Chris, when I did my 'Earthquake Dart Board' prediction run several years ago, there were two ways of looking at the results.

One way is to look at the number of quakes. In that case, there were 18,952 candidate quakes, and 17,450 of them were within a prediction window. 92.075% success rate.

The other way is to look at the predictions. There were 5,489 predictions, and 3,688 of them had at least one hit. 67.189% success rate.

Brian