Abstract
Australian Rules Football is a field invasion game where two teams compete to score the most points. While complex machine learning models can predict match outcomes post-game, their lack of interpretability limits understanding of the factors influencing team performance. Using data from the male Australian Football League (2015–2023), we estimate team strengths and their determinants by fitting flexible Bradley-Terry models. We identify teams significantly stronger or weaker than average, with stronger teams placing higher in the previous season’s ladder and leading in Forward 50 activity, goal shots, and scoring. Playing at home consistently creates an advantage, regardless of strength. When used for forward prediction, models incorporating team-specific, time-variant features correctly anticipate up to 71.5% of outcomes. Our approach thus enables interpretable strength estimation and competitive forecasting, supporting data-driven strategies and training.
Get full access to this article
View all access options for this article.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
