The Viewing-Angle Error: When Football Conclusions Are Built on Empty Data

In July 2026, in the 35th minute of the World Cup final between France and Croatia, referee Nestor Pitana stepped away from the monitor and pointed to the penalty spot. The ball had struck the hand of Ivan Perisic. Inside the VAR room, he was shown exactly one camera angle, lasting one minute and forty-seven seconds. I sat in front of six screens in Chengdu, re-measured every frame, and found what no one wanted to say on that broadcast: five of the six angles were insufficient to conclude that Perisic's arm was extended in an unnatural manner. Only one angle showed it, and that was the only angle the referee was allowed to see. I retell this story not to reopen an old decision. I retell it because it is the cleanest example of a disease that afflicts both the VAR room and the newsroom: building a very confident conclusion on a very thin foundation of data. We tend to think the problem lies in the conclusion. For me, the problem always lies in the input. The law we all argue about carries a grand name: clear and obvious error. It sounds simple. But I have spent nearly thirty years rereading the rulebooks, and I can tell you that no definition quantifies the phrase clear and obvious. How much deviation counts as clear? Fifteen centimetres? Twenty? No one can answer, because the answer changes with every match, every referee, every camera angle. VAR was introduced officially at the 2026 World Cup with a promise to reduce errors. It did reduce them. But it also moved the argument from the pitch into a closed room, where spectators, quite literally, are not allowed to watch along. At many tournaments, the number of cameras serving VAR is only a fraction of the number serving commercial broadcasts. That means the referee never sees everything the audience sees. The empty stadiums of 2026 showed me this: VAR did not save football, it exposed football. When the noise of the crowd no longer covers it, you hear every silent second of a referee waiting on a headset, and you realise that most of that time is spent not finding the truth but finding an angle convincing enough. In 2026, while working as a VAR data analyst for a sports television channel in Chengdu, I had the chance to do something few people bother to do: review everything. Across 28 rounds of the Chinese Super League, I dissected 147 controversial refereeing incidents. Not to count how often referees were right or wrong. I wanted to know something else: when a decision was wrong, why was it wrong. The result kept me awake for several nights. There were 12 incorrect offside decisions, and all 12 were directly linked to camera placement. Not the assistant referee's fault. Not the technology's fault. A fault of geometry. A camera placed off the horizontal plane of the last defender will create an illusion of distance. In round 25, in the match between Guangzhou Evergrande and Shanghai SIPG, a goal by Wu Lei was disallowed for a fifteen-centimetre deviation, yet no camera sat on the plane that needed measuring. That fifteen centimetres was not a fact. It was the output of a measurement taken under the wrong conditions. I found the error not in the centre of the pitch, but at the edge of the frame. From there, I began building my own database, which I called viewing-angle error. Every controversial incident was tagged with three parameters: the camera position relative to the decision plane, the estimated measurement error, and the time the referee actually spent watching the monitor. When I laid the 147 incidents on a single board, a pattern emerged as plain as daylight: the shorter the review time, the higher the probability of missing an important angle. No one can watch six angles in forty seconds. But the time pressure of television demands exactly that. It took 37 replays before I understood that the human eye is not a measuring stick. This is where I want to pause, because it connects straight to my craft as a writer. When an analyst sits down before a match with no data, he has two options. The first is to tell the truth: I do not know. The second is to build a story that sounds plausible out of thin air. In this industry, the second option is always rewarded. A data-free piece full of assertion will spread faster than an honest piece admitting the evidence is insufficient. It is the exact paradox of the VAR room: the more decisive the decision, the fewer people demand to know what it was based on. I have witnessed an analysis system return an empty result: no title, no source, no data points, no entities, no assessment of time sensitivity. Technically, it was a failure at the extraction stage. But professionally, it was a perfect warning: when the input is empty, every conclusion is fabrication, no matter how beautifully it is presented. I have never seen a football analysis with nothing to say that was nonetheless honest. Emptiness is always filled with tone. Look at the metrics we celebrate every day. Distance covered and sprint counts are packaged as measures of effort. A player covering twelve kilometres in a match sounds impressive. But ineffective running also produces pretty numbers. A midfielder who runs twelve kilometres while his team keeps losing the ball in the zone he just left is not measuring effort; he is measuring waste. The problem is not the data. The problem is that we assign meaning to data before checking whether it qualifies to carry that meaning. By the same logic, look at the metrics presented as truth. Expected goals, known as xG, measures the quality of shooting chances. The metric called passes allowed per defensive action, known as PPDA, measures pressing intensity, where a lower value means stronger pressing. Both are useful. But both depend on a human-made definition of what counts as a chance, what counts as a defensive action. When someone says a team controlled sixty percent of possession, the right question is not whether sixty percent is high or low, but where on the pitch that sixty percent was counted. Possession in your own half is not the same as possession in the opponent's half. Same number. Different meaning. Look at how injuries are announced. A player's return schedule is rarely decided by the doctor. It is controlled by the club's communications department. When you read that player X will return at the weekend, there is a high probability the injury has not healed, and that the message simply needs to keep the market calm, especially during the transfer window, when a player's value is tightly bound to his publicly stated fitness. This is not a conspiracy. It is information management. But it reminds me that in modern football, many facts are in reality press releases. So what distinguishes a trustworthy conclusion from a constructed one? Not the confidence of the speaker. It is the number of camera angles, in the metaphorical sense. A good judgement is one verified from several independent directions, accepting that each direction has its own error, and stating that error clearly. In the VAR room, we call it multi-angle verification. In writing, I call it honesty about the limits of what I know. Within the decision chain of a single incident, there are four links I always check in order: time, camera angle, the player's motion data, and the referee's cognitive limits. Time decides how much the referee is allowed to see. The angle decides what the referee sees. Motion data shows what the player did in the moment before. And cognitive limits remind me that even with sufficient data, the human eye still has blind spots. These four links never operate independently. They are a chain, and one weak link breaks the whole chain. There is a paradox I always carry with me: we think we are searching for justice, when in fact we are only searching for a prettier camera angle. That paradox explains why the same incident, reviewed ten times by two people, still yields two opposite conclusions. One looks at the arm. The other looks at the distance between ball and body. Both are seeing with their eyes. But seeing with your eyes was never evidence; it is only the starting point of an argument about who is looking in the right place. What I want to say to young people entering the analysis profession is this: the greatest temptation is not to say something wrong. The greatest temptation is to say too much from too little. A table with three rows can still generate a two-thousand-word article. But length is not proof of depth. When you find yourself writing enthusiastically about something you actually have no data on, stop and ask yourself: if I strip out every adjective, how much truth remains? In this transfer window, that question is worth even more. Every day brings hundreds of rumours, each packaged with a number. But which number has a source, and which is merely the writer's estimate? What motive does an agent have for leaking information? How long does the contract still run, and how is the release clause written? Those are the questions that can be verified. The rest, the sources close to, the reported to be, the closing in on, are blurred camera angles. They may be true, but they cannot be checked, and information that cannot be checked cannot serve as the foundation of a firm conclusion. I am not saying you must wait until you have enough data before writing. If you waited for that, no one would write anything. I am saying you must be transparent about where you stand on the evidence axis. A VAR decision based on one camera angle can be correct, but it must be noted as based on one camera angle. An analysis based on three rows of data can be useful, but it must state clearly that it has only three rows. That transparency does not weaken the argument. It makes the argument trustworthy, because it tells the reader exactly what they are trusting. Perhaps this is what I have learned most after nearly thirty years: the value of a conclusion lies not in whether it sounds decisive, but in whether the reader can walk back along the path that led to it and check every step. When a conclusion can be checked, it becomes trustworthy. When it can only be believed, it becomes dogma. And dogma is the most dangerous thing in both football and journalism, because it needs no evidence to survive, only enough people repeating it. At forty-five, I have softened noticeably compared with my younger self. I no longer believe it is possible to remove all error from football. What I believe is that error can be made visible, measurable, arguable. A good refereeing system is not one that never errs. It is one where, when it errs, anyone can see where and why. That is the difference between a mature football culture and one that only knows how to react. And I believe the future of both refereeing and football analysis lies there: not in producing fewer errors by declaring yourself less wrong, but in opening the process wide so that anyone can see the error margin. More cameras will not save us if we still choose to show only one angle. More data will not save us if we still use it to decorate a conclusion already written. Technology does not manufacture truth by itself. It only manufactures more ways to look, and more ways to look wrong. The question I leave behind, for both writers and readers: next time you see a conclusion too confident about a match, a transfer, or a refereeing decision, will you dare to ask how many camera angles it was built from? And if the answer is one, will you have the courage to call it by its true name: a guess presented as fact?