Yes, Roger, they do. But I see in that a seed of future obfuscation - not what you want when you're attempting to design a scientific experiment. The predictor's upper bound can go in the notes. The form can disallow the entry in the parameters section.

You know darn well that, if a very large quake fulfills a predictor's time and location parameters but is above a moderate specified range, he or she is going to howl if the failure of that prediction keeps the method from scoring above chance - and maybe rightfully so.

Meanwhile, a hit within a narrow range of moderate magnitudes will boost the significance rating of the overall score (unjustifiably?) higher. Since the occurrence count drops off very rapidly as magnitude increases, the inclusion of ranges might skew the results, and, at the very least, complicates the analysis.

Mike Williams