Roger:

Actually the first priority, IMO, is to have the ring probabilities for just one earthquake (i.e. no ring overlaps from others).

In this way, the background probabilities (per mag ranges) for the entire set of rings will be known and will be the benchmark against whether or not any prediction using FFA rings exceeds random chance.

For example, lets say that somebody predicts mag 5-5.5 to happen somewhere in the FFA rings after a strong earthquake within say 10 days or so. The background probability using all the rings will give the "random chance" probability.

Then we'll have a better handle (especially over time) on whether or not the FFA concept (with the specified ring distances) is statistically significant (over time) as a method of prediction.

Of course, Lowell has been producing statistics regarding the FFA concept and has been accumulating them. The tool your building will also allow a predictor to tweak a prediction to adjust the probability of the event (to < 5 % for example...).

The ring overlap discussion is a related matter but secondary at this point in time, IMO.

$.02