canie;

Yes, we have discussed this before.

I do have a "best fit" program which will become part of the one I'm thinking about. It selects the best fitting quake for each parameter separately. That might mean the same quake in each case (or not). The question could then becomes one of probability; which one is the least likely?

Or maybe I should just list all 3 and let you take your pick?

No, because the idea is to obtain a cumulative score which reflects the predictor's ability, the idea being to eliminate the "shotgun" approach and the "count only hits" types and the "sure thing" group so we can focus on those with real ability.

Like Don.

Or maybe the whole thing is a waste of time.

Roger