Clustering performance in the European Basketball according to players’ characteristics and contextual variables

Abstract

The aim of this study was two-fold: (i) to identify differences in the game performance profiles of basketball players in Euroleague and European national championships; and (ii) to analyse the influence of player-related (e.g. age or court-position) and contextual variables (e.g. distance travelled or game break) in players’ game-related statistics. Archival data from 2014/2015 and 2015/2016 seasons ( n = 16,268 individual game performance records) of Euroleague and European national championships were separately analyzed using a two-step cluster model and then a discriminant analysis to differentiate the obtained groups. The clustering process allowed identifying three and five different performance player profiles for Euroleague and national championships, respectively. The discriminant analysis identified differences in Euroleague playing profiles, particularly related to performers’ age and court-position. In national championships, this trend was not observed. Regardless of the competition, a relationship was perceived between players’ performance and minutes played; conversely, the players’ performance seems to not be affected by contextual variables. The current results provide evidence of different players’ profiles in elite basketball to coaches regarding the influence of player-related and contextual variables. Coaching staffs may account for this information for better management and understanding of their team composition when playing in different competitions.