Hi all;

just out of curiosity I checked Amit's predictions for date only hits, where any quake in his window of mag 6+ was a hit.

As I said before, chances of success were 0.698 and he got 147 hits out of 151 predictions but his score was still negative (-2.913).

This is because hitting a high probability event isn't worth much but missing one hurts a lot.

One weakness of this method is that it can't handle unlikely hits properly.
Getting a hit where the probability is zero gets such a high score it can't be overcome. Such a case can happen when the NEIC catalog contains no quakes for a particular area and one does happen there. The Myanmar quake was such a case.

Roger