Thanks, Roger,

You wrote: "Right off the bat you need to rule out single predictions. Otherwise someone will enter 1000 tight predictions for mag 7.8+ quakes and claim victory when one hits."

See-that's where I need the professional help (OK -make your jokes!!) - To me, it seems, if somebody made "1000 tight predictions," and only one (or a few) succeeded, then the large number of predictions would work against their having achieved a statistically significant result. In other words, it would be entirely fair for them to do exactly as you suggest they might.

I assure you, Roger, I have no reason to exclude you as the analyst - other than the fact that you've already been highly controversial (due to others simply not liking your results - nobody has made any specific complaints as to your methodology), and I want to minimize accusations of bias. I am considering posting to a mathematics newsgroup for their input on whether such analysis as I propose is feasible, what the methodology should be, and who might be a good candidate for analyst.

Could you explain here, or direct us to a link to explain, just exactly what your methodology is?

Michael F. Williams
Arroyo Grande, CA USA