Your point is taken on the 3.0 example. My prior counter to that was that it would take many hits on a 3.0 prediction to become statistically significant. I also don't think that people would take too much notice on somebody hitting a bunch of 3.0 predictions in a seismically active area.