Data / entry 09 / May 25, 2026 / 6 min read

Arsenal's title, broken down at the underlying-numbers level

Manchester City actually scored more goals than Arsenal this season. Here's the xG model, the shot-suppression data, the pressing numbers, and the corner-routine breakdown that explains how Arsenal still won by 7 points.

Start with the counterintuitive part

Manchester City scored six more league goals than Arsenal this season. If the title were decided on attack alone, City wins it. It wasn't decided on attack. Arsenal conceded eight fewer goals than City, and that single number, the defensive gap, is larger than the seven-point margin that separated the two clubs at the final table. This wasn't a title won by outscoring the division. It was won on the other end of the pitch, and the underlying data shows it wasn't luck either.

Is the defensive record repeatable, or was it variance?

Expected goals (xG) measures shot quality allowed, independent of finishing. If actual goals conceded tracks closely with xG against, the defensive record reflects a repeatable process rather than good fortune: missed sitters, an overperforming keeper on chance quality, and so on.

Arsenal conceded 27 goals against a model expectation of 28.5, a gap of 1.5 goals across 38 matches. That's about as close to a 1:1 match between expected and actual as a defense gets over a full season. The takeaway: this wasn't a defense riding a hot goalkeeper or a run of opponents missing open nets. The underlying shot quality they allowed was genuinely elite, and the actual results confirm the model rather than deviating from it.

Shot suppression: what opponents were actually allowed to do

That 70.7% figure isn't just a low-shot-volume defense getting outshot but bailed out by blocks. It's a defense that forced nearly 3 of every 4 attempts wide, over, or into a block before they ever became a clean look at goal. Six times this season, an opponent finished a full 90 minutes without a single shot on target against Arsenal, a mark matched by only one other side in Premier League history's on-record era.

The other half: goals manufactured from dead balls

Roughly a third of Arsenal's entire scoring output this season came from dead-ball situations, and the corner-kick number specifically broke a record that had stood since before most of this squad was born. In a season where City's attack outproduced them by six league goals in open play and general play combined, Arsenal closed most of that gap through a single, coachable, repeatable phase of the game: the corner routine.

Pressing intensity: who's actually forcing the tempo

Arsenal's own PPDA of 10.05 is the second most aggressive pressing rate in the league. The PPDA opponents managed against Arsenal was 18.05, nearly double, meaning teams routinely dropped into a low block and let Arsenal have the ball rather than risk pressing them. Arsenal spent the season imposing tempo far more than they had it imposed on them, a real asymmetry in game control, not just a record on paper.

Where the corner goals actually came from

The routine itself is a repeatable zone pattern, not chaos: Arsenal overload the near post with runners to drag markers away, while Gabriel, the Premier League's top-scoring defender since 2021/22 with 15 goals, deliberately holds back from the initial cluster and arrives late into the space that opens up at the back post once defenders have followed the near-post decoys.

Putting the model together

Three separate data points point the same direction. The xG model says the defensive record was earned, not fortunate, with actual goals conceded landing within 1.5 of the model's own expectation across a full season. The shot-suppression numbers say opponents were denied clean sight of goal more than seven times in ten. And the set-piece data says the attack that finished six goals behind City's in open play closed most of that gap through one specific, repeatable mechanism. None of that is a story about who wanted it more. It's a defense that the underlying numbers actually back up, paired with a manufactured, quantifiable scoring edge from set pieces. That combination, not raw attacking firepower, is what a seven-point title actually looked like under the hood this season.