VAR's Silent Failure: When the System Validates a Decision That Was Never Checked

Minute 67 at the bet365 Stadium: A Void Nobody Named

Minute 67, bet365 Stadium, May 2026. Theo Walcott burst into the box, Bruno Martins Indi grabbed his shirt and pulled him down. Mike Dean stood less than ten metres away and did not blow his whistle. I sat in a small newsroom in Manchester and replayed twelve camera angles for forty straight minutes. The camera behind the goal showed Martins Indi's hand on Walcott's shoulder for exactly 0.4 seconds. The match finished 2-2. English football had no VAR in that season.

That night I wrote an analysis built on IFAB Law 12, and the piece was shared more than five thousand times. But I skipped one detail, and it took me years to realise it mattered more than the shirt pull itself. Nowhere in the match record was there any note showing the referee had seen the incident and dismissed it. Nor was there a note saying he had missed it. The gap sat quietly inside a valid document, and the document remained valid.

A whistle can change a fate, but it cannot change the truth on the pitch. What I never wrote in that piece was this: on some nights, the truth was never brought into the room.

Context: VAR Is Not Wrong, But Sometimes It Never Runs

The Premier League brought VAR into operation from the 2026-20 season. Across the first two seasons I built my own tracking sheet in Excel: every intervention, every check, every "no clear and obvious error" verdict, each with a column for time, a column for available camera angles, and a column for the decision-maker. By May 2026, when the league paused for COVID-19, my sheet covered 92 matches. One figure stopped me cold: 34 per cent of goals reviewed by VAR involved offside offences, and most of those were handled by the semi-automated offside tool with margins measured in millimetres.

But the column that cost me sleep was a different one. I named it the empty-input column. Those were the cases where the system returned a completed check while the record showed only one or two usable angles, or angles blocked by players, or an unclear audio channel between the VAR hub and the on-field referee. Formally, the process ran. Substantively, there was nothing to run.

People hate VAR because it is slow; I value it because it is not in a hurry. But slowness only has value when there is content inside it. A slow and empty VAR is worse than a fast referee who is wrong, because it manufactures the illusion that everything was examined.

This is the point most football commentary ignores. Fans argue about whether a referee was right or wrong. The better question is whether the referee was ever given enough evidence to be either. A referee's mistake does not vanish with the whistle; it lives on through every season. And one kind of mistake lives longer than all the others: the mistake born from an empty input that was still validated.

I call it the silent failure.

The Core: Four Layers of a Refereeing Decision

To explain silent failure in football, I borrow the structure that data engineers use to test an automated process. A process can fail in two ways. The first is a loud failure: it throws an error, stops, and everyone knows. The second is a silent failure: it runs to completion, returns a correctly formatted result, and nobody knows that there was never any data inside it.

Football has plenty of both. And the second kind is far more common than we think.

Layer one: input validation

Before a referee issues a verdict, that person needs a minimum dataset: standing position, line of sight, speed of the passage of play, direction of contact, and, for VAR, the number of usable camera angles. When that dataset is complete, errors still happen, but they are errors of judgment, and they can be corrected through professional debate.

When that dataset is empty, we are no longer debating judgment. We are debating a void.

I tracked a large number of Premier League and La Liga matches through the 2026-20 season for comparison. What I found was that the most explosive controversies were rarely the incidents with a clean angle and a hard conclusion. They were incidents with poor angles where a firm conclusion was still delivered. Sheffield United lost seven points that season to marginal decisions, and in at least two of those cases the angle log recorded fewer than three usable views of the point of contact or the ball's position.

A referee without enough angles still has to conclude. That is the fracture point.

Layer two: the clear-and-obvious threshold

The Premier League's VAR protocol uses a very high intervention threshold. A referee is only recommended to review the monitor when the VAR hub identifies a clear and obvious error, or a serious missed incident. This threshold is designed to protect the flow of the game, and in principle I support it.

But the threshold contains a logical hole. Without enough camera angles, the hub cannot assert a clear error. And if it cannot assert a clear error, the protocol-correct conclusion is to uphold the on-field decision. In other words: insufficient data automatically produces confirmation of the original call.

That is precisely the mechanism that generates silent failure. The system does not lie. It simply has nothing to say, and is then recorded as consensus.

In my Excel sheet I separated this group and called it the empty-confirmation group. Its share across the 2026-20 season, as I logged it, was roughly 11 per cent of all checks. That is not a large statistical share. But it clustered in exactly the situations fans care about most: penalties inside the box, handballs, and contact incidents that decide scorelines.

Layer three: semi-automated data and the limits of precision

Semi-automated offside technology is the cleanest example of this paradox. It draws a line from body-tracking points captured by a multi-camera system. The output looks deeply scientific: a thin line, a toe, a distance in millimetres.

I have reviewed hundreds of these graphics. There is one thing the graphic never tells the viewer: which frame was chosen for the line, and why. The moment the ball leaves the passer's foot is usually determined by a camera running at a limited frame rate. When the ball travels at speed, being one frame out at 1/50th of a second translates into a meaningful distance on the grass.

Put differently, the millimetre line is built on a foundation that is not entirely millimetre-grade.

I say this not to dismiss the technology. I say it to show that even football's best data layer carries a fuzzy input zone. And when that fuzzy zone is concealed behind a beautiful graphic, fans lose the ability to doubt. All they retain is the ability to be angry.

Layer four: the psychology of the person who must conclude

This is the part I learned last, and paid for most dearly.

In July 2026, ten minutes after the European Championship final ended, I published an analysis of Bukayo Saka's penalty. I wrote that Saka took a five-step run-up, that his centre of gravity leaned slightly left, and that Gianluigi Donnarumma therefore had grounds to dive right. Every technical detail was correct. Saka's family and many Arsenal supporters reacted furiously, and they were right on one point I had ignored: a nineteen-year-old standing twelve yards out in front of an entire nation does not run up according to a biomechanical model. He runs up according to fear.

I had to reread five books on athlete psychology to understand that data does not lie, but data also does not say everything. And that loops straight back to referees.

A referee in the 88th minute of a derby, in front of seventy thousand people, in a passage of play with the VAR hub buzzing in his ear, is also a human being under pressure. In interviews with senior European referees in the years that followed, one phrase kept recurring: the hardest thing is not making the wrong call, but being forced to conclude while knowing you do not have the foundation for it.

Nobody is trained to say this sentence on a pitch: I do not know.

A referee holds three powers: to blow the whistle, to show the card, and to stand firm under pressure. The third is the most misunderstood. Standing firm does not mean always concluding. Sometimes it means tolerating the discomfort of not concluding.

Cross-Reference: World Cup 2026 and the Cost of Certainty

In June 2026, at the World Cup in Russia, I published a series defending VAR after Diego Godín's handball went unpunished in the Uruguay match against Saudi Arabia. My argument was tidy: VAR was not wrong, the operators were. I cited an average of 6.8 minutes of stoppage per match for reviews and argued that this was the price of accuracy.

A BBC editor replied bluntly that the piece was too rigid. Social media called me an emotional machine. I kept my technical position, but I conceded one thing: the way I wrote had closed the door on dialogue.

Six years later, looking back at the 2026 World Cup data, I see something different. In many of the handball reviews at that tournament, the problem lay not with the operators. It lay in the fact that the handball law was in transition, and referees were forced to apply a version of it they had not been trained on deeply enough. That, too, was an empty input, sitting at the training layer.

When the cathedral falls silent, only the laws speak. But the laws only speak clearly when the person holding the whistle has been prepared to understand them. Otherwise what speaks is simply a version of the law recited by reflex.

The Counter-Intuitive Angle: The Problem Is Not That VAR Is Slow

Most public debate about VAR revolves around time. Fans in the stadium wait, fans at home wait, players stand in the middle with hands on hips. It is a visible, easily hated experience.

My counter-intuitive view is this: waiting time is not the core problem. It is merely the most visible symptom.

If a VAR check lasts three minutes and ends in a decision built on five clear angles, fans will accept it, grumbling aside. If a check lasts twenty seconds and ends by confirming a decision that nobody actually examined, fans will lose faith. And faith lost that way is very hard to recover.

This is why I believe competitions are measuring the wrong indicators. They measure average review duration, minutes of stoppage, successful interventions. They do not measure how often a conclusion was delivered without sufficient input evidence. That is the most important metric, and almost nobody publishes it.

There is an argument I hear often from defenders of the status quo: if referees said they lacked data, matches would never end. I understand the concern, but I think it asks the wrong question. Nobody is asking referees to stop the game and hunt for more data. All that is needed is a small change in how we record things: when a conclusion rests on incomplete evidence, the record must say so.

One line of annotation. Nothing more.

But that one line would change the entire public conversation. It turns an apparently absolute verdict into a conditional one, and it makes transparent the grey area that everyone in football knows exists but nobody officially admits.

The View From the Stands: Pain Has No Column in the Record

I always keep one passage in every piece to step away from the spreadsheet and look up at the stands, because I once wrote a piece made only of numbers and was taught a lesson by my own readers.

What the data layer cannot record is how a supporter feels when their team drops points because of a line nobody can explain. That person is not angry at technology. That person is angry at being forced to trust a process whose interior they are never allowed to see.

The cultural comparison here is fascinating. In Spain, where I was born, debates about refereeing are openly emotional and almost a secondary sport. In England, where I work, debates are usually framed in the language of law and precedent. Each culture sees the same incident through a different lens, and both are partly right.

But there is one shared trait. Supporters in Madrid and supporters in Manchester react the same way to silent failure, once they notice it. They do not react with an explosion of rage. They react with a prolonged scepticism that erodes their trust in the sport they love.

That is why I frame this at the level of process, not at the level of individual referees. When I built my Excel tracking sheet in the 2026-20 season, I did it not to catch out individual referees. I did it to find a pattern. And the pattern I found was not located in the name of the person holding the whistle.

It was located in the gap.

Mistakes Do Not Vanish With the Whistle

Over many years in this trade, I have learned that the biggest refereeing controversies are not settled on the night of the match. They are settled, if at all, weeks later, when the data is thick enough and the emotion has cooled.

That creates an occupational problem for me personally. Modern football pushes journalists to file overnight. But if I write immediately after the final whistle, I risk reproducing the very error I criticise: concluding when the input is not yet sufficient.

I learned that lesson painfully. My piece on Saka at Euro 2026 went out ten minutes after the match. Ten minutes. I had enough technical data, but not enough human context. And I wrote with a certainty I had no right to.

Since then I have set two hard deadlines for every refereeing analysis. The first is an emotional draft written immediately after the match, purely to record what I saw and felt, never published. The second is the finished piece, written after I have replayed every angle, cross-checked the law, and verified every number against its origin.

I never file before asking myself: where does this number come from. If I cannot answer, the number does not make the page.

A referee's mistake does not vanish with the whistle; it lives on through every season. Silent failure does the same. It survives in the form of decisions nobody remembers, entered into history as verified fact.

A Proposal: One Validation Gate for the VAR Room

If I could put one proposal in front of the people who write the protocols, it would not be a law change. It would be a validation gate.

The gate works on a very simple principle. A check may only be closed when it meets a minimum input threshold: the number of usable angles at the key frames, the frame quality at the moment of contact, and the clarity of the communication channel. If the threshold is not met, the outcome must be recorded as a conclusion on limited evidence, with a specific reason attached.

Nothing in this proposal changes the final decision on the pitch. It only changes how we record that decision.

The first benefit is transparency. Fans would know why an incident was upheld, instead of having to guess.

The second benefit is better training. If a VAR hub routinely delivers conclusions beyond its input threshold, that is a signal for coaching intervention, not a signal to replace referees.

The third benefit, and to my mind the most important, is protecting the referee. Referees are currently judged on the final outcome of a process over which they control very little of the input. A validation gate would allocate responsibility more honestly between people and systems.

A Forward-Looking Thought

Football changes its laws once every three years, but the trust of its audience changes very slowly. That is the sport's greatest asymmetry, and no technology fixes it automatically.

What I hope for in the coming seasons is not a new camera, or an offside algorithm accurate to the last millimetre. What I hope for is a generation of referees and protocol-builders willing to say the hardest sentence in the VAR room: we did not see enough.

People hate VAR because it is slow; I value it because it is not in a hurry. But if that unhurriedness does not come with the courage to admit its own limits, it is nothing more than a delay dressed up in graphics.

And when the cathedral falls silent, only the laws speak. The question for those holding the whistle this season is this: if the laws speak while you were never given enough to hear them clearly, will you conclude, or will you dare to say that you do not yet know?