The red card did not create United’s attacking problem. It removed the excuse for hiding it.
The game-state audit
The officiating error affected the result. The rest of the evidence still asks why United could not make the extra player decide the match.
The officials were wrong. That is where the analysis begins.
Erling Haaland’s 60th-minute goal should not have survived without an on-field review. Haaland was onside, but Enzo Fernández attacked the same cross from an offside position. Lisandro Martínez reacted to that movement, the ball continued to Haaland, and the only goal of the derby followed.
Pro Ref later contacted Manchester United and acknowledged an error of judgement. Its explanation was specific: the VAR did not recognise the likely impact of Fernández’s position and should have recommended that Michael Oliver review the incident at the monitor.
United supporters are entitled to be furious. The goal decided a Manchester derby and three league points. Pretending the mistake was irrelevant would not be balanced analysis. It would be a different kind of dishonesty.
But an official apology cannot be allowed to perform tactical witness protection. Remove Haaland’s goal and United still spent more than an hour at home against ten men without scoring. The decision explains why City won 1–0. It does not explain why United gave Donnarumma so little serious work.
The referee error belongs in the case file. It cannot close the case.
Having the extra man is not the same as using him
Phil Foden was sent off in the 23rd minute. From that moment, the original derby plan became obsolete. United no longer needed only to survive City’s pressure or attack the space behind an aggressive side. They needed to move a reduced block, create overloads and keep returning to the weak side before City could reset.
That requires speed of thought more than speed of running. Move the ball early. Make the spare player visible. Stretch the back four, then attack the gap created by the stretch. Put runners around the penalty area so that a cross, cut-back or blocked shot has a second action waiting behind it.
United produced moments. Marcus Rashford struck the post from a free-kick before half-time. Kobbie Mainoo hit the woodwork early in the second half. Bryan Mbeumo forced Donnarumma into an excellent late save. The match was not empty.
It was scattered. United found openings without making the openings reproduce themselves. The ball travelled around City more often than it travelled through them. We gave the ball a stadium tour and kept forgetting to show it the goal.
Opta recorded only two United shots on target after the dismissal. That means City’s disadvantage existed on the team sheet far more clearly than it existed in their penalty area.
City changed the match they were playing
The most uncomfortable comparison is not between United’s eleven players and City’s ten. It is between the two benches.
City dropped into a 4-4-1 after the red card. That gave Rúben Dias and Marc Guéhi protection in front of the box and denied United the transition spaces they had expected before kickoff. At half-time, Enzo Maresca introduced Iliman Ndiaye for Rayan Cherki. The substitution kept a ball-carrier on the pitch who could win fouls, carry City upfield and give Haaland somewhere to run beyond.
Opta’s numbers make that adjustment visible. Ndiaye attempted six dribbles and won five fouls. City did not merely retreat and wait for the clock. They built an escape route.
Maresca’s own explanation was revealing. City wanted to keep the ball when possible and attack when they recovered it. They understood that blindly running forward would make the second half feel endless. So they protected their shape, chose their exits and preserved enough ambition to threaten United.
United needed to make ten-man City solve a fresh problem every few minutes. Instead, City solved their emergency once and United spent the evening submitting the same question in slightly different wording.
Carrick’s second plan never became clear
Michael Carrick began with Mainoo and Youri Tielemans behind Bruno Fernandes, with Mbeumo, Rashford and Matheus Cunha forming the attacking line. That selection made sense for a level derby with spaces to attack. The red card demanded a different use of those players.
The spare man should have appeared somewhere City could not comfortably ignore: an extra midfielder stepping beyond the block, a full-back arriving on the outside, or another penalty-box target forcing the centre-backs to defend facing their own goal. Instead, too many attacks arrived with City organised and United still looking for the perfect pass.
Benjamin Šeško replaced Tielemans in the 67th minute. That at least added a recognised penalty-box reference. Joshua Zirkzee replaced Rashford in the 83rd. By then United were chasing the match, but the pressure never became the sustained siege the situation invited.
This is not an argument that one magical substitution was sitting on the bench. Personnel can limit a manager. Execution can ruin a good instruction. Posts can reject good shots. Goalkeepers are allowed to make great saves.
The criticism is simpler. When the contest changed in the 23rd minute, United’s attacking relationships did not change quickly or decisively enough with it. Carrick admitted afterward that United could have created more. That is fair. The next question is whether the staff can identify the mechanism, because “more” is a quantity. United first need a clearer method.
The transfer window was supposed to create answers
This derby also returns us to the risk register United carried out of the transfer window.
The club rebuilt the midfield around different functions. Mainoo could receive and connect. Tielemans could direct play. Andrey Santos offered another progression profile. Mbeumo and Cunha gave the attack more variety. Šeško supplied the penalty-box reference the side had often lacked.
On paper, that is the beginning of adaptability. On Sunday, the pieces did not yet become a convincing second plan.
That distinction matters. Recruitment does not finish when a player holds up the shirt. The football department still has to turn several useful profiles into relationships that survive an unexpected game state. Otherwise, depth becomes a list of names waiting beside the manager while the same attacking problem stays on the pitch.
The derby does not prove the summer failed. It gives the summer’s promise a practical test: can United change the function of the team without needing a different starting eleven and a full week of preparation?
January gave United a result. September asked for repeatability.
Carrick’s first derby in January produced a 2–0 United win at Old Trafford. United were compact, disciplined and ruthless enough to punish City in the second half. It was an important result and legitimate evidence that the team could execute a demanding plan.
September asked a different question. Could United lead the game rather than disrupt it? Could they manufacture space against a side that had surrendered territory? Could they turn an unexpected advantage into authority?
The answer was no. That does not erase January. It defines its limit.
A rebuild cannot live forever on proof that one plan worked one night. The next stage is repeatability across different problems: a low block, a transition game, a physical away match, a European turnaround, or a derby transformed by a red card.
January showed Carrick could prepare United for a contest. September showed that changing the contest from the touchline remains an open part of his audit.
The fair defence of United
There is a reasonable alternative reading. United hit the post twice. Donnarumma made the decisive late save. City’s goal was awarded after the refereeing body now accepts the VAR process failed. On another afternoon, one of those efforts goes in, Haaland’s goal stays disallowed, and the same performance produces a United win.
Results contain luck. Analysis should be able to admit it.
The problem is that a more favourable scoreline would not transform two shots on target after the red card into convincing control. It would make the result easier to live with while leaving the structural question unanswered.
This was also only United’s fourth league match of the season. It is too early to turn one derby into a final judgment on Carrick, the new midfield or the entire recovery. The proportionate verdict is a warning, not a sentence.
United were unlucky in moments. They were also too easy for ten men to manage. Both truths fit inside the same match.
The verdict
The worst part of the derby was not simply losing to Manchester City. It was watching the match hand United a new problem and seeing City answer it first.
The red card offered control. United treated it like possession. City lost a player but kept their distances, their outlet and their nerve. United kept the ball without making the numerical advantage feel inevitable.
That is the next recovery test. Not another statement about mentality. A visible response when the game changes: faster circulation, clearer width, more bodies occupying the box and substitutions that alter the problem rather than decorate it.
Pro Ref can explain the goal. Only Manchester United can explain the football around it.
United had the extra man. City had the plan. Until that sentence stops being true, this rebuild remains qualified—not fixed.