Training Load and Match-Play Performance in Collegiate Division I Basketball

Abstract

Purpose: This study examined the relationship between training load and match-play team performance in the subsequent game. Methods: Training load for a NCAA Division I Basketball team was tracked over a 3-year period; 225 practices were observed and recorded. Training load was classified as total duration of training and duration of full-court 5-versus-5 (5v5) during training. In-game performance was tracked for 92 matches during this period. Training load was organized into 48-hours prior to competition (MD-2), 24-hours prior to competition (MD-1), Total Duration, Average Duration, Totals 5v5, and Average 5v5. Performance was determined on how the team faired against the closing point spread differential (CPSD). If the team performed above the CPSD it was considered an ABOVE performance, whereas if the team performed below the CPSD it was considered BELOW. A linear mixed model was used to assess the differences between ABOVE and BELOW performances. Cohen’s d effect sizes (ES) were calculated to determine the magnitude of the differences. Results: Statistical significance was set at p ≤ 0.05. Both 5v5 MD-1 (p = 0.03; ES = -0.43) and Average 5v5 (p = 0.01; ES = -0.55) leading into matches was significantly greater for BELOW than ABOVE outcomes. Conclusion: 5v5 MD-1 and the Average 5v5 in preparation for competition had a significant impact on ensuing performances, whereas no significant difference was found in training MD-2. These findings could have implications on the sequencing of training sessions leading into competition.   Keywords: basketball training; team sports; duration; team performance; points spread