It may be that in order to arrive at technique for predicting the "big" ones, it has
to be proven with the "small" ones first. There is certainly value in predicting large
earthquakes in a cultural sense, but there is value in predicting small ones in a
scientific sense. I think for the initial stages, unadulterated probability should
be the main determining issue. But clearly, if someone predicts an 8 and it happens
within reasonable limits, we will all be more interested in how that was done
than if someone predicts a 2 and it happens. But if someone predicts 10's or
100's of 2's correctly, then that may lead to a technique whereby the big earthquakes
can be forecast. Don't throw out the baby with the bathwater. This is not a problem
that will be solved overnight.