On the validity of team performance indicators in rugby union

Abstract

We conduct a validation study to determine which historically statistically significant rugby union team key performance indicators (KPIs) continue to be valid differentiators between winning and losing teams. We examine KPI validity within and among five competitions across 2013, 2014 and 2015 by determining which KPIs continue to achieve statistical significance, as well as provide actionable information to stakeholders (coaches, players, teams, performance analysts, researchers). While most of the historically significant KPIs differentiate between winning and losing teams in at least one competition, few do so across all the competitions considered. We argue that the handful of KPIs that reliably differentiate between winning and losing teams across competitions do not yield much actionable information for stakeholders. We discuss some contributing factors to this result and other important methodological issues in KPI performance analysis in rugby union