A deterministic prediction model for the American game of football

This article describes a simulation model of the sport known as football. It was created to predict results of post season football games, most notably college bowl games. By constructing a totally deterministic model of the game, one in which there are no random events, and by using data that is directly accessible to represent the participants, an objective prediction can be reached. The remainder of this paper is divided as follows: Section 2 discusses more about the model's internal workings; Section 3 describes the all-important play selection strategy; Section 4 presents the results collected for verifying the model's accuracy; and finally Section 5 presents some observations and shortcomings of the model along with possible directions for further investigation.