Answers are grounded in this paper and cite their sources.
Introduction Badminton training studies evaluate several physical-conditioning approaches, but their comparative effects remain uncertain. Methods We searched PubMed, Web of Science, Scopus, SPORTDiscus, China National Knowledge Infrastructure, and Wanfang Data through 30 May 2026 and synthesised controlled interventions for change-of-direction ability, dynamic balance, lower-limb power, linear speed, and badminton-specific technique. The review was not prospectively registered. Of 190 included studies (202 reports), 80 were quantitatively eligible (58 randomised and 22 non-randomised), and 53 randomised controlled trials contributed 86 study-outcome effects to the primary Bayesian random-effects networks; non-randomised studies, single-trial leaf nodes, and effects with |Hedges g|>3 were examined in sensitivity analyses. Results For change-of-direction ability, posterior means versus Control were 1.33 (95% credible interval 0.83 to 1.83) for combined training, 1.33 (0.19 to 2.47) for core training, 1.16 (0.68 to 1.63) for plyometric training, and 0.28 (-0.21 to 0.77) for resistance training. Prediction intervals crossed zero for all four nodes, heterogeneity was substantial (I-squared = 71.2%), and incoherence was present. Combined packages that added balance or strength work to plyometrics showed larger effects (2.20, 1.38 to 3.03) than multicomponent functional-strength packages (1.10, 0.61 to 1.59), and no study-level covariate explained the heterogeneity. Secondary networks were sparse, and sensitivity analyses showed that single extreme trials could dominate estimates. Discussion All principal change-of-direction comparisons were of very low certainty. Structured training may improve change-of-direction ability on average, but the evidence does not support a reliable hierarchy of modalities.
No takes yet. Share an insight, caveat, or question.
Liu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: