Brian;

Easy enough, you just repeat the process for each prediction, taking into account date range, location range and mag range and recording the odds and hit or miss.

Try doing that for a few million random predictions.

Then we can get into the evaluation math.

Roger