Visualizing Rugby Game Styles Using Self-Organizing Maps

Abstract

Rugby coaches and analysts often use notational data describing match events to assess their team's performance and to devise strategic plans for upcoming matches. However, given the volume and complexity of the data available, it is difficult for them to recognize high-dimensional relationships among the available performance variables. A nonlinear approach using self-organizing maps (SOM) can help visualize the performance of a team and its opponents as well as the subsequent suitability of certain game styles, given the style of the opponent.