Evaluators should note in their risk assessment reports that researchers have found only moderate effect sizes regarding the accuracy of existing risk assessment tools and, therefore, that predictions of future sexual offending are, on average, better than chance but can be incorrect