Footballers' hamstrings fail at speed, not in the tackle

Muscle injuries are among the most frequent and costly problems in professional football, and the hamstring carries a reputation as the most vulnerable muscle of all. How often these injuries happen is well documented, but prevention depends on something different: knowing exactly what a player was doing in the moment the muscle failed. Video analysis can capture that inciting event directly, yet individual video studies have been small and inconsistent in how they classify actions. This review set out to pool them for the first time and ask whether a consistent pattern exists across the sport.

The researchers systematically searched five databases up to October 2025 for video-based studies of lower-limb muscle injuries in professional football, following PRISMA guidance. Twelve studies qualified, covering 513 players, of whom 434 were male and 79 female, and 538 injuries in total. They extracted the type of inciting event, the speed of the action, the timing within the match, the location on the pitch and the tactical situation. Dichotomous outcomes were pooled as odds ratios, multi-category outcomes through a Bayesian model, and study quality was appraised with three dedicated tools.

Hamstring injuries dominated at 74.7% of the total, followed by adductor injuries at 14.3%, quadriceps at 7.1% and calf at 3.9%. Non-contact injuries, 322 in all, were about eight times more likely than the 108 indirect contact injuries. Injuries occurred overwhelmingly during moderate-to-high horizontal speed actions rather than slow or stationary ones. Running, stretching or reaching, and kicking were the most common actions at the moment of injury, with no credible difference between them. Notably, nothing consistent emerged for match time, pitch location or tactical phase, and only male data were sufficient for meta-analysis.

The findings confirm the long-standing suspicion that fast running, not player contact, is the main setting for hamstring injury in football, and they add pooled numbers to what were previously scattered single-club reports. They also complicate some popular ideas: if fatigue late in matches or specific zones of the pitch drove injury, a timing or location signal should have appeared, and none did consistently. The big gaps are exposure adjustment, since the studies could not account for how much time players spent in each action, and women, who contributed 86 injuries but too little data to analyse separately.

For anyone who plays football or another sprint-based sport, this is a reason to think of high-speed running itself as the moment of greatest hamstring risk, rather than tackles or collisions. It also reframes where prevention attention plausibly belongs: on how the body handles sprinting, stretching for a ball and kicking, since those were the actions most often on screen when a muscle failed. The absence of a timing effect suggests the risk is not confined to the tired final minutes. What the data cannot yet say is how risk changes per minute of sprinting, or whether the same pattern holds for women.