Repositorio Académico

The aim of the study was to identify the basketball performance indicators that best discriminated winning and losing teams according to game location and final score differences. The sample used was composed by 306 games of the ACB basketball league from the 2007-2008 regular season. A cluster of k-means was used to split the sample into 3 groups: balanced (differences between 1 and 12 points); unbalanced (differences between 13 and 28 points) and blowout games (differences above 28 points). The results obtained from the discriminant analysis showed that home teams were discriminated from away teams by assists (SC=.41), successful 2-point field goals (SC=.40) and defensive rebounds (SC=.36) during balanced games when home teams won. Conversely, when visiting teams win, assists (SC=.33) and steals (SC=.42) discriminated home and away teams. In unbalanced games, when home teams win, home and away teams were discriminated by assists (SC=.30). On the other hand when away teams win, home teams were discriminated from away teams by assists (SC=.31) and 3-point field-goals (SC=-.33). These findings may help coaches to prepare and design specific programs focused on game dynamics.