Manchester Derby: Pro Ref Admits Error on Haaland's Goal — and the Price of an Offside Law That Cannot Keep Up With Its Own Technology

Minute 60 at Old Trafford. Joško Gvardiol whipped a cross in from the left, the ball flicked off a defender's boot and changed direction, and Erling Haaland tapped it in from a distance that a top striker never misses. The assistant referee raised his flag. The stadium erupted. Then the flag came down, the earpiece crackled, and the big screen flashed the familiar line: checking offside.

The semi-automatic offside system drew its lines, measured, and concluded within forty seconds: Haaland level with Patrick Dorgu's leg. No offside. The goal stood.

I was sitting thirty miles from Old Trafford, in Liverpool, a stopwatch next to my coffee. From the moment the referee touched his ear to the moment he pointed at the centre circle: one minute and forty-two seconds. Nearly two minutes for one decision. In those two minutes, a stadium of 76,000 went cold, a defensive line lost its shape, and the rhythm of a derby was snapped in half.

Then Pro Ref spoke. English football's professional refereeing body admitted Enzo Fernández had been in an offside position on the play, and that his position could have impacted the outcome. They added that VAR should have recommended an on-field review. The three points stayed in Manchester. The crack, however, had already formed — and it was not located in Haaland's foot.

Three layers of decision in a single phase

People call me reckless, but the data has never lied to me. What the data shows here is one phase handled by three separate layers of decision-making, each answering a question the layer below never asked.

The first layer is technology. The semi-automatic system did its job: it measured Haaland against Dorgu, and it measured accurately. This is where many people are getting it wrong. The machine did not fail. It answered precisely the question it was built to answer.

The second layer is the VAR team. They saw Enzo Fernández in the box, in an offside position, standing between Lisandro Martínez and Haaland. But they applied a threshold the Premier League itself has published: when an offside player does not touch the ball, the situation becomes subjective, and the bar for VAR intervention rises sharply.

The third layer is Pro Ref, issuing a verdict after the final whistle. They acknowledged an error of judgement and committed to a review.

Three layers, three answers, one result. That is why I say the problem sits outside Haaland's boot.

The technology measured correctly. The law is asking the wrong question.

Law 11 of the IFAB Laws of the Game states that a player in an offside position is penalised if they become involved in active play. Involvement is defined in three ways: blocking an opponent's line of vision, challenging an opponent for the ball, or performing an act that clearly impacts an opponent's ability to play the ball.

Fernández did not touch the ball. Fernández did not collide with anyone. But Fernández moved toward the ball, and by Pro Ref's own reading, that movement may have affected Martínez's ability to attack the cross, or the goalkeeper's ability to collect it.

The semi-automatic system cannot measure intent. It cannot distinguish a player running to receive the ball from a player running to cut off a defender's recovery path. To cameras and AI, both appear identical: a point on a map, a movement vector, a distance in millimetres.

In other words, the technology answered the question of whether Haaland was offside, and the answer was no. But the real question was whether a different player in an offside position affected the play. That is where the system goes silent.

I spent most of the following morning re-reading similar incidents from recent seasons. A pattern emerged: whenever the law produces an objective criterion, technology is immediately cited to defend it. Whenever the law slips into subjective territory, technology is declared to lack the authority to intervene. That logic protects the process. It does not protect the result.

My own worst mistake, and why I keep repeating it

I was wrong about the 2026 World Cup. It remains the most expensive lesson I have.

On live radio during Croatia versus Denmark in the round of sixteen, I mispronounced Ivan Rakitić's name three times in a row. Not once. Three times. I realised my carelessness the moment the producer's voice came through the headset.

How I handled it is the part worth telling. I spent a full month rewatching footage, logging every pass from Croatia's midfield, and writing a piece arguing against myself: Modrić, Rakitić and Brozović completed 89% of their passes in the knockout rounds, and that is why Croatia reached the final.

I bring this up not to apologise. I bring it up because it explains how I read what happened at Old Trafford. When a process fails, fixing the process is the small job. Fixing the habit of trusting a process that has just admitted its own failure is the big one.

Number 91 was never a lucky number. It was the endpoint of a plan. In 2026 I published a blog post claiming Salah, Firmino and Mané would score at least 84 goals across all competitions. I was mocked. They finished 2026-18 with 91: Salah 44, Firmino 27, Mané 20. That taught me one thing sufficient for the rest of my career: a specific number always beats a vague judgement, provided the number is right.

At Old Trafford, the right number is being used in the wrong place.

What nobody wants to say out loud

Manchester United played with a man advantage from midway through the first half, after Phil Foden's red card. They had more than forty-five minutes with an extra body, at home. They lost 1-0.

That is the most important number of the match, and it has nothing to do with offside.

I do not have expected goals data for this fixture, and I will not invent it. But I have one historical fact sufficient to raise the question: when a major club plays a half with a man advantage at home and fails to score, the problem is rarely the referee. It is the structure of the attack, the quality of the final decision, the number of solutions available against a compact block.

Martínez called it a collective injustice. I understand the feeling. Players sense errors more sharply than anyone, and they have every right to speak. But a club that builds its identity around injustice runs into a very specific problem: when you believe the world is against you, you stop auditing your own work.

Home advantage is dead. The pandemic proved it, and eleven players in a second half at Old Trafford have just proved it again.

Ten men and a shield

Set the controversy aside for thirty seconds and look at what Manchester City actually did. They played half a match with ten men at one of the most hostile grounds in England. They kept a clean sheet. They protected the most fragile lead football offers: one goal.

I have watched English football for close to fifty years, and I can tell you that big clubs are usually poor in this situation. When they lose a man, instinct pulls them deep; when they sit deep, they lose the ball entirely. City did the opposite at Old Trafford: a low block that stayed connected, ceding territory without ceding the centre.

That is a genuine tactical achievement, not luck. Anyone using the offside row to erase it is reading the wrong match.

Data does not kill emotion. It gives emotion a frame. This match's frame has two bars: a contested goal, and forty-five minutes in which the team with the extra man created nothing.

Where I could be wrong

First possibility: I am wrong and Pro Ref is right. If Fernández's movement genuinely had no bearing on Martínez's actions, the on-field decision was correct and the post-match admission was merely public relations. In that case, everything since is noise with no technical value.

Second possibility, and the one that worries me more: I am right, and being right makes this dangerous. A refereeing body publicly conceding an error after the final whistle sets a precedent. Once that precedent exists, every club knows it can demand an apology after any adverse decision. When apologies become part of the process, the finality of football results begins to wobble.

I do not want to live in a game where results get re-opened on a Monday.

Third possibility: both sides are right within their own scope, and the real problem is that the offside law was written for a sport that no longer exists. A rule drafted when referees had two eyes and a flag is now applied by a system with dozens of cameras and skeletal-tracking algorithms. That mismatch cannot be fixed in a meeting.

And I will admit my own biggest weakness: I have no process data for this match. No expected goals, no touches in the box, no post-loss pressure index. Every tactical claim above rests on the result and on observation, not on measurement. A serious writer has to say that plainly.

The opposite view: the goal may have been entirely legitimate

I will push my own argument one uncomfortable step further.

Law 11 requires an act that clearly impacts an opponent's ability to play the ball. The word clearly is the most important word in the sentence, and it is not decorative. If a player merely stands there, moves with the momentum of the phase, and appears in a defender's eyeline without blocking a path, obstructing vision or contesting the ball, then under a strict reading there is no offence.

On that reading, the on-field decision may have been correct, and Pro Ref's admission may be a second error stacked on the first.

That is why this article is not a demand for justice for Manchester United. Justice in football has no unit of measurement. What concerns me is process: a result established by three layers of decision-making, then denied by one of those layers the following day, while the result itself stands untouched.

That asymmetry is the real price of this derby.

What I will be tracking

First, the formal conclusions of Pro Ref's review. If, within two to four weeks, they issue specific guidance on offside players who do not touch the ball but move toward it, that will be the first procedural change of its kind in the Premier League.

Second, the scope of semi-automatic offside technology. The current system cannot assess the impact of an offside player who makes no contact. If the league announces an expansion of the system's scope before the winter fixture pile-up, I will treat that as evidence that Old Trafford generated real pressure rather than a press release.

Third, the rate at which VAR recommends on-field reviews in non-contact interference cases. If that number rises next month, the published subjective threshold has effectively been lowered, whatever the written protocol says.

Fourth, and most important to me: United's next three fixtures. Seven points from nine and the injustice story becomes fuel. Fewer, and the story is a blanket — and every blanket eventually gets pulled off the person sleeping under it.

A belief that has expired

Home advantage is dead. Not because crowds stopped singing, but because visiting teams learned how to play inside the noise. Old Trafford used to be a place where away sides surrendered before kick-off. Now it is a place where a team with ten men comes to keep a clean sheet.

That is the only thing in this derby I am absolutely certain about. Everything else — the offside line, the red card, Pro Ref's admission — will fade within weeks. But a major club losing at home while a man up for an entire half does not fade.

The data will record three points for Manchester City and not a single line about the feeling of injustice. The league table has no column for grievance.

Football waits for no one. It only waits for those willing to ask the question. Tonight's question is not whether Haaland was offside — that answer arrived long ago and it was accurate. The question is whether we accept a system in which technology can measure millimetres but not an action, in which a governing body can publicly admit an error but cannot return the points, and in which a major club can play a man up for forty-five minutes and fail to score.

I wrote this at two in the morning, and I still believe every word. But as always, if I am wrong, I will be the first to say so. My reputation was not built on predictions that came true. It was built on the fact that I was willing to make them.