Back in 1995 and 1996, Dennis made lots of predictions in the sci.geo.earthquakes forum. Immediately after each prediction, I posted the probability of success by chance based on catalogs and, if there had been a recent main shock, on the probability of an aftershock in at least one of his windows.

I have not been following the forum much lately due, in part, to the lack of clean predictions. A clean prediction is one that states the time window, the space window, and the magnitude window. Dennis has been consistent in doing this and hardly anyone else has.

For his latest predicition, I have computed that, based on historical catalogs, that the probabability of success is only 0.4%. However, as Roger Musson has pointed out, due to the Hector Mines event, the probability is much higher. In fact, I compute it as 99.9%. This is based on a paper in Science by Paul Reasonberg and Lucy Jones (no kin).

Dennis' first region (101 miles) does not extend to the Hector Mines aftershock region but his other two do.

If folks want, I will post the method I use to evaluate earthquake predictions.

Plainly stated, I attempt to compute the probability that a prediction will come true base only on chance.

And, as expected, there have been no events in his first window (it was too small) and several in the other ones:

1999/10/18 06:35:47.0 34.3500N 116.1500W 1.4 4.6
1999/10/18 11:02:20.0 34.8500N 116.3500W 6.0 4.4
1999/10/18 19:01:59.0 32.1900N 115.1800W 5.0 4.5
1999/10/19 12:20:44.0 34.7100N 116.3400W 5.0 4.6
1999/10/18 06:35:47.0 34.3500N 116.1500W 1.4 4.6
1999/10/19 12:20:44.0 34.7100N 116.3400W 5.0 4.6

Alan Jones