Two matchdays, one strange pattern
Keep the ball, lose the game. That is roughly what the opening two rounds of the 2026-27 Premier League season have produced, and Opta put figures on it this morning.
Teams finishing a match with more than 50% possession have won 30% of the time. Teams finishing with less than 50% have won 50% of the time. Two rounds is a small sample, obviously. It is also a straight inversion of how the league has read for most of the past decade.
The weekend supplied the evidence
Chelsea produced the extreme case. They beat Brighton 4-3 at Stamford Bridge with 25.9% of the ball and completed 146 passes, the fewest they have managed in a Premier League match in Opta’s records, which stretch back to 2003-04. They led 3-0 inside 32 minutes. Their two games have brought 12 goals, more than any other team in the division.
Try it for $0.99.
Newcastle won at Tottenham with 42.5%. Hull won at Coventry with 31.5%. Sunderland beat Fulham with 41.4%. Four wins, four losing possession counts, and none of them looked like accidents.
The mechanism is not mysterious. Pressing has improved to the point where building patiently against a compact block is a slow route to almost nothing. Dropping off, absorbing pressure and breaking at speed forces defenders to run backwards. Chelsea did it with a front three in which Cole Palmer, Morgan Rogers and Joao Pedro have each scored or assisted in both matches.
The exception proves the rule
Arsenal and Manchester City ignored all of it. Both won twice, both with over 60% possession. When a side is good enough to control a game by keeping the ball, keeping the ball stays the safest way to close it out.
So this is not a rejection of possession football. It reads more as a reminder that possession is a tool rather than a virtue, and that most of the division is not equipped to use it against the best pressing sides in Europe. Anyone reading possession projections into a match preview should treat the figure as descriptive rather than predictive at this stage of the season.
Two matchdays can easily be noise. If teams are still winning half their games with a minority of the ball in November, it stops being noise and starts being a league-wide shift.
Discover the rest of our Football news on MathODDS!
in your pocket.
Real-time notifications & alerts
Leave a Reply