The 13 minutes between the 68th and 81st were the most consequential 13 minutes of City’s season — and they hinged on a single back pass from Marc Guehi.
At the 68th-minute mark City had 75% possession and a 1-0 lead from Jérémy Doku’s curled finish into the top corner four minutes before the interval. The first half had been so City-controlled that the only open question seemed to be whether the 43’ goal got a partner before the table maths got serious. Three Everton yellows had already piled up by 53’ (Keane 45’, Beto 48’, Tarkowski 53’) — a side already hanging on by reputation, not pressure.
Then Guehi played the back pass.
Substitute Thierno Barry intercepted it and slipped it past Donnarumma. Four minutes after coming on for Beto, on a transition the home side had not won all night. 1-1. The first crack.
How the 13-minute crater opened
The shift between the 68th and 81st minutes was as fast as any City have endured under Guardiola. Three goals from a side that had spent the prior hour without one:
- 68’ — Barry, off the Guehi back-pass error.
- 73’ — Jake O’Brien, glancing home a James Garner corner at the near post.
- 81’ — Barry’s second, after a shot deflected into his path.
By the 81st minute the visiting side were 3-1 down with nine minutes left. The home crowd had had a comeback they hadn’t seen in a decade. The two stats most useful to read together: City’s 75% possession had produced only 4 shots on target out of 20 attempts and Everton’s expected-goals tally — 2.77 by full-time, against City’s 1.37 — said the home side were the better attacking team on quality, even though they touched the ball a quarter as often.
How City got it back
Two goals in stoppage time. Haaland at 83’, clipping his 25th of the season over Pickford after storming forward. Doku at 90+7’, curling a finish with almost the last kick of the match.
A point either side could argue they didn’t deserve. Pep called it “an outstanding first half” and a result that “is better than a loss”. David Moyes, going the other way, said only “when you’re 3-1 up, you’re thinking you’re in with a great chance of winning”.
Both are right. Both also know the table doesn’t care.
What the underlying numbers say
City out-shot Everton 20 to 14 and out-possessed them 75-25, but the expected-goals tally tells a different story: 2.77 for Everton, 1.37 for City. Everton attempted fewer shots from better positions; City attempted more from worse ones. The home side’s 6 on target out of 14 attempts vs City’s 4 on target out of 20 is the cleanest summary of who actually threatened.
Pickford made one save. One. City had the ball; they did not have the goalmouth.
There was also the Keane-on-Doku tackle reviewed for a potential red card in the first half, downgraded to a yellow on review. Moyes, after the match, defended a separate non-decision: “If that doesn’t get given as a penalty then it’s an absolute free-for-all.” That flashpoint is harder to score from the data alone, but the Sky reporting flags it as the moment the home dugout will be most aggrieved by.
What this does to the title race
Coming in: Arsenal 76 points (35 played), Manchester City 70 points (33 played). The arithmetic that mattered was simple — City’s two games in hand, both at 100% conversion, would have made it a one-game gap with parallel runs.
Going out: Arsenal 76 (35 played, three to go). City 71 (34 played, four to go). City’s maximum reachable total is now 83. Arsenal’s is 85.
In practice that means Arsenal need any combination worth eight points from their last three games to win the title regardless of City’s results. Two wins and a draw does it. Three draws does not, but only just. The cleanest path is: West Ham at the Emirates next weekend, then a tour of the run-in’s two easier away days, and the title is settled before the final fixture round even kicks off.
There is a scenario where City catches Arsenal — Arsenal lose two of their last three; City win all four — but the probability assigned to that combination by any reasonable forecast is in low single digits.
Verdict
Two points dropped in a game where the chasers had to win is the kind of result that looks, in retrospect, like the moment the league was decided. The xG model thinks City were the marginally better side overall; the scoreboard says otherwise; the table says it doesn’t matter.
Three goals in 13 minutes will live in the highlight reels. The title race, on this evidence, will not.