You still didn't own up to the fact that it was you who said the PDE had a greater magnitude of completeness. I realize that this isn't what you intended to say. But it IS what you said and by the way, my name is Dennis, not David.

As to magnitude scales, it all depends on the size of the quake. Local magnitude isn't a good source for the larger quakes where Mw and even Ms would be a better indication of the quakes size as opposed to Ml. But then a lot of the quakes on the PDE were being reported with the body wave Mb scale (which isn't all that bad either).

As to the issue on M4 quakes occurring everyday, is that the scientific way of getting a point across by skewing facts through a speach mannerism? If that wasn't your intention, lets just keep it to the actual facts. Thank you.

Now lets get to the every 7.8 days. During my days of posting predictions on sci.geo.earthquakes, we got into a discussion on how probability should be calculated. Also, the data I had presented a week or so ago from Berkeley was calculated after declustering of the catalog (meaning that aftershocks were not part of the result set used in determining probability). The discussion on SGE included people from a variety of fields including geologists, seismologists, geophysicists, statisticians, etc. After much discussion, Alan Jones had come up with a method (though still not perfect) of calculating probability. Since the catalog Alan was using wasn't declustered, the method he came up with, and to which everybody agreed to (at least for the time being), was to calculate how many windows had a hit over the period of the catalog he was using. I think that the data he had was for about 20 years or so. At any rate, the window was the time window of the prediction and he would see how many windows during that 20 year period had a hit for the predicted magnitude in the area of the prediction. This method still doesn't handle all aftershocks but then does miss some quakes that should have been counted. So it evens out. The NOA catalog goes from 1963 - 2000 (or 1999 if we want to use complete years) so in Antonio's case, with his window being 11 days for an M4+ event anywhere in Greece, how many of the 11 day windows would be satisfied with an M4+ event. For 1963 - 1999 we have 13,514 days (9 leap years) or 1,228.55 (rounded) eleven day windows.

Another point that averages don't show, that the above formula partially helps with, is that there is a slight clustering of quakes. Meaning that they don't occur every 7.8 days on average. I haven't looked over the NOA to check this, but I wonder how many periods there are with no M4+ events in periods longer then 11 days? What is the longest period? Hopefully you understand my point here.

Dennis