Brian;

I'm back.

> Perhaps when Roger get's back I'll talk to him about the possibility of doing evaluations with fuzzy boundaries. I can see in my mind one way to do it, but even then, there are still clearly defined rules.

I have such a program now. BESTFIT checks for quakes which exactly fit, but also for a range of each parameter. it selects the quake which is closest in all parameters and in each individual parameter as well. It's intended to be a training tool, showing what you should have predicted.

But for real evaluations, I use what is predicted. So if you say 5.6-6.2 then 6.3 or 5.5 is a miss. If you mean 5.2+ then say so.

Roger