The most quantified institution in professional sport is not a Silicon Valley franchise but the one played in stadiums on Saturday afternoons. Association football has completed its analytics transition: tracking cameras record every player's position twenty-five times a second, models grade every action, and the “expected goals” that once sounded like satire are now the lingua franca of commentary, recruitment and tactics.
What the models changed
Recruitment first, and most profoundly. Scouts once judged what their eyes could see in person; clubs now screen entire continents' worth of match data before anyone boards a plane. The effect has been a genuine market correction: players whose contributions were invisible to spectators — pressing forwards, duel-winning defenders, progressive passers — repriced upward, and stylish volume shooters repriced down. The bargain bin of undervalued skills is smaller than it has ever been.
- Recruitment: data screening before scouting trips
- Tactics: pressing schemes and rest defence designed from tracking data
- Training: load management reducing soft-tissue injuries
- Contracts: performance models appearing in negotiation rooms
What the models did not change
The score. The sport's analytics community is refreshingly candid about the ceiling: football remains a low-scoring, high-variance game where championships are decided by moments the models call noise. Data found the edges — set pieces, finishing selection, fatigue management — worth real goals across a season. It did not find a formula. The tactical meta oscillates as it always has, now simply with better instrumentation on the pendulum.
The part fans actually feel
Two visible changes bear the data's fingerprints. Matches are more pressing-intensive and, purists complain, more positional. And the discourse has changed for good — a generation of supporters now argues with shot maps, which is either enlightenment or noise depending on the pub. The most honest summary of the revolution is the one the analysts themselves give: the data did not solve football. It finally wrote the sport down in a language precise enough to argue with.