Clustering Performances in Elite Basketball Matches According to the Anthropometric Features of the Line-ups Based on Big Data Technology.

Abstract

The aims of this study were: 1) to conduct a descriptive analysis of the anthropometric features of the line-ups of strong teams (top 16) in the 2019 FIBA Basketball World Cup; 2) to group the line-ups mentioned above into different clusters based on their average height, weight, and body mass index (BMI); and 3) to explore the performance variables that discriminate between various line-up clusters. The play-by-play statistics were collected from 104 team objects in 67 games and 525 line-ups were analyzed using two-step cluster and discriminant analysis. Line-ups were classified into four groups: low average height and weight with middle BMI (LowH–LowW–MiddleBMI); high average height and low average weight with low BMI (HighH–LowW–LowBMI); low average height and high average weight with high BMI (LowH–HighW–HighBMI); high average height and weight with middle BMI (HighH–HighW–MiddleBMI). The results of the discriminant analysis demonstrated that LowH–LowW–MiddleBMI line-ups had the least time played and the lowest offensive rating, but the best offensive rebounds, turnovers, and fastest game pace performance; HighH–LowW–LowBMI line-ups demonstrated the best defensive rating but performed poorly with a low value of assists and a high value of turnovers; the LowH–HighW–HighBMI group achieved the best time played statistics but had the lowest number of free throws made; the HighH–HighW–MiddleBMI group had a higher number of assists and a higher offensive rating and 2-point field goal performance, while also achieving the lowest number of offensive rebounds and ball possessions. These results provide novel insights for coaches and performance analysts to better understand the technical characteristics of different line-ups in elite basketball competitions.