make a series of random prediction for 7+ quakes in a 3 day window (for example), a single probability is all that's necessary. But if each prediction has a different window size and/or mag range then you need individual odds.

Roger