The Empty VAR Room: When Football Verdicts Are Built From Nothing

The big screen in the VAR operations room is black. Not a power cut. There is simply nothing left to show. The clock on the wall keeps counting, the referee stands in the penalty area with a hand on his earpiece, waiting for a signal from the room where I sat for many years. And what I learned there, after all of it, is uncomfortable: when there is no footage, people still blow the whistle.

That is the cruellest paradox of modern football. We build dozens of cameras, sensors inside the ball, probability models, tracking systems for every stride a player takes — all to serve one purpose: to turn judgement into evidence. Then comes the moment the system falls silent, and the verdict still has to be delivered within seconds. Nobody allows a referee to say he does not have enough data. Nobody allows an analyst to say he does not yet know. That void is where football's most expensive mistakes are born.

Context: every meeting room has a screen

From the 2026/18 season, Serie A entered the VAR era. I was there at the beginning, when the system was crude, when referees were still unused to being watched from behind, when the stands were still unused to a goal being chalked off after three minutes of waiting. Those days taught me that VAR shifted power on the pitch: from the man with the whistle to the man with the footage.

At the same time, a second shift happened in silence, behind the dressing-room door. Clubs began hiring people who never touch the ball: data scientists, engineers, modelling specialists. They brought handsome dashboards, forecasting models, metrics nobody named a decade earlier. They spoke of expected goals, of passes allowed per defensive action, of the expected value of a pass. And they delivered conclusions, very fast, very confidently.

I want to stop here. Modern football has two kinds of decision-maker. One is the referee, who must conclude within three seconds using slow motion. The other is the analyst, who can sit for three weeks with a spreadsheet. Both face the same temptation: to conclude before the evidence is sufficient. The difference is that the referee is abused live on television, while the analyst is quoted as truth.

Based on my experience watching matches across many seasons, the most dangerous error is not one of the eye. It is one of emptiness: when people fill a data gap with guesswork, then present that guesswork as a discovery.

The night at San Siro

In late October 2026, at San Siro, Milan hosted Juventus. I was in the operations room as assistant VAR. The match drifted towards the visitors, and Gonzalo Higuaín led the way with a brace. There was one passage of play where the feed showed me the Argentine striker standing just below the home defence — a gap of about two handspans, something the naked eye at an assistant referee's angle could never distinguish.

I hesitated. I was afraid of being wrong. I did not recommend a review. The match ended in a Juventus win, and that evening, in front of the whole team of officials, I heard the referee supervisor tell me plainly that I had failed at the one job the system paid me to do.

I tell this story to talk about method. Over the following month I sat down and reviewed 47 similar offside situations — not any 47, but situations sharing the same geometry: a small gap, an angled camera, the running speed of the defensive line. I built a checklist of 37 criteria. From then on, whenever I sat in front of a monitor, I had a decision tree instead of a hunch.

My problem that year was not weak eyesight. The problem was that I let a data gap fill itself with fear. Football is a game of error margins, but the winner is the one who knows which error is worth making.

The Mbappé lesson

In the summer of 2026, at the World Cup in Russia, I was invited to comment on VAR for a television channel during the round-of-16 match between France and Argentina. Before kick-off I did something many colleagues considered pointless: I pulled data from Kylian Mbappé's last 14 matches in Ligue 1 and the Champions League, and reconstructed his acceleration phases.

The data showed his peak sprint speed at roughly 36.5 km/h, while the average sprint speed of Argentina's defenders was nearly 3 km/h lower. I wrote an analysis arguing that Argentina's defensive structure would break between the 60th and 70th minutes, once the defensive line lost its legs and Mbappé began being released into the space behind.

He won a penalty and scored twice. France won 4-3.

I am not telling this story to boast. That piece was only possible because I had real data, from real matches, with figures that could be checked. Had I only had an empty dataset and a belief, I could still have written the right conclusion for entirely the wrong reason. And next time, with a different player, I would be wrong. Belief is not reusable. Method is.

The trap of the empty dataset

This is the part I consider most important, and the part least discussed inside professional football analysis rooms.

There is a type of failure I call the silent failure. A data-collection process fails without reporting an error. It does not crash. It shows no red warning. It simply returns an empty set: no passages of play, no metrics, nothing at all. The user on the other end, seeing no anomaly, carries on as usual.

Inside a dressing-room environment, this kind of failure is far more dangerous than an obvious one. An obvious error forces you to stop. An empty dataset does not. It lets you fill the gap with imagination. And human imagination, placed inside an analytical framework that looks professional, produces conclusions that sound extremely convincing.

I have seen it in many forms. A transfer report built on a single post from an aggregator account, which three days later becomes a well-placed source in ten other articles. A tactical breakdown built on two matches, one of which the club played with a patched-up XI because of injuries. A forecasting model trained on data from a season when the club was still in a lower division.

In every case the structure of the failure is the same: there is a gap, and the gap is filled with a story. The better the story, the fewer people check it.

The analyst in the dressing room

I hold a fairly hard line on the wave of data analysts who moved into clubs over the past decade. They bring real value. But their conclusions often sit apart from the actual rhythm of a match.

The reason is not competence. It is the data source. A model is only as good as what is fed into it. And football data, however carefully collected, still misses the decisive things: the mental state of a defender who has been beaten twice, the weight of a roaring stand, the breathing of an entire back line in the second period of extra time. None of that lives in a spreadsheet. And when it does not live in the spreadsheet, the analyst must know what he is missing.

This is where I return to the VAR lesson. A good referee is not one who gets every situation right. He is one who knows which situation he does not have the angle for, and calls for a review. A good analyst is the same: the one willing to say the available data is not enough is the trustworthy one, not the one who always has an answer ready.

The problem is that our profession does not reward silence. A newspaper needs copy. A club needs a report. Fans need a verdict. In that noisy market, the person who says he does not yet know is seen as weak, while the person who invents a tidy story is seen as visionary.

When the stands fall silent

The season played in empty stadiums because of the pandemic was a natural experiment nobody designed. A variable that had always existed but was never fully measured — crowd noise — suddenly vanished from the equation.

I spent weeks analysing one club's transition data during that period, and the results forced me to rewrite a good number of my old assumptions. Their counter-attacking defensive capacity dropped markedly once there was no crowd to drive the press. Models that had forecast well suddenly drifted off, not because the algorithms were wrong, but because an important column of data had disappeared from the real world.

The lesson is clear: we tend to believe our models cover everything, until a variable nobody noticed abruptly leaves the table. At that point, an honest analyst must admit the model is running on an incomplete dataset, rather than pretend it is still correct.

The contrarian view: silence is also a verdict

There is something both referees and analysts learn late: not deciding is also a decision. When a referee stands still because he is unsure, he has chosen to let play continue. When an analyst submits a report that says nothing clearly, he has chosen to let the coach guess. Neutrality does not exist. There are only choices that are spoken and choices that are hidden.

And here is the counter-intuitive part. We usually assume errors come from emotion: the referee pressured by the crowd, the analyst pressured by his employer. But in most cases I have witnessed, the error came from a missing column, and from the fact that nobody dared point at that gap during the meeting.

I set myself a rule, and I recommend it to anyone whose job is reading matches. Before reaching any conclusion, I ask: if tomorrow someone demanded the evidence, what would I show? If the answer is a feeling, I delete the conclusion and write it again. Before blowing the whistle, I review myself. Every verdict needs one review, including the verdict of the data.

The arms race and where the small clubs stand

Recent transfer windows keep showing me the same thing. The giants spending hundreds of millions of euros are not really buying a player; they are buying a statement to the market. It is a brand arms race, in which a high fee becomes part of the communications campaign.

The real value of the market sits elsewhere. It sits with smaller clubs working with their own data, buying cheap, developing for two seasons, then selling at a multiple. There, every decision has to rest on evidence, because nobody can afford to be wrong. And precisely for that reason, those clubs are often where analytical method is respected most.

If I had to choose one place to learn football analysis, I would not choose the meeting room of a champion. I would choose the small office of a mid-table club, where every dataset is tied to a decision that matters.

Academies and the branding trap

At the same time, another wave has risen: former stars opening youth academies. Most of them operate more as a personal brand than as a training institution. They sell image, short courses, and the feeling of being near someone who was once famous.

The systemic shortfall sits lower down: the training of grassroots coaches. A country can have hundreds of academies bearing the names of stars, but if the people teaching children in schools and amateur clubs are not properly trained, the talent pipeline stays as narrow as before. This is an area where the data shows a serious lack of investment, yet it is rarely discussed because it generates no glamour.

What technology cannot fix

People often say technology will gradually erase controversy from football. I do not believe it. Technology does not kill football, it kills blind faith. But it cannot erase the void. The more data there is, the subtler the void becomes: no longer a case of seeing nothing at all, but of seeing enough to think you know.

That is why VAR, after almost a decade, still generates fiercer arguments than before it existed. The system does not make a verdict clear. It makes a verdict checkable. And once a verdict is checkable, people start to notice how many past verdicts were issued with nothing behind them.

I think of transfer reporting. Every window, thousands of fees, wages and contract lengths circulate across news sites. Most can never be verified, because no club publishes contract details. Yet fans still argue, still rage, still celebrate, on the basis of facts of unknown origin. That is the blacked-out VAR room again, except this time the whistle-blower is us.

What a serious process looks like

If I had to describe a trustworthy football analysis process, I would start with the least attractive parts.

It starts by stating the source. A fact without a source is a meaningless fact. It continues by stating the date. A conclusion that was right in October can be wrong in March, because squads change, coaches change, fixture lists change. It requires a clear separation between event and inference: a player running 36 km/h is an event; that he will break the opposing defence is an inference. The two must not be set in the same type size.

And it ends with what I consider the most important element: a column reserved for what is missing. If my report does not have that column, it gets sent back. I do not trust my eyes, I trust slow motion. But when there is no slow motion, I must record in the minutes that there is no slow motion, rather than close my eyes and rule.

A closing thought

Football is entering a new cycle with a denser calendar, more competitions, more data and more voices. In that environment, the greatest value of an analyst is not the ability to conclude quickly, but the ability to know when to stop and say that here, there is not yet evidence.

Emptiness is not shameful. It is data. And the person who knows how to read it will never be fooled by a conclusion built out of nothing.