One way of approaching this might be to look at the Flinn-Engdahl region of the
quake/forecast. If they occur in the same region, probabilities could be calculated.
If in different regions only calculated if within certain distance criteria.
Normally, if someone told me there was going to be a moderate earthquake
off the coast of Northern California on a certain date and an event did occur
in that region on that date, I would not reject immediately from consideration
that forecast without checking further - just a kind of an intuitive thing.