Speed, on its own, wins nothing. What it does is compress the decision window. A forward running at 36 kilometres per hour through a defensive line forces the centre-back to commit a half-second earlier than the geometry of the pass demands; that half-second is the difference between a goal and a covering tackle. The Premier League’s official tracking data for 2025/26 gives us the ten highest peak speeds recorded this season, and the question worth asking is not simply who is fast, but what their speed buys their team in the channels where goals are scored.
Gabriel Martinelli tops the list at 36.74 kilometres per hour, a figure logged against Wolverhampton Wanderers on 25 April 2026. The context matters. Martinelli’s sprint came in Arsenal’s right half-space, the channel where Mikel Arteta’s 3-4-2-1 overloads opposition full-backs. Djed Spence, at 36.62 kilometres per hour, reached his peak against Wolves a year earlier; Anthony Elanga hit 36.35 against Chelsea in December; Anthony Gordon touched 36.26 at Brighton in October. Newcastle United, remarkably, placed three players in the top ten, a concentration that reflects Eddie Howe’s insistence on wide players who can transition in under two seconds from a mid-block to a full sprint behind the opposition back line.
The raw numbers, though, need a frame. Speed without end product is a training-ground metric. What the 2025/26 data shows, when cross-referenced with goal contributions, is a split between players whose pace functions as a direct attacking weapon and those whose pace serves a structural role.
Erling Haaland’s 36.09 kilometres per hour, recorded against Everton on 4 May 2026, is perhaps the most instructive entry. Haaland’s peak speed is not his primary threat; his movement in the box, his aerial dominance, and his finishing with both feet account for the bulk of his 0.94 expected goals per ninety minutes this season. But the sprint against Everton was different. It came from a Kevin De Bruyne through-ball that split the centre-backs, and Haaland’s ability to cover the final twelve metres before the covering defender could recover is what turned a 0.12 xG chance into a goal. The StatsBomb model assigned that sequence an xG of 0.12 at the moment of the pass; Haaland’s speed doubled the probability of conversion by eliminating the defender’s angle.
Martinelli’s case is the inverse. His 36.74-kilometre-per-hour sprint did not end in a goal; it was a recovery run from a failed Arsenal corner, covering thirty-two metres in 3.8 seconds to prevent a Wolves counter-attack. The value was defensive. Arsenal’s PPDA (passes allowed per defensive action) in the final third this season is 8.3, the second-lowest in the league, and that press relies on wingers who can close down in under four seconds. Martinelli’s speed is not a luxury; it is the mechanism by which Arteta’s counter-press functions on the right side.
Elanga and Gordon present a paired case study. Both have registered peak speeds above 36 kilometres per hour, both operate as wide forwards in Howe’s 4-3-3, and both have contributed directly to Newcastle’s rise to fourth in the table. Elanga has seven goals and four assists in 2,180 minutes; Gordon has nine goals and six assists in 2,410 minutes. Their combined 26 goal contributions from wide positions are the highest of any pair of wingers in the league, and the common factor is the speed with which they attack the space behind the full-back. Newcastle’s average width in possession this season is 61 metres, the third-highest in the Premier League, and the wide channels are where Elanga and Gordon do their damage.
The structural argument extends to the clubs themselves. Manchester City’s possession model, which averages 64.2 per cent of the ball this season, does not depend on raw speed in the same way. Haaland’s pace is a finishing tool, not a build-up mechanism. Arsenal’s counter-pressing system, by contrast, treats speed as a defensive asset first and an attacking one second. Newcastle sit between the two: they press at a PPDA of 9.7, higher than Arsenal but lower than City, and their wide players’ speed serves both phases.
What the data does not show is fatigue cost. The Premier League’s tracking system records peak speed, not the number of sprints above 34 kilometres per hour per ninety minutes. A player who hits 36.74 once in a season is in a different physical category from one who sustains thirty sprints above 34 across ninety minutes. That metric, the sprint volume, is what separates a pace-dependent system from one that merely benefits from it.
The forward-looking constraint is this: as the league’s tracking infrastructure improves, the conversation will shift from peak speed to sprint density, the number of high-intensity runs per match. Managers like Arteta and Howe, whose systems are built on the assumption that their wide players can recover in under four seconds, will need to manage that load across a thirty-eight-game season. The fastest player in the Premier League is not necessarily the most productive. The most productive is the one whose speed fits the geometry of the system around him, and whose body can sustain it from August to May.
