Erling Haaland was onside. The technology got that part right: Manchester City’s striker had legally arrived at the back post, finished the chance and appeared to have been rescued from an incorrect assistant referee’s flag by VAR. Yet the goal should still never have counted, and the reason has almost nothing to do with Haaland himself.
Ten days after his strike decided the Manchester derby at Old Trafford, the Premier League’s independent Key Match Incidents Panel has delivered the clearest verdict yet on one of the most controversial VAR decisions of the young season. All five members agreed that the intervention which turned a disallowed goal into Manchester City’s winner was wrong. The player who committed the offside offence was not Haaland but Enzo Fernandez, who never touched the ball at all. FootyRoom previously explained why the goal was initially allowed, but the Premier League panel’s verdict now provides a definitive assessment of what the officials missed.
It sounds contradictory, but the incident is actually a useful demonstration of a part of the offside law that is still widely misunderstood. A player does not have to touch the ball to become actively involved in play, and the technology used to determine an offside position cannot decide everything that happens afterwards.
VAR solved the wrong problem
The controversy began in the 60th minute at Old Trafford, with City already down to ten men following Phil Foden’s first-half red card. Josko Gvardiol sent the ball across the Manchester United penalty area and Haaland finished at the back post, only for the assistant referee to raise the flag. Semi-automated offside technology subsequently established that Haaland had timed his movement correctly: Patrick Dorgu was playing him onside by approximately three centimetres.
Ordinarily, this is precisely the sort of situation VAR is designed to correct. A marginal position can be measured, an incorrect flag overturned and a legitimate goal restored. The problem was that while the officials concentrated on Haaland’s position, another City player had become relevant to the decision.
Fernandez was beyond United’s defensive line when Gvardiol delivered the ball and made a clear attempt to reach the cross before it continued towards Haaland. He failed to make contact, and VAR used that fact as part of the justification for allowing play to continue. According to the Premier League’s independent panel, that was where the review went wrong.
The panel voted 5–0 that the goal should not have been awarded and 5–0 that the VAR intervention itself was incorrect. Its conclusion was that Fernandez attempted to play a ball close to him and, in doing so, affected the ability of Manchester United’s defenders and goalkeeper to play it. The correct process would therefore have been to identify Haaland as onside but then consider whether Fernandez had committed a separate, subjective offside offence.
How can a player be offside without touching the ball?
The incident exposes one of football’s most persistent misunderstandings. Simply standing in an offside position is not an offence, but touching the ball is also not the only way to become actively offside. Under the Laws of the Game, a player can be penalised for interfering with an opponent even if the ball never reaches him.
That can happen when an offside player prevents an opponent from playing or being able to play the ball, clearly attempts to play a nearby ball and affects an opponent in the process, or makes an obvious action which clearly influences an opponent’s ability to play it. Fernandez therefore did not need to head Gvardiol’s cross for his position to matter.
Watch the move while ignoring Haaland and the issue becomes much easier to see. Fernandez attacks the cross from an offside position, meaning United’s defenders have to react to the possibility that he will make contact. The goalkeeper must also account for a potential touch that could change the direction of the ball. Fernandez misses it, but by the time the cross reaches Haaland, the defensive situation has already been influenced by his movement.
Howard Webb specifically highlighted the effect on Lisandro Martinez and the United goalkeeper when discussing the incident. This is also why a freeze-frame cannot settle the argument. The lines can establish where Fernandez was standing when Gvardiol played the ball, but they cannot determine whether his subsequent movement affected an opponent. That remains a subjective judgement for the officials.
The technology was right — and VAR was still wrong
What makes the incident particularly interesting is that semi-automated offside technology did exactly what it was supposed to do. It established that Haaland was onside by roughly three centimetres. The failure came afterwards, when the officials treated that answer as though it resolved the entire attacking phase.
Howard Webb has described what happened as a form of “tunnel vision”. Once the VAR had been presented with the evidence showing Haaland onside, the review became concentrated on correcting the assistant referee’s original flag. Fernandez’s position was noticed, but the focus shifted towards whether he had actually touched the cross rather than whether his attempt to play it had affected United’s defenders.
That distinction is important in the wider debate about VAR. Offside technology is extremely good at answering factual questions about position: it can establish whether a shoulder, knee or foot is beyond a defensive line to a degree of precision impossible for the human eye. It cannot make every judgement required by the offside law, because concepts such as interfering with an opponent still depend on interpreting movement and its effect on other players. In this case, the technology accurately answered the question it was asked, but the officials failed to properly address the next one.
Even the original offside decision was technically wrong
There is another unusual detail in the Premier League panel’s findings. The assistant referee was not vindicated simply because the final outcome should have been “no goal”. The original decision was also judged incorrect because the flag had effectively identified Haaland as the problem when Haaland was actually in a legal position.
That produced a strange chain of events. The officials initially arrived at what should have been the correct outcome — disallowing the goal — but for the wrong reason. VAR then correctly established that the original reason was wrong, only to miss the separate offence that should still have prevented the goal from standing.
Had the review been completed correctly, the process should have established that Haaland was onside, identified Fernandez’s possible interference and recommended that referee Michael Oliver review the subjective element at the pitchside monitor. Instead, the review stopped too early and City were awarded the goal. It is an unusually good example of why VAR decisions cannot always be reduced to whether the lines were drawn correctly: the three-centimetre measurement was accurate, but it represented only one part of a much larger decision.
What does the Premier League’s verdict actually change?
Pro Ref had already acknowledged an error after the derby, while Webb subsequently explained publicly how the review had failed. VAR Matt Donohue and assistant VAR Blake Antrobus were also left without Premier League appointments for the following round of matches. The September 23 findings now provide a formal independent assessment of both the original decision and the subsequent intervention.
The Key Match Incidents Panel consists of five members who independently review major refereeing decisions. From the 2026/27 season, the Premier League is publishing its findings every week as part of an attempt to provide greater transparency around officiating. Across Match Round 4, the panel examined 19 key incidents and judged 95 per cent of the on-field decisions correct; Haaland’s goal was the only VAR overturn from the round that the panel subsequently considered incorrect.
The unanimous nature of the verdict is also significant. This was not a borderline 3–2 interpretation on a highly subjective incident. All five members concluded that the goal should not have stood, and all five considered the VAR intervention incorrect. What the verdict does not change is the result: Manchester City’s 1–0 victory remains in the record books, Haaland keeps the goal and City keep the three points, because the KMI process assesses officiating decisions rather than retrospectively altering results.
Haaland was onside, but the goal was still offside
For years, arguments about offside have increasingly been reduced to lines, body parts and millimetres. The introduction of semi-automated technology has pushed the game even further in that direction because factual positioning can now be established faster and with extraordinary precision. The Manchester derby showed why that will never eliminate interpretation from the law completely.
Haaland’s position could be measured, while Fernandez’s impact on Martinez and the goalkeeper had to be judged. One question belonged to technology; the other still belonged to the officials, and concentrating too heavily on the first caused them to miss the second.
That is ultimately why four statements that initially seem incompatible can all be true: Haaland was onside, Fernandez never touched the ball, the technology worked correctly and the goal should still have been disallowed. Far from being another argument about whether an offside line was drawn in the right place, the incident was a reminder that the most difficult part of the modern offside law often begins after the lines have disappeared.
Published by Patrick Jane
23.09.2026