I think the key to Alan's statement (correct me if I'm wrong Alan) is:

"However, some folks think that earthquake predictions can be
evaluated by computing what percentage of predictions have a hit."

Berkland is predicting percentages, not earthquakes.

The evaluation of earthquake predictions have to include an analysis of the prediction(s) occuring by random chance for the area, and I see no such analysis with Berkland's predictions. Alan's currently batting an excellent 100% for his prediction, but, that doesn't mean anything as the quakes would have occured anyways by random chance. Berkland is not comparing his predictions to random chance, which is why they are virtually useless. He is only comparing his predictions to himself, and lets face it, if you can't prove a significant improvement over random chance, your not predicting anything.

The quality of the prediction is everything.

An example of a useless prediction is Alan's example.

An example of a usefull prediction is an 8.0+ within 10 miles of downtown LA in the next 6 hours.

Anyways, that's how I see it. I'm not the michael you think I am, but I do live in SCV :)

Michael