Hi Roger. Thanks for your response. I'm still trying to think of a way to use probability to determine the most likely location for large events. You have to admit, it would be much more useful than any current use of the probability calculation if it could actually predict a large earthquake.

Although you mentioned that it is used to determine if someone can predict better than chance, there are scores of examples on the internet of probability being used to prove that no one is better than chance.

Can't you do a probability of a large event for typical locations to see which one is the most probable? Forget about any locations that are not occupied by humans for now. Maybe you could come up with a short list of major cities, calculate the probability for each and show it in a chart form. No matter how small the probability, it still might indicate a trend. If so, you would really be on to something.

Keep an open mind and anything is possible. I have a spreadsheet with locations reported by the USGS. Although I am sure that the USGS does not report all events worldwide