Like many areas of psychology, aggression research has been impacted by psychology's replication crisis. Until recently, aggression was understood largely from the perspective of social cognitive models, wherein aggression was conceptualized as learned scripts. However, more recent rigorous research has suggested that cognitive scripts may play only a small role in aggression. By contrast, genetic and evolutionary inputs are powerful and combine with both developmental and situational life stress to increase the risk for serious aggression. This article examines aggression from the Catalyst Model and how diathesis-stress approaches to understanding aggression can help reduce error in predicting clinically significant aggression. This article also addresses related issues of measurement error and smallest effect size of interest (SESOI) as they relate to aggression research.