Semi-Automated Offside and the Blind Spots of VAR: When Three Stories Do Not Match

Minute twelve. A long ball over the top. A striker breaks clear. A yellow flag. Then a whistle. On the big screen, the crowd waits for the familiar offside graphic - a horizontal line, a vertical line, a red dot marking the heel. But this time, no graphic appears. Only a short notice: check complete. The goal is disallowed. No one explains further.

I sit in a small analysis room in Tokyo, in front of a screen showing six detached frames of the same play. On the pitch, the referee touches his earpiece. In the stands, sixty thousand people are screaming. In this room, no one screams. Data never panics. People blind themselves when they rush into emotion.

Every play tells three stories: one of the camera, one of the technology, one of history. My profession, across seventeen years since I left my seat as an economic analyst to step into the world of sport, is to sit down and decode where those three stories diverge and by how much. That night, they diverged by exactly 11.2 centimetres. A number smaller than the width of a smartphone. A number enough to cancel a goal, to change the fate of a group, and to open a debate that would last for months.

This is not the story of a referee who was wrong. It is the story of a system designed to eliminate human error, but which unintentionally created a new kind of error that no one prepared to face. And as always, people look for someone to judge, instead of looking for a mechanism to understand.

Context: From Human Eye to Machine Eye

To understand why 11.2 centimetres caused such a controversy, we must return to the starting point of VAR. In 2026, the International Football Association Board - the body that keeps the laws of the game - began testing video technology to assist referees in official matches. The spirit of the law at that time was clear: VAR would intervene only in four situations that could lead to clear mistakes - goals, penalties, direct red cards, and mistaken identity. The golden principle lay in the phrase high intervention threshold. That is, the video referee does not overturn decisions that are technically correct but emotionally contested.

By 2026, at the World Cup in Russia, VAR appeared on the biggest stage for the first time. That was where I worked as an analyst for an Asian broadcaster. In the France-Australia match, I discovered a calibration error in the semi-automated offside system that caused a penalty to be awarded based on desynchronised data. I wrote a five-thousand-word analysis pointing out the algorithm error. The newsroom refused to publish it, citing that it was too technical. Instead of softening my tone, I dived into twenty-five other matches and found four similar errors.

That was when I understood something that later became the foundation of how I view every VAR controversy: the problem rarely lies in the final decision. The problem lies in the gap between what the technology sees and what the technology communicates to the audience. A system that measures an error of 11.2 centimetres but shows no one which starting point it calculated from, what its sampling frequency was, and what the latency of each frame was, is a system that starves public trust.

In 2026, the English Premier League brought VAR into full operation. Immediately a series of controversies appeared over offside lines drawn by hand on low-frame-rate images. Fans began calling it football of the toes. Strikers such as Raheem Sterling repeatedly saw goals disallowed by a millimetre. Matchday attendance dipped slightly in some rounds. Organisers realised that a technology correct in mathematical terms can still be wrong in experiential terms.

Core: Three Layers of Analysis Within a Single Line

To properly decode a semi-automated offside, it must be split into three layers. The first layer is physics. A player's position at the moment the ball leaves the foot is a point in three-dimensional space, determined by dozens of optical cameras around the pitch sampling at high speed. These cameras do not take photographs; they record points. From that point data, the system reconstructs movement and identifies the valid body part of each player nearest to the opponent's goal.

The second layer is timing. This is the most error-prone and least discussed area. The system must determine exactly which frame corresponds to the ball leaving the passer's foot. If the sensor inside the ball is working, the moment is fixed by a signal pulse. If the sensor fails or a backup ball is used, the moment is inferred by the human eye or by an interpolation algorithm. A half-frame error at thirty frames per second, when a striker sprints at nine metres per second, equals fifteen centimetres of displacement. Multiplied by the expansion factor of movement, it can cross the decision threshold.

The third layer is the interface. This is the layer the public sees, and also the most undervalued. A line drawn on a television frame, at broadcast resolution, is not the resolution of the source data. That line is an illustration, not a measurement. But the audience is not taught to distinguish the two. They see a decisive line, and they believe that behind it lies a decisive truth. When the truth is not as decisive as the line, belief turns into outrage.

I spent three hundred hours reviewing footage of a single season just to understand a team's zonal defensive mechanism, and I learned a lesson applicable to VAR as well: every system has a structural blind spot. It is not a temporary error. It is a limit built into the design. Semi-automated offside has its blind spot where it handles clear plays well but leaves a grey zone for borderline situations. And that grey zone is where football currently plays out most of its controversies.

In 2026, when the entire global fixture calendar was suspended and I lost my broadcasting contract, I spent eight months building a database of one thousand two hundred matches without spectators from twenty countries. The data showed something unexpected: the success rate of penalties rose by roughly fourteen percent when the stadium was silent. When crowd noise disappears, psychological pressure shifts from the feeling of being watched to the feeling of being measured. Players no longer fear the crowd, but they begin to fear the number. The same mechanism applies to referees. When every decision is recorded and measured, referees no longer judge by instinct, but judge by fear of being overruled by the system.

Contrarian Angle: Emotion Is Not the Enemy of Law

There is a popular view in analytical circles, and I once leaned toward it: emotion is the enemy of consistency. Fans want football to flow, while VAR wants football to be accurate. These two desires are in conflict, and people often conclude that fans need to be educated to accept accuracy.

But looking closely at the data, I see the opposite. Fans do not oppose accuracy. They oppose unexplained accuracy. The outrage in millimetre offsides does not stem from the goal being disallowed. It stems from the feeling that a machine decided in place of a human, and that machine refuses accountability. In a game organised by humans, for humans to watch, the feeling of losing control to a black box is socially unacceptable, no matter how correct the black box is.

This brings me to another rarely mentioned blind spot. VAR was designed to correct clear mistakes, but it unintentionally created a new kind of mistake: the mistake of being right and still causing controversy. A referee correctly determines a correct decision, and then the whole stadium boos. In legal terms, he cannot be judged. Emotionally, he has lost. And a system that makes the right person suffer like the wrong one is an incomplete system, even if its results are accurate.

I once argued for five hours with an editor on exactly this topic. He wanted me to write that VAR needs improvement. I refused to write the conclusion, because I believed the data spoke for itself. Colleagues began avoiding working with me. It took me years to understand that the silence of data only convinces those already ready to read data. For everyone else, a system needs a voice. And that voice must be human, not algorithmic.

There is a memorable historical fact here. Before technology, offside controversies also existed, but they had a different quality: they ended. A wrong decision was acknowledged, an apology was issued, and the story closed. Technology has changed that. Now a single play can be dissected infinitely, because there is always another camera angle, another frame, another interpretation. The permanence of data has turned every controversy into a wound that never fully heals.

Tactical Blind Spot and the Trap of Perfection

What troubles me most is not the technology, but how technology reshapes behaviour on the pitch. When a system can detect a hand touching the ball at fifteen centimetres, defenders change how they defend. They keep their arms tighter, drop their positions lower, and accept giving strikers space rather than risk an accidental touch. The consequence is that teams play deeper, goals from set pieces rise, and improvised plays inside the box become rarer.

In other words, the pursuit of absolute accuracy has quietly changed the nature of the sport. And this does not appear in any data report, because data measures what happens, not what was prevented from happening. I tracked a domestic league for three seasons and noticed the boundary of what is called rational risk behaviour shifting markedly. Players learned that risk is punished rather than assessed. And in a sport where creativity comes from daring to be wrong, teaching players that being wrong is untouchable is a cultural loss.

This is where my stance on professionalisation becomes clear. When I built a prediction model for eighteen teams in one season and the model correctly predicted the champion with eighty-two points, I did not feel as happy as I expected. I felt afraid. I felt afraid because a system had captured a team to the point of predicting their exact point total. But football is not an equation with a solution. It is a continuous process of choices that cannot be fully predicted. When every choice can be modelled, the most beautiful part of the sport - the surprise - becomes a computable variable.

Takeaway: A Thought Moving Forward

My eyes return to the screen with six detached frames. Three stories are unfolding in parallel, and they will never fully align. That is not a failure of technology. It is the nature of humans trying to measure a game created by humans.

If I could choose a single change for the future of VAR, I would not demand more cameras, nor faster algorithms. I would demand something far smaller: a mandatory disclosure protocol. Every high-threshold decision must be explained publicly in human language, accompanied by a confidence index. Not to reassure fans, but to give them back the right to understand. A system transparent about its own limits will be more trustworthy than one claiming perfection.

Ultimately, I believe football does not need an absolutely correct VAR. It needs an understandable, verifiable, and forgivable VAR. Because what makes this sport alive was never that no one errs. What makes it alive is that after every mistake, the whole world still sits down to watch the next match.

The remaining question is not whether technology is right or wrong. The question is: when technology is right but trust is wrong, do we dare to judge trust, or will we keep judging those who try to explain the truth?