Predicting the Brownlow Medal winner

The Brownlow medal is the highest individual honour that can be achieved by Austra-lian Football League (AFL) players. It is based on the umpires votes to the three best players (3 for first, 2 for second, 1 for third) in each of the 176 home and away matches for a season. An ordinal logistic regression model retrospectively applied to past data has been used to identify specific player performance statistics from each match that can aid in the prediction of votes polled. By applying this model to present data it is possible to objectively assign leading players a probability of winning the Brownlow medal. Over the past two AFL seasons (2000 and 2001), the authors have successfully used this approach to identify the leading contenders for the Brownlow medal.