The Empty Frame and the Irreversible Verdict: The Real Limits of VAR

Minute 38, France against Australia at the 2026 World Cup. The ball struck Josh Risdon's arm inside the penalty area. The referee, less than ten metres away, did not blow his whistle. From where he stood, the Australian defender's arm was tight to his body, the moment passed too fast, and a correct decision made in a split second could not be undone by a hunch. Then the earpiece crackled. The VAR room in Moscow had a different frame. The decision was reversed, France were awarded a penalty, Antoine Griezmann opened the scoring, and the match turned on a different axis. I filed an analysis that night. The 2026 World Cup was the first to deploy VAR at scale, and across the group stage alone the number of penalties generated from the video room ran three times higher than at the 2026 World Cup. Many readers took that as proof the technology had beaten human error. Seven years on, I am not so sure. The trouble is not whether VAR catches mistakes. The trouble is the occasions when a VAR team is forced to deliver a verdict while the evidence set on the table is close to empty. In Vietnam, VAR does not operate with the same hardware as the Premier League. It arrived in the V.League along a multi-year rollout, with the number of covered matches rising season by season, and each match covered by a limited number of camera angles. In Europe's top divisions a single fixture can be captured by dozens of lenses, plus ball-tracking and virtual offside lines. In many other matches, the number is three or four. The law is identical. The standard of adjudication is not. That is the point I want to state plainly from the start: football does not lack rules, it lacks people who read the rules in the language the rules are written in. Law 12, covering physical challenges and handball, sits in the same FIFA document sent to every member association. But when a referee in Melbourne, a referee in Shanghai and a referee in Hanoi watch the same incident, they are not watching the same quantity of information. The paradox is that the verdict is identical. A penalty is a goal that may have been taken away. A red card is a player leaving the pitch. An offside is a goal wiped out. None of those outcomes has an undo function. The minimum input gate In data analysis there is a principle any serious system must obey: if the input set is insufficient to answer the question, the only correct output is insufficient data. A system forced to generate a conclusion from an empty set does not produce knowledge. It produces the illusion of knowledge, and that illusion is more dangerous than silence. Referees operate on the same logic, except nobody bothers to name it that way. FIFA's VAR protocol contains a mechanism called clear and obvious error. In essence it is a minimum input gate: only when the evidence set is thick enough to prove the referee was incontrovertibly wrong may the VAR team intervene. If the incident sits in the grey zone, the on-field decision stands. That is a sound design. The problem is that the grey zone has no clear border. And when pressure from the stands, from broadcasters and from the competition organisers themselves pours into a sealed room, the grey zone has a habit of shrinking until it disappears. The offside line and an error larger than the measurement Take the most common example: offside. A capture system running at 50 frames per second records a player sprinting at roughly 9 metres per second. Between two consecutive frames that player moves close to 18 centimetres. At the same moment, the defender's foot may not have touched the ground, the attacker is mid-stride, and both anchor points used to draw the offside line are geometric reconstructions produced by software. A conclusion is only trustworthy when the evidence set is thicker than the error of the measurement used to produce it. Here the reverse happens. The measurement error — roughly 18 centimetres from frame rate, plus the error in identifying the point of contact with the ball, plus the error in the line-drawing algorithm — is larger than the distance actually being declared. A line drawn to the millimetre does not create precision. It creates an image clean enough to convince viewers that precision exists. Based on my experience following matches, I have rebuilt the same offside incident from three different image sources: the main broadcast feed, a recording from the stand on the opposite side, and a clip circulating online. Three sources. Three lines. Three different outcomes, from one passage of play. None of those three sources was deliberately wrong. They were each answering a question that exceeded the capability of the data they held. What supporters feel in that moment is entirely understandable. They do not want to hear about frame rate error. They want an answer. But supplying an answer before the data supports one is the fastest way to destroy faith in the very system built to protect that faith. The paradox of too many frames The common assumption is that more cameras mean more fairness. That holds only when the system retains the capacity to say no. Every additional frame raises the probability of finding some detail that can be read as an error. A hand tucked against the body from angle A can look extended from angle B. A collision from angle C can become a legal challenge from angle D. When a VAR team has thirty frames and forty seconds, they are not facing a clearer truth. They are facing a mountain of contradictory evidence and must select the version that fits a decision. VAR does not uncover the truth; it only exposes what the referee chose to overlook. That is why the arguments did not shrink after VAR arrived. They simply moved from the referee on the pitch to the referee in the sealed room. The same level of grievance, a different address. When IFAB changed the definition instead of the procedure In 2026, IFAB tightened Law 12 to deal more harshly with handball incidents leading to a goal, including unintentional contact. On paper it was a major change. In operation it pushed referees into a harder place: they must judge intent, distance, ball speed and body shape while the law demands an objective standard. Law 12 does not explain the incident; it only assigns who carries the responsibility. Every shirt pull inside the penalty area leaves a mark in the ink of the match's verdict. And that ink cannot be erased. What needs to change I propose a procedural change, not an emotional one. Competitions should publish in advance the number of camera angles and the capture frame rate for each fixture. A match with four cameras should not be run as though it had thirty. That information belongs to the public, because it determines the ceiling of certainty any verdict can reach. A recognised error band for offside and handball should be defined. When a measurement falls inside that band, the on-field decision stands and the VAR team is permitted to declare insufficient data to overturn. This is not a concession. It is a correct conclusion. VAR teams must be protected from the pressure to find an error. A VAR team that stays silent all match is not a failed VAR team. An outstanding referee is only remembered after everyone has had to watch it again. But for people to replay it, the decision must rest on an evidence set thick enough to withstand that scrutiny. Otherwise what gets remembered is not the referee but a system failure. Football will never run out of arguments. The only thing that can change is which kind of argument we are forced to endure: an argument about decisions made from a sufficiently thick data set, or an argument about verdicts issued from an empty frame.