Basketball is an intermittent sport with high neuromuscular and metabolic demands. To optimize specificity, training tasks should replicate competitive loads, but little is known about how drills compare to official matches. This study compared the physiological and biomechanical load of training tasks with official competition in elite U18 basketball players. Twelve male players (16.9 ± 0.8 years) were monitored across two seasons (179 training sessions, 21 matches). A total of 3136 individual records were collected using Catapult Vector S7 LPS units. Training drills were classified by specificity (0–5). Physiological (distance and intensity zones) and biomechanical variables (accelerations, decelerations, jumps, explosive efforts, PlayerLoad™) were analyzed using cluster analysis and linear mixed models. Competition imposed the highest physiological and biomechanical loads. Non-opposition drills (1v0–5v0) showed limited transfer, though 1v0–2v0 accumulated higher jump density. Among opposition formats, 3v3 full-court was the best at replicating match demands. Continuous opposition tasks (3v3v3, 4v4v4, 5v5v5) elicited lower physiological but comparable biomechanical load. Small-sided formats, particularly 3v3 and 4v4, are the most effective training tools for reproducing competition demands, while non-opposition drills are better suited for technical or rehabilitation purposes.
Marín et al. (2026) studied this question.