Quantifying Collective Performance in Rugby Union

Objectives: The aim of this study was to quantify collective experience based on cumulative shared selections of players and to assess its impact on team performance in international rugby union. We assume that the greater the experience, the better the group will perform. Methods: Scoresheets of all games involving at least one of all 10 nations participating at the Rugby Championship and the Six Nations Championship were collected from the end of the 1999 Rugby World Cup (RWC) up to the 2015 RWC. A single indicator quantifying the cumulative shared selections (CSS, the number of selections that each player has shared with the other ones) was computed for each match as a key collective experience indicator. The World Rugby Ranking points of each nation and the percentage of victories were used to estimate team performance. The study period was divided into sequences of 4 years corresponding to the period between two consecutive RWCs. For each sequence and nation, slopes and intercept of CSS trends were computed along with victory percentage and mean ranking points. Multiple linear regression analysis was used to establish the associations between team performance and experience. Results: In regards to the CSS trends, both intra- and inter-nation variability appears to exist. Positive and negative slopes can be observed for the same team from one 4-year cycle to the next. Still, CSS slope is found to be significantly associated with both ranking points (p value = 0.042, R-2 = 0.13) and victory percentage (p value = 0.001, R-2 = 0.42). Conclusion: The evolution of the CSS that quantifies the collective experience of a team is linked to its performance. Such an indicator could be helpful in the decision-making process of national coaching staff.