Abstract
BACKGROUND: Sided games (i.e., small sided, medium sided, large sided) involve tactical, technical, physical, and psychological elements and are commonly implemented in soccer training. Although soccer sided-games research is plentiful, a meta-analytical synthesis of external load exposure during sided games is lacking. OBJECTIVE: The objective of this systematic review and meta-analysis was to: (1) synthesize the evidence on high-speed and sprint running exposure induced by sided games in adult soccer players, (2) establish pooled estimates and intra-individual reliability for high-speed and sprint running exposure, and (3) explore the moderating effects of game format and playing constraints. METHODS: A literature search was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidelines. Four databases (PubMed/MEDLINE, Scopus, SPORTDiscus, Web of Science Core Collection) were systematically searched up to 25 January, 2022. Eligibility criteria were adult soccer players (population); training programs incorporating sided games (intervention); game manipulations including number of players, pitch dimension, and game orientation (comparator); and high-speed, very high-speed, and sprint relative (m[Formula: see text]min(-1)) running distances and associated intra-individual reliability (outcome). Eligible study risk of bias was evaluated using RoBANS. Pooled estimates for high-speed and sprint running exposure, and their intra-individual reliability, along with the moderating effect of tracking device running velocity thresholds, pitch dimension (i.e., area per player), and game orientation (i.e. score or possession), were determined via a multi-level mixed-effects meta-analysis. Estimate uncertainty is presented as 95% compatibility intervals (CIs) with the likely range of relative distances in similar future studies determined via 95% prediction intervals. RESULTS: A total of 104 and 7 studies met our eligibility criteria for the main and reliability analyses, respectively. The range of relative distances covered across small-sided games, medium-sided games, and large-sided games was 14.8 m[Formula: see text]min(-1) (95% CI 12.3-17.4) to 17.2 m[Formula: see text]min(-1) (95% CI 13.5-20.8) for high-speed running, 2.7 m[Formula: see text]min(-1) (95% CI 1.8-3.5) to 3.6 m[Formula: see text]min(-1) (95% CI 2.3-4.8) for very high-speed running, and 0.2 m[Formula: see text]min(-1) (95% CI 0.1-0.4) to 0.7 m[Formula: see text]min(-1) (95% CI 0.5-0.9) for sprinting. Across different game formats, 95% prediction intervals showed future exposure for high-speed, very high-speed running, and sprinting to be 0-46.5 m[Formula: see text]min(-1), 0-14.2 m[Formula: see text]min(-1), and 0-2.6 m[Formula: see text]min(-1), respectively. High-speed, very high-speed running, and sprinting showed poor reliability with a pooled coefficient of variation of 22.8% with distances being moderated by device speed thresholds, pitch dimension, and game orientation. CONCLUSIONS: This review is the first to provide a detailed synthesis of exposure and intra-individual reliability of high-speed and sprint running during soccer sided games. Our estimates, along with the moderating influence of common programming variables such as velocity thresholds, area per player, and game orientation should be considered for informed planning of small-sided games, medium-sided games, and large-sided games soccer training. CLINICAL TRIAL REGISTRATION: Open Science Framework available through https://osf.io/a4xr2/ .