Supplementary Table S31. Estimated between-study heterogeneity (τ2) for each outcome-specific network and 95% prediction intervals for key contrasts Outcome No. of studies No. of interventions in network Assumed heterogeneity structure τ2 Key contrast MD (95% CI) 95% prediction interval 5-m sprint 10 4 Common τ2 across comparisons within this outcome-specific network 0.0006 Traditional Strength Training vs Regular Training -0.09 (-0.11, -0.07) -0.13 to -0.05 20-m sprint 13 6 Common τ2 across comparisons within this outcome-specific network 0.0018 Traditional Strength Training vs Regular Training -0.13 (-0.20, -0.06) -0.26 to 0.00 30-m sprint 13 6 Common τ2 across comparisons within this outcome-specific network 0.0032 Speed Training vs Regular Training -0.16 (-0.34, 0.01) -0.42 to 0.10 40-m sprint 8 4 Common τ2 across comparisons within this outcome-specific network 0.0041 Endurance Training vs Regular Training -0.21 (-0.45, 0.03) -0.56 to 0.14 SJ 15 5 Common τ2 across comparisons within this outcome-specific network 5.92 Traditional Strength Training vs Regular Training 4.40 (2.07, 6.74) -0.02 to 8.82 CMJ 28 6 Common τ2 across comparisons within this outcome-specific network 4.68 Traditional Strength Training vs Regular Training 3.58 (2.00, 5.15) 0.11 to 7.05 COD 17 5 Common τ2 across comparisons within this outcome-specific network 0.0054 Traditional Strength Training vs Regular Training 0.13 (0.08, 0.18) 0.01 to 0.25 1RM 16 4 Common τ2 across comparisons within this outcome-specific network 16.74 Traditional Strength Training vs Regular Training 20.86 (10.73, 30.98) 1.84 to 39.88 Network meta-analyses were conducted separately for each outcome. For each outcome-specific network, a common between-study heterogeneity variance (τ2) was assumed across all treatment comparisons and estimated under a random-effects model. The term “multivariate” refers to the contrast-based formulation implemented in the Stata network suite for treatment comparisons within a given outcome network, rather than to a joint model across different outcomes. Key contrasts were selected according to the most decision-relevant comparisons reported in the main text, generally contrasting the highest-ranked intervention against Regular Training. For sprint outcomes, negative mean differences indicate faster sprint times. For SJ, CMJ, and 1RM, positive mean differences indicate better performance. For COD, the direction is presented as reported in the current manuscript and should be interpreted accordingly.