Table 1.
Game-related statistics.
Fig 1.
A schematic representation of the procedures adopted for the analysis conducted in this paper, illustrating the potential application of the proposed method in a real-world setting.
Table 2.
Distribution of demographic and performance characteristics across four functional classes in the considered regular season.
Fig 2.
Profile plots of seven game-related statistics normalized by play time.
The dashed blue line represents the midpoint between the minimum and maximum values. First-round championship data was used for the analysis. PTS, total points made per match (PTS = FTM + 2*P2M + 3*P3M); P2M, number of two-point field-goals made; P3M, number of three-point field-goals made; FTM, number of free-throw points made; REB, number of rebounds; AS, number of assists; ST, number of steals.
Table 3.
Distribution of demographic and performance characteristics across subgroups identified by cluster analysis applied to data of the first round.
Fig 3.
Distribution of athletes across the four functional classes within the two clusters.
Table 4.
Linear regression models evaluating the association between team’s net performance as the outcome variable, and the minutes played by players of the cluster (Model 1) and in the functional classes (Model 2).
Fig 4.
Performances (plus-minus) of the teams in the second round according to the minutes played by the players belonging to the two clusters identified in the first round.
The bars denote the proportional composition of each team’s clusters, arranged in ascending order according to the percentage representation of Cluster 2. The grey line depicts the teams’ net performance (PlusMinus), presented in the same order as the bars.
Fig 5.
Performances (plus-minus) of the teams in the second round according to the minutes played by the players in the four functional classes.
The grey line depicts the net performance (PlusMinus) of the teams.