47 Seconds and 523 Matches: What VAR Data Says About the Gap Between Law and Emotion

Minute 55, June 16, 2026, in Kazan. Referee Andrés Cunha stopped the Group C opener between France and Australia, walked to the pitchside monitor, and the big screen replayed the moment defender Josh Risdon turned his shoulder into the shot. I was sitting in a Valencia radio studio with the microphone open and the crowd still ringing in my headphones. I told more than four million listeners that Cunha had got it wrong, that the ball had struck the armpit, and that the armpit is not part of the arm. I said it in the flat, certain voice of a man who had been reading the laws for thirty years. I believed I was right.

Seventeen seconds later, the colleague to my left touched my shoulder and read the amendment out loud. Since 2026, the boundary of the arm in Law 12 had been redrawn, and the armpit fell inside the handling zone. I was wrong on live radio. The editorial desk issued a correction within fifteen minutes. It was the first time in thirty years of work that a sentence of mine was publicly overturned, and I had no data on hand to defend myself. I once got one sentence wrong and lost an entire reputation. I wish I had known this back then.

I first learned the laws of football in 2026, in a county referee course in Ohio. My instructor, a former state-level official, opened the book and read the handling clause aloud: contact from the shoulder upward is not an offence. The version I carried in my head for the next two decades was the 2026 edition, when the arm boundary was still a grey zone that local referees interpreted for themselves. I believed I knew the law. I knew one version of it, and I called that version the law.

On June 1, 2026, the International Football Association Board (IFAB) issued a circular clarifying that the boundary between shoulder and arm sits at armpit level. On June 1, 2026, Law 12 was rewritten: a goal scored by hand, or from a ball that touches an attacker's hand or arm, is disallowed even when accidental. On June 1, 2026, a further circular clarified the case of a handball that creates a goalscoring opportunity. On July 1, 2026, the "unnaturally bigger" criterion was redrawn, and the automatic rule treating an arm above shoulder height as an offence was removed.

After that broadcast I built a law index organised by year and colour-coded it. Green means the clause is in force. Amber means IFAB has approved it but the effective date has not arrived. Red means it has been repealed, and the 2026 edition I once quoted sits in the red column. The law does not live in memory. It lives in data. Memory has no effective date. Data does.

In August 2026 I started a project many people at the time considered pointless: logging every VAR intervention in La Liga and the Champions League. For each one I recorded the fault code, the match minute, the distance from the offence to goal, an estimated ball speed, the number of camera angles the VAR team reviewed, and the time from the referee's hand reaching his earpiece to the final signal. By March 2026, when the pandemic stopped football, I had 523 matches in my database. Every one was a competitive fixture. No friendlies, no youth matches.

The number that stopped me was 47 seconds. Among offside incidents where the assistant raised the flag on the field and the VAR team later overturned it, the average interval from the start of the review to the final signal was 47 seconds. Seventy-four per cent of that group were marginal offsides, meaning less than 20 centimetres of separation. On a 40-metre pass travelling at roughly 25 metres per second, a 20-centimetre positioning error equals about eight thousandths of a second. The human eye cannot read that interval. The system can, but it takes 47 seconds to say so. VAR's latency is a match-management problem, not a technology problem.

I published a 48-page report on my personal blog, with the spreadsheet attached so anyone could audit every row. Its central proposal was simple: each review gets a maximum of 30 seconds. If the VAR team has not found clear and obvious evidence within 30 seconds, the on-field decision stands and the clock stops running. That principle comes from the language of the VAR protocol itself: intervention exists for clear errors. A clear error is something you can see in seconds. If it takes seventeen viewings from six angles, the incident has stopped being clear. The Valencia football federation invited me to advise on process reform after that report.

Three years later the hardware of the argument had changed. At the 2026 World Cup in Qatar, semi-automated offside technology entered official operation, cutting positioning time to a few seconds. From the 2026/23 season, the Champions League adopted the same system. From 2026, referees at several major competitions began announcing decisions over a body-worn microphone. All three changes move in the direction my data pointed to: shortening the silence. A crowd can absorb a decision it disagrees with. A crowd cannot absorb prolonged ambiguity.

Technology only solves half the problem. The other half sits at the intervention threshold. Across my 523 matches, VAR teams intervened in handball incidents more often than in offside incidents, even though marginal offsides occur at a higher density per match. The reason is not in the law. It is in the image. A frozen frame showing a ball on an arm is visually persuasive to a television audience. An offside line two centimetres off is not. VAR teams are pulled toward the incidents that are easy to present, not the incidents that are most serious. I call that VAR's media bias.

When I split the data by club wage rank, a sample of 523 matches is not enough to conclude bias, and I will not say otherwise. It is enough to conclude latency. Incidents involving clubs with home crowds above 50,000 took an average of 9 seconds longer than the rest of the dataset. Stadium noise shapes decision time. That is an environmental variable, not a conspiracy.

Alongside the VAR project I tracked a second group: players returning from cruciate ligament injuries. Across the 523 matches I flagged 41 cases of a player starting within three matches of returning from a long-term injury. That group received their first disciplinary card an average of 11 minutes earlier than their own previous-season baseline. The mechanism lies in approach distance: a returning player tends to be half a step slower, and that half step turns a legitimate challenge into an offence. The referee is correct. The price the player pays is not in the law.

This is where I part company with most commentators. When a player returns from injury and is told to "prove himself", the public is applying a standard that appears in no clause anywhere. No clause says a player in his first match back must perform as he did before the injury. Yet the pressure exists, it is measurable in minutes played and substitute appearances, and it raises the probability of re-injury. I do not write this to defend anyone. I write it because I counted.

There is one point data cannot settle. The pressure that makes referees treat big clubs and small clubs differently is real, and it does not come from the VAR room. It comes from the stands, from the evening bulletins, from the number of cameras pointed at a manager's face after every whistle. A referee has about three tenths of a second to decide. In those three tenths he is not calculating. He is reacting, and the reaction was shaped by everything he absorbed in the week before. That is why I do not treat "big-club bias" as a personal accusation. I treat it as a variable to be measured.

The larger blind spot sits with the audience. We remember a match. We do not remember a season. One match is a story. Five hundred matches are a law. A shocking decision is not reckless if it is built on five hundred foundations. A decision that looks reasonable inside one match can be a statistical outlier across five hundred. Emotion is not wrong. Emotion simply has no sample. Without a sample, people are forced to believe in motives. And once they believe in motives, every decision becomes evidence for what they already believed.

Referees do not need protecting. They need to be understood through correct data. In fifty-one years of watching football on two continents, I have never seen an argument end with the full data being published. I have seen many arguments end with a single figure lifted out of context. The difference between those two things is my entire profession.

At 67, I do not need to remember everything. I need to know how to find what is right. Tomorrow there will be another controversial incident, and another man will stand in front of a microphone and state with certainty that he knows the law. I hope he opens the dataset before he opens his mouth. The rest, leave to the numbers.