VAR, the Handball Law, and the Price of One Wrong Assertion

In June 2026, at the Kazan Arena, I sat in a small studio in Valencia, headphones just tight enough, in front of me a sheet of paper with only a few lines about the France-Australia match. In the 55th minute, referee Andres Cunha stopped play, ran outside the penalty area and drew a rectangle in the air. The ball had struck the hand of Josh Risdon inside the box. I said into the microphone, clearly and firmly: the ball hit the armpit, not a foul. Four million listeners heard that sentence. None of them knew that the man explaining the law to them was citing a version that had been out of force for two years.

Andres Cunha awarded the penalty. Antoine Griezmann opened the scoring. By the 62nd minute, Mile Jedinak equalised, also from the spot, after another VAR review. In the 81st minute, Paul Pogba put France ahead with a move that goal-line technology had to confirm. The match ended 2-1. Three situations, three technologies, three times a referee had to stand between two waves. But what I remember most is not the scoreline. I remember the colleague sitting beside me, after the first half, leaning in and saying: "Since 2026, the armpit zone has been part of the definition of the hand." I said nothing. In thirty years in the trade, that was the first time I had been contradicted directly, in public, by a single short sentence.

That night, back home, I reopened the 2026 IFAB law book I still kept in a drawer. The pages had yellowed. The clause I read belonged to another time, another world, where referees had only their eyes and a whistle. Football had passed that milestone long ago, while I had stayed behind. I folded the paper, put it away, and began to understand that what I had lost was not my memory. I had lost the habit of verifying before speaking. The law does not live in memory, but in data. I had forgotten that at the exact moment I was most confident.

A law that does not live in memory

The story of the armpit does not begin in Kazan. It begins in the closed meetings of the International Football Association Board (IFAB), where the lawmakers sit down and try to define something the naked eye cannot see: the boundary between the shoulder and the arm. Throughout the 20th century, the handball law had only one keyword — "deliberate". Referees were asked to judge intent, something not even machines can do, let alone a human running twenty metres away with his view blocked by players.

In 2026, IFAB introduced a geometric definition. The arm, under the new law, begins at the bottom of the armpit. Above that line is the shoulder, and the shoulder may touch the ball. Below that line is the hand, and a hand touching the ball inside the box is a penalty. It sounds simple, but this was the first time football law accepted such a vague physical limit — a line that does not exist on the human body, only in the imagination of the lawmakers and in the eye of the referee. Anyone who has tried to draw the armpit line on a photograph understands the problem: the same incident, three people draw three different lines.

In June 2026, IFAB amended it again. This time they added two notable changes: an accidental handball by an attacking player in the build-up to a goal is an offence, and a ball touching the hand after touching that same player's head or body is not an offence. Each amendment is another occasion on which commentators, experts and fans must relearn from scratch. And here is what I want to stress: the effective date of a law matters as much as the law itself. Quoting the right words but the wrong year of enactment is a mistake just as destructive as quoting it entirely wrong.

I say this not to excuse myself. I say it because I am the man who committed exactly that mistake, in front of four million people, and was corrected by a colleague with a sentence so quiet I only heard it clearly because the microphone was already off.

Why a law expert can still be wrong

There is a question I am asked many times at seminars in Valencia: how could a man with thirty years in the trade mistake a clause amended two years earlier? The answer lies in how the human brain stores law. We do not store clauses; we store conclusions. When I was young, I learned the handball law by memorising situations: this one a foul, that one not. After thousands of situations, I no longer remembered the original clause; I only remembered the conclusion nailed into my head. Each time the law changes, the old conclusion does not evaporate. It sits there, waiting for the right moment of pressure to resurface.

France-Australia was that moment. VAR had just been introduced to the World Cup for the first time, referee Cunha had just made the rectangle sign, and in my head the old conclusion fired before I could look anything up. That is the mechanism I call "confident memory" — more dangerous than ignorance, because it comes without any accompanying sense of doubt.

After the night in Kazan, I built a law reference updated year by year. Every clause has a date of enactment, a date of amendment, and a column stating which version is still in force. I forced myself to cite specific clause numbers, with the note "under the 2026 amendment", instead of speaking in generalities. At first this habit made my writing heavy. Readers complained that I stuffed in too many dates. But I kept it. I once got one sentence wrong, and lost an entire reputation. If only I had known this back then. A reputation in the law-explaining trade is built on thousands of correct sentences and can collapse on one wrong sentence broadcast live.

Out of that fear came a decision I myself did not expect: to stop relying on memory, and start counting.

523 matches and an error codebook

In August 2026, I began a personal project that at the time I treated as a way to repay my debt. I logged every decision involving VAR in La Liga and the Champions League: type of error, match time, spacing between incidents, ball speed, the camera angle the VAR team used, and the time from the referee's signal to the final decision. I built an error codebook of five groups: offside, handball, contact inside the box, severity of the foul, and the red-card threshold.

By March 2026, when the pandemic forced European football to a halt, I had logged 523 matches. I did not select the sample on a whim. I took every match with VAR in that period, including matches I found dull, including matches with no controversy at all. That is the difference between a dataset and a highlight reel. One match is only a story. Five hundred matches are the law.

The first finding made me stop. In the offside group, 74% of overturned decisions came after an average wait of 47 seconds. The figure 47 seconds does not mean referees are slow. It means the process is waiting for something it does not need to wait for. Most of that time is not spent drawing the offside line, but choosing frames, synchronising frames, and arguing over which player touched the ball first.

The second finding concerns handball — precisely the subject I got wrong in 2026. In the handball group, the rate of overturned decisions was markedly higher than in the offside group, but the waiting time was shorter. In other words: with handball, referees decide faster but also err more. With offside, they decide more slowly but more accurately. This is a paradox I have never seen anyone analyse: the accuracy and the speed of VAR do not move in the same direction, and we are optimising the wrong variable.

The third finding concerns the probability of a decision changing by match time. The later the match, the higher the probability that a contact inside the box is reviewed, while the probability that the referee's original decision is overturned actually falls. In other words, referees run to the monitor more often in the 85th minute, but change their minds less often than in the 25th. I do not yet have enough data to assert a cause. I can only state a degree: the available data suggests a trend, and I refuse to turn that trend into a firm conclusion.

47 seconds and the value of slowing down

I published a 48-page report on my personal blog, with an Excel file so that anyone could check it themselves. In it, I proposed a 30-second limit for each review, for a simple reason: if a situation cannot be clarified within 30 seconds using the available angles, it is not clear enough to overturn an on-field decision. I called it the "reasonable-doubt threshold" — if the evidence is not strong enough to cross that threshold within the allowed time, the original decision stands.

The interesting part was the reaction. Not from referees, but from fans. Many wrote to me that they would rather wait 47 seconds for a correct decision than 30 seconds for a fast but wrong one. That is a perfectly reasonable objection, and I recorded it in the second version of the report. I did not withdraw the proposal, but I added a condition: the time limit should apply only to frame selection, not to analysis. Because what irritates viewers is not the waiting, but the feeling that time is being spent on an argument that should not need to exist.

The Valencia football federation invited me to advise on process reform. In those meetings, I learned something data cannot teach: most of the tension in the VAR room does not come from the law, but from communication. The on-field referee cannot hear clearly what the VAR team is saying, the VAR team does not know what the on-field referee has seen, and both sides are trying to defend their own decision rather than jointly seek the truth. When I proposed standardising the language of exchange — each sentence containing only one piece of information, each piece accompanied by image evidence — average processing time fell markedly in drills.

When I counted every passage of play, I understood that the law does not judge anyone. It only waits to be applied correctly. The fault does not lie in the clause. The fault lies in the gap between the clause and the person applying it, under pressure, within seconds.

Why referees treat giants and small clubs differently

This is the part that has cost me the most friends in the trade. Many hold that saying referees favour big clubs is pseudoscience, a surrender to the conspiracy theory of the stands. I do not think so. I think there is a very concrete mechanism, and that mechanism is measurable.

In my dataset of 523 matches, I separated situations by crowd noise at the moment the incident occurred — estimated through stadium capacity and the intensity of the match. The preliminary result: in stadiums with large, loud crowds, the referee's initial decision time was shorter, and the probability that this decision was overturned by VAR was higher. Referees are not biased because they like big clubs. They are under the pressure of an enormous wall of sound, and under that pressure the human brain tends to choose the socially safe conclusion — the one the crowd wants to hear.

In the opposite direction, when a small club benefits from a controversial decision, I observed what I call "default suspicion": major media outlets spend more airtime dissecting that decision than when a big club benefits from a similar decision. This is an effect at the media level, not the referee level, but it feeds back into the pressure on referees in later matches.

I do not claim there is a conspiracy. I claim there is a mechanism. And a mechanism that can be measured can be fixed. Referees do not need to be protected. They need to be understood through correct data. Protection through emotion only makes both sides more convinced of their own story.

Injury and the price of demanding that players "prove" themselves

During the same period in which I logged VAR data, I tracked another group of situations: players returning from long injuries. I group these two topics into one article because they share the same error of thinking. With VAR, we conclude from a single frame. With injury, we conclude from a single match. Both are conclusions from a sample that is far too small.

In the 2026-2026 season, I recorded cases of players returning from anterior cruciate ligament injuries and playing a full 90 minutes in their very first match back. My sample is small, and I say so clearly. But the trend is clear: the group of players demanded to "prove themselves" immediately — usually those who had just signed a big contract or were in a period of criticism — had a higher rate of re-injury within a few months than the group rotated in gradually. I do not have enough data to speak of biological causes. I can only speak of pressure.

The problem is that the pressure to "prove" is created by us — the writers, the commentators, the people asking questions in press conferences. When a player returns from injury and we ask whether he still deserves his salary, we are planting a seed that will sprout weeks later into a new injury. Demanding that a player prove himself in his comeback match is not being strict. It is cruel, and it is counterproductive in professional terms.

The way I write about this subject is directly shaped by the lesson of 2026. I do not say "this player is finished". I say: the available data suggests a trend, the sample is still small, more monitoring is needed. That is not a compromise. It is honesty about my own limits.

VAR does not reduce controversy. It relocates it.

This is the counterintuitive view I want to devote a section to making clear.

When VAR was introduced, people promised it would end controversy. Six years later, controversy has not disappeared. It has only moved. We used to argue about the on-field referee's decision. Now we argue about the decision of the person sitting in a closed room, whom we cannot see, cannot hear, and cannot question directly. Controversy does not shrink. It becomes more opaque.

That is why publishing the audio of exchanges between referee and VAR team — which some leagues have begun to do — matters so much. Not for entertainment. But to shift controversy from "what he thought" to "what he said". Once the reasoning behind a decision is brought into the light, controversy can focus on the clause rather than on the person. And controversy about the clause is the healthy kind, because it can be resolved by amending the clause.

A second counterintuitive point: I believe VAR is being judged by the wrong yardstick. People measure it by how many times it correctly overturns a wrong decision. But the right measure must be the consistency of similar decisions across many matches. A system can overturn correctly many times and still be inconsistent, if each overturn follows a different standard. Consistency is what makes fans accept a decision, even when they disagree with it.

I tested this hypothesis in my dataset. Geometrically similar handball situations in different matches often received different rulings, and the inconsistency rate was higher in leagues with several different VAR teams rotating. I do not conclude that one VAR team is better than another. I conclude that diversity in training produces diversity in judgement, and that diversity, not technology, is the source of most modern controversy.

The trap of the empty conclusion

There is a professional lesson I want to tell, because it relates directly to how I work now.

In an internal analysis project I once took part in, my team received an almost empty input dataset — no title, no source, no events. The requirement was to "fill in the report template". This is the most dangerous moment in the analysis trade. An empty template always calls to be filled. And the human brain, especially the brain of a long-serving professional, is always ready to fill the gap with whatever sounds most plausible.

I saw that happen. When data is empty, people do not say "insufficient information". They say "in my experience". They invent a match, a club, a transfer figure, and present it with the same confidence as when they have real data. That is exactly the mistake I made in 2026, only different in scale.

From then on, I apply a hard principle: when data is insufficient, the only correct product is a structured null result — stating clearly what is missing, where it is missing, and what is needed to fill it. An honest null conclusion is worth more than a complete but fabricated one. In football, this is even truer. A shocking decision is not reckless, if it is built on five hundred foundations. But an ordinary decision, built on a single frame, can be a disaster.

I remember once a reader sent me a short email: "You say the available data suggests a trend. Why don't you just state the conclusion?" I replied that I had once stated it plainly, and I had been wrong. Caution is not weakness. It is the consequence of a specific scar. At 67, I do not need to remember everything. I need to know how to find the right thing.

What is coming at the next World Cup

Looking ahead, I see three changes that will reshape how we argue about referees in the coming years.

First, semi-automated offside technology. After being used at the 2026 World Cup and gradually spreading across European leagues, it cuts the time to determine offside to a few seconds, but it concentrates all the controversy into a single point: the moment the ball leaves the foot. This is a new paradox — we will argue over a frame at millisecond speed, while the human eye is incapable of that. I expect this to reduce arguments over "offside or not" but increase arguments over "should we intervene at millisecond level at all".

Second, on-site publication of exchange audio. When viewers hear the reasoning live, they will begin to judge the process rather than only the outcome. This is the biggest cultural change, and it will force VAR teams to standardise their language, much as I proposed in Valencia.

Third, amending the handball law toward simplification. I believe that within a few years, IFAB will be forced to accept a cruder but more consistent definition — perhaps dropping the notion of "deliberate" entirely and replacing it with a fixed geometric limit, or ruling that any handball in the box is an offence regardless of intent. That will annoy many people. But consistency matters more than subtlety, because a subtle law applied inconsistently produces more injustice than a crude law applied evenly.

What I want to leave the reader is not a forecast, but a question. The next time you watch a controversial decision and feel certain that the referee was wrong, ask yourself: what are you relying on? One frame, or five hundred matches? One memory, or a law book with a date of enactment?

In my trade, that question is not a rhetorical device. It is the line between a judgement and a lie beautifully presented. And I have stood on the other side of that line, in front of four million people, to know how thin it is.

There is one thing I learned after all these years: football does not lack people who deliver conclusions. Football lacks people willing to spend ten minutes looking it up before speaking. If this article makes even one reader pause for a beat before asserting something about a referee, then my 523-match project has repaid part of the debt I incurred on that night in Kazan.