I'll respond to the part addressed to me.

> With regards to Rogers old database of predictions: Roger? any comments? Remember what happened to them?

I have no certain knowledge.

The programs would have been archived in the tape library but that is eventually recycled

The original letters, magazine articles, etc. were in a file cabinet in my office. They may have been sent to Federal archives or may have been destroyed.

A tragic loss IMHO, but I wasn't asked!

> You mentioned that confidence level is used on your system. Could you describe how that parameter is applied to your analysis? While on the subject, how is analysis done (in general)? Roger may have more information on that subject ....

No, I don't know how it is done at NGDC, didn't even know it was being done there until now.

My program calculated a form of probability when evaluating predictions. I don't have a program listing and don't recall all the details but I do know it calculated e^(-seismicity) as part of it.

I don't know why the predictor's confidence level is used anyway. It's just an escape clause to explain failure as far as I can see.

Roger