Reply to comment on "Assessing CN earthquake predictions in Italy"

Abstract

Molchan et al. [2018] raised concerns on the reliability of the main Taroni et al.’s [2016] conclusion that reads “Considering the data available so far, the Molchan Test does not show that CN prediction performance is significantly better than predictions based on the stationary Poisson model.” In particular, Molchan et al. [2018] discuss two main issues: 1) the Taroni et al.’s [2016] results are based on too few data to achieve robust conclusions, and 2) the parimutuel gambling score (PGS) produce unfair results in comparing predictive models. We thank Molchan et al. [2018] to give us this opportunity to clarify further some aspects of our paper, but we anticipate that we do not see any compelling reason to modify our original conclusion.