A0431
Title: An explainability approach for identifying winning strategies in rugby union
Authors: Sebastien Dejean - University of Toulouse (France) [presenting]
Arnaud Odet - University of Toulouse (France)
Cristian Pasquaretta - University of Toulouse (France)
Abstract: Interest in predicting sports match outcomes has grown significantly. However, the utilization of predictive models in enhancing tactical team performance remains relatively limited. A methodology is proposed that combines machine learning and algorithm explainability techniques. The case study on rugby union unfolds in two phases. First, the best modeling approach is identified for the data by establishing a prediction model based on performance indicators observed during games. Then, SHapley Additive exPlanations (SHAP) values are used to interpret the predictions of this model. Findings serve three purposes: (i) from a global standpoint, identifying performance indicators that potentially determine match outcomes; (ii) from an aggregated point of view, highlighting strengths and weaknesses of any given team; and (iii) from a local perspective, offering technical staff diagnostic analyses of past games.
