Right. Where my proposal had limits to it.

I don't see the reason for the log part of it either. Sure the mags are logarithmic, but is what the predictor is using logarithmic as well? Don't know.

Also, the predictor is trying to come as close as he or she can to something that is subjective where for most quakes, the average of all readings is used for the final magnitude. In addition, there are several magnitude scales reported for each quake. I don't think that there is any correct scale to use since they are all measuring something different about the quake - after the fact. The predictor is using/measuring something that occurs prior to the quake which is entirely different. So we're comparing apples and oranges. Maybe some other method can eventually be decided on.