Question for all;

Which is more important, a very low probability prediction of a mag 2.0 or an equally low probability prediction of a mag 8.0?

It seems to me that the 8.0 is much more important to predict but how do we "score" predictions to reflect that, given both are equally unlikely?

Roger