Amit;

Suppose there were enough quakes that we had one every day. What are the chances of a hit then, even if you make only one prediction that year?

That's the flaw in your reasoning.

I determine the odds by dividing the year up into 3 day windows and counting how many of them contain quakes. That determines the odds on a window chosen at random containing a quake. And for the first quarter of 2010 it's 80%

Roger