Space and Some Other Things: Point Process Models for Hockey Data

Statistical methods for the analysis of baseball are well established. Each game is essentially a series of trials with a finite number of setups and outcomes. Each play begins the same and each player can only make one decision. After the play, all of the players are reset and another trial begins. Because of this, there are established statistical methods that can be directly applied. These methods have lead to a revolution in the study of baseball (e.g. James (1987)).