Large-sided games do not replicate the positional worst-case scenarios of official matches in professional football

The aim of this study was to determine if positional Worst-Case-Scenarios (WCS) during official games (OG) are replicated in Large-Sided Games (LSG) with a similar Area per Player (ApP) as OG. Twenty-one football players from the second team of a professional German football club were monitored during OG and LSG. WCS of 1 min (WCS1), 3 min (WCS3), and 5 min (WCS5) were calculated for OG and LSG using rolling average methods for Total Distance (TD), Medium-Speed-Running (MSR), High-Speed-Running (HSR), and Sprint Distance (SPD), in addition to low-, medium-, and high-intensity Accelerations (ACC1, ACC2, ACC3) and Decelerations (DEC1, DEC2, DEC3). LSG displayed significantly lower values ( p < 0.05) for TD, MSR, HSR, and SPD for all positions compared to OG. ACC1-3, and DEC1-3 were also significantly ( p < 0.05) lower in LSG compared to WCS in OG, with exceptions observed for Full Backs (FB) in ACC1 (WCS3), ACC3 (WCS1, WCS3, and WCS5), and DEC1 (WCS3); for Strikers (ST) in ACC1 and DEC1 (WCS3 and WCS5); and for Central Midfielders (CM) in ACC2 (WCS3). These findings suggest that LSG are not suitable for replicating OG' WCS1, WCS3, and WCS5. Coaches should therefore supplement with running-based drills or adjust training games to ensure that WCS demands are met.