Hi all;

This latest go-round has pointed up some problems, aside from my programming errors.

These are twofold and insoluable.

1) No one catalog is totally complete.

2) The answer you get depends on the dataset you use.

I can't seek out all possible sources of location information the way Lowell has. I don't have the time or resources. The best I can do is the CNSS catalog which is about as complete as there is. But the recent contest has shown it does lack some information.

For probability calculations I use the CNSS data for 1997-2001. If the catalog was longer the answer would be different. If I used another source such as NEIC, the answer would be different. There's not much I can do about it. Making it longer would make it slower and not necessarily better.

So what it comes down to is that this is the best I can do as a practical matter. Bear in mind too that the smaller the quake you predict, the less likely the data will appear in the catalog and the worse my answer will be. That's just the way it is.

I'll continue doing the evaluations for anyone who wants it but it should be considered as an aid and not as a final answer.

Roger