Hi Roger,
Thanks for the warning regarding the problems inherent in your sampling technique.
Those who make predictions will be happy to have this aid. It is certainly a giant
step in the right direction, albeit less than perfect.
I think for contests on this board, the decision of the judges should be allowed
to be queried in case the earthquake is found in a catalog other than CNSS, otherwise
the decision of the judges should be considered final. If the earthquake is found
in another catalog but not in CNSS, no probability can be assigned, and thus the
forecast cannot be completely evaluated. In this case, the winner will remain the
one whose event was in the CNSS catalog with the caveat that only CNSS events
will be assigned probabilities.
Hence a new rule for the contests:

Only events found in the CNSS global catalog will be considered for evaluation
of forecasts/predictions.