Can We Replicate the Most Demanding Periods of Official Football Matches in Large-Sided Training Games?

<b>Objective:</b> This study aimed to compare the external load demands of large-sided games, with and without regulatory interruptions, to the average (FOOTmatch) and most demanding scenarios (MDSmatch) observed in official football matches, taking into account players' pitch positions. <b>Methods:</b> Large-sided training games were implemented with regulatory interruptions (FOOTtask) and without interruptions (MURDtask), during which a coach continuously introduced new balls to maximise effective playing time. GPS data were collected from eight official matches and six training sessions, involving a total of 30 young male football players. <b>Results:</b> MURDtask surpassed FOOTmatch, FOOTtask, and even MDSmatch in distance covered (Effect Size = 0.6-2.5), and equaled MDSmatch in the number of decelerations > 3 m·s^-2. Both FOOTtask and MURDtask exceeded the average match demands in most locomotor external load variables, except for sprinting efforts > 28 km·h^-1. However, none of the training games reached the peak values of MDSmatch in high-speed running or high-intensity accelerations and decelerations (>4 m·s^-2). Positional analysis revealed significantly higher values in MDSmatch for most variables, except for decelerations <-3 m·s^-2. Conclusions: MURDtask is a useful strategy to overload players, but additional tasks are required to replicate high-speed and high-intensity accelerations and decelerations typical of MDSmatch.