The previous chapter discusses anomalies that arise from conflicts with our predictions about actors’ plan choices. However, we also depend on many other types of predictions. In order to anticipate others’ actions and their effects, we predict the course of steps in a chosen plan and also predict the plan’s final outcome. To guide our behavior and control our environment, we predict the course of chains of events in the physical world and predict the behavior of devices. In order to decide when to believe new information, we predict whether information is likely to accurate. When the world deviates from any of these expectations, we need to explain, to predict better next time. For explanation to be feasible, we need to focus the explanation effort: Our anomaly vocabulary must characterize these types of anomalies as well.