Half past six on a wet Tuesday evening, the last group of an elite squad jogs off the training pitch, and the first thing the fitness coach checks is not how tired anyone looks. It's a spreadsheet. Each player peels off a compression vest with a black pod tucked between the shoulder blades — a device roughly the size of a phone battery, holding a GPS chip, an accelerometer and a heart rate receiver — and hands it to a member of the sports science staff, who dockets it, syncs it, and has full session data on a laptop within ninety seconds. Somewhere in that data sits a single number that will quietly decide whether that player trains at full intensity on Thursday, works alone in the gym, or doesn't train at all. Twenty years ago that decision was made on gut feel and a stopwatch; now it's made by a graph.
What the Vest Actually Measures
The pod itself is unremarkable. STATSports' Apex unit, worn by squads at Liverpool, Arsenal and the Irish rugby team, weighs under 50 grams and sits flush against the upper back, sampling GPS position ten times a second. Catapult Sports, the Melbourne-based supplier used across the AFL, much of the NBA's practice-court analytics and a long list of Premier League academies, runs a comparable setup through its Vector system, and between the two of them they have effectively owned elite sport's movement-tracking market since the International Football Association Board formally approved electronic performance and tracking systems for use in competitive matches back in 2015. What the pod records is deceptively simple: total distance covered, distance covered above 5.5 metres per second (the standard threshold sports scientists use for 'high-speed running'), the number of accelerations and decelerations above 3 metres per second squared, and a composite 'player load' score drawn from the accelerometer rather than the GPS chip itself. None of those figures means much on its own. A midfielder who covers 11 kilometres in a match but does almost none of it at sprint speed carries a completely different injury profile to one who covers 9 kilometres but produces forty hard accelerations, and treating the two identically in a Tuesday session is a mistake most sports science departments stopped making a decade ago.
Football didn't arrive at this by accident. Bolton Wanderers, under Sam Allardyce in the mid-2000s, were among the first Premier League clubs to hire a dedicated performance director and build training decisions around data rather than habit, at a time when most top-flight squads still ran preseason off a printed schedule and a manager's memory of what worked the year before. That early, unglamorous investment is why a mid-table club with no European football ended up years ahead of the game on load management, and why so much of the vocabulary now standard across elite sport — 'high-speed metres', 'load monitoring', 'player readiness' — traces back to a training ground in Greater Manchester rather than a Silicon Valley lab.
Why a Full-Back's Numbers Mean Nothing for a Centre-Back
Position-specific benchmarking is where a lot of clubs still get it wrong, usually by accident rather than laziness. Wide players and full-backs routinely rack up the highest high-speed running totals on the pitch, covering repeated overlapping runs that a target man or a holding midfielder will never come close to matching, so judging a centre-back against a winger's weekly high-speed total tells a coach nothing except that they play different positions. The better departments build a rolling personal baseline for every player instead — comparing this Tuesday's numbers against that specific player's own last six weeks, not against a squad average — and flag deviations from that individual pattern rather than deviations from a generic target. Skip that step and a club risks two failure modes at once: resting a wide player who is actually well within his own normal range, and clearing a central defender whose load has crept up 30 per cent against his personal baseline even though it still looks unremarkable next to a winger's raw numbers.
The Ratio That Rewired Squad Rotation
One number now decides more matchday squads than any tactics board.
It's called the acute:chronic workload ratio, and it compares a player's training load over the past seven days against their rolling average over the past twenty-eight. The research behind it, published by sports scientist Tim Gabbett in the British Journal of Sports Medicine, identified a 'sweet spot' between 0.8 and 1.3 — inside that band, injury risk stays low; push a squad member above 1.5, spiking their load well beyond what their body has adapted to, and soft-tissue injury rates climb sharply. Any club still building preseason around a single weekly mileage target, ignoring how sharply that load rose from the week before, should scrap that approach outright; it's the sports science equivalent of judging a diet by calories alone while ignoring what happens when you go from 1,500 a day to 3,000 overnight. The method hasn't gone unchallenged, though. A wave of methodological critiques published in the very journals that popularised the ratio argued it suffers from mathematical coupling — comparing a number against a rolling average that already contains itself skews the result — and several sports scientists now treat a raw ACWR figure as a rough early-warning flag rather than a diagnosis on its own. What hasn't changed is the underlying logic: bodies adapt to load that rises gradually and break down under load that spikes suddenly, and no amount of academic argument over the exact formula has dented that basic principle.
Where the Numbers Get It Wrong
Cricket's fast bowlers illustrate the limits better than almost anyone. The England and Wales Cricket Board has monitored bowling workload for years using overs-per-week caps layered on top of GPS and video data, because a quick bowler's injury risk tracks bowling volume far more tightly than it tracks general running load — a metric built for football and rugby doesn't translate cleanly to a sport where the stress sits almost entirely in one repeated action. If you've ever wondered why a fast bowler who looks fresh in the nets still gets rested for a low-stakes match, that's usually the reason: the workload model flagged a spike three overs before anyone watching could see it in his run-up. Rugby union has its own version of the same problem. Forwards absorb collision loads that a GPS accelerometer registers as generic 'player load' without distinguishing a tackle from a hard change of direction, so several Premiership clubs now layer instrumented mouthguards and separate collision-tracking systems on top of the standard vest to fill that gap. Tennis and boxing barely use GPS at all for the same reason — a five-metre court or a boxing ring generates almost no distance data worth tracking, which is precisely why those sports lean on accelerometer-based swing and punch counters instead of anything satellite-driven.
None of this works without someone who actually watches the players. The best-run performance departments treat a player's own subjective wellness score — logged each morning on a phone before a single GPS number is pulled — as being at least as important as anything the vest produces, because sleep quality, muscle soreness and simple mood swings show up there days before they show up in movement data. A department leaning on hardware alone, and skipping that five-minute morning questionnaire because it feels unscientific next to a satellite feed, is behind where the sport actually is. Heart rate variability trackers worn off the pitch — Whoop and Oura are the two brands most commonly seen on elite athletes' wrists and fingers away from training — add a further layer, flagging when a player's nervous system hasn't recovered from a hard session even though every GPS metric from that session looked unremarkable. Put those three data streams next to each other and the picture sharpens considerably; leave any one of them out and a club is, in effect, working from a photograph with a third of the frame cropped off.
What the Best Programmes Do Differently Now
A sports scientist for every eight to ten first-team players is now close to standard practice among the Premier League's top six, up from ratios closer to one per twenty a decade ago, and that staffing shift is the real story behind the technology rather than the hardware itself. Return-to-play protocols after a hamstring or calf injury increasingly hinge on the same GPS baseline rather than a fixed calendar: a player is cleared to train fully only once he can match his own pre-injury high-speed running numbers in a controlled session, not simply because six weeks have passed since the scan. That shift alone has done more to cut reinjury rates than any single piece of hardware, because it replaces a guess dressed up as a deadline with an actual physical benchmark.
The vest generates the numbers, but it's the person cross-referencing those numbers against a coach's eye test, an ice-bath log and a player's own account of how his hamstring feels on a Tuesday who actually decides whether he starts on Saturday. That's the part no algorithm has replaced, and — for all the money clubs have poured into GPS suppliers over the past decade — probably never will.