The emerging basketball discipline: unpacking game outcomes in the 3 × 3 basketball professional league based on performance indicators and contextual variables

Abstract

ABSTRACT 3 × 3 basketball has become a popular urban team sport, and there has been a noticeable growth in research focused on this emerging basketball discipline. The aim of this study was to unpack the game outcome based on performance indicators and contextual variables. For this purpose and to achieve an objective evaluation, 13 performance indicators and the quality of opponents were fitted into logistic regression, decision tree, and neural network. Results showcased that the accuracy of the classification of neural networks markedly outperformed others under different game types and the stage of the tournament. Four key performance indicators that significantly impacted all game outcomes under two contextual variables were the percentage of 2-points and 1-points, defensive rebounds and turnovers, and the positive influence of the quality of the opponent on the game outcome was detected in the four sub-datasets. Furthermore, the performance indicators of ball possession and key assists can support classifying winning and losing teams from the games in regular and playoff seasons, respectively, while team fouls and extra free throws can facilitate discrimination of the game outcome in balanced games. This study will serve as a foundational resource to enhance the decision-making processes for participants in 3 × 3 basketball.