[SIPTA] interesting discussion on use of imprecisely known probabilities in earthquake prediction