When There Is No Camera Angle: VAR, the Laws, and the Trap of the Premature Verdict

When There Is No Camera Angle: VAR, the Laws, and the Trap of the Premature Verdict

I opened the analysis file at 2:40 in the morning, after the second half of a knockout tie had ended. The file had seven fields. The first held the match name: blank. The second held the source: blank. The third listed the information points: an empty list, not one line. The fourth recorded the entities involved: unidentified. The last three fields, time sensitivity, source quality, confidence level, all sat at not assessed.

Seven empty boxes. Nothing to hold on to.

I sat there for another twenty minutes staring at that skeleton. In my trade, an empty file is usually treated as a technical fault, a transmission error, something to delete and redo. The longer I looked, the more it resembled a VAR shift I once worked: the monitor had signal, the screen was lit, but inside there was no camera angle for the incident that needed reviewing. And the referee still had to decide. The whistle does not wait for data.

Emptiness in football is never neutral. It is always filled with something, with instinct, memory, crowd pressure, the referee's gut, or worse, with the verdict of someone who has no authority to deliver one.

Football built a system that assumes data always exists

Over two decades the sport moved from a game governed by the human eye to a system run on evidence. On 5 July 2026 the International Football Association Board approved goal-line technology, and by the 2026 World Cup in Brazil it was mandatory at every finals match. Four years later the video assistant referee entered the Laws of the Game and debuted at a major tournament at the 2026 World Cup in Russia. At the 2026 World Cup in Qatar, semi-automated offside technology tracked 29 data points per player with 12 cameras, alongside a motion sensor in the ball sampling 500 times per second.

Read that timeline and an unspoken assumption appears: that data will always be there. That there will always be an angle. That a camera will never be blocked, a feed will never drop, an assistant referee will never have a sightline obscured, a frame rate will never be too thin to separate the moment the ball leaves the foot from the moment a shoulder passes the last defender.

No article of the Laws says what to do when there is no data. The four reviewable categories, goal or no goal, penalty or no penalty, direct red card, mistaken identity, are all written around a clear and obvious error standard. But clear and obvious compared to what, when all you hold is one blurred frame and three seconds of silence in your earpiece?

I have sat in a referee's chair five times across different competitions, and each time I learned something about the limits of proof. The hardest part of the job is not reading the law. It is recognising when you do not yet have the grounds to read it.

Three kinds of failure that look identical on screen

When an analysis file comes back empty, the cause is usually one of three: transmission failure, extraction failure, or interpretation failure. Football commits all three, and they produce the same visual result, a blank space.

Transmission failure is a signal that never arrives: a dead camera, a severed cable, a touchline lens blocked by a crowd at the decisive instant. At club level it is an analysis department that never logged the match because nobody was assigned. At federation level it is a league that does not publish the audio between referee and VAR.

Extraction failure is a signal that arrives but is deformed in transit, and it is the most dangerous because it never announces itself. A corrupted ingest, a copied column in the wrong order, a truncated recording. Each produces a file that looks complete while being hollow. A hollow file that looks full is far worse than an obviously empty one, because it deceives the reader.

Interpretation failure is when the data is complete but the reader takes a shortcut. It is the most common failure in sports media, and the one I have committed most often.

None of the three ever names itself. Nobody in a VAR room announces an extraction failure. Nobody on air says they are interpreting incomplete data. People simply state conclusions. And a conclusion, once released, outlives the truth by a wide margin.

My mistake on live television became the foundation of a new system.

Law 11 and the frame problem the eye cannot solve

Start with Law 11, offside, where the data gap is clearest.

A player is offside if any part of the head, body or feet is nearer to the opponents' goal line than both the ball and the second-last opponent, at the moment the ball is played by a teammate. Arms and hands are excluded, including the goalkeeper's, since no part of the body may score with the hand.

The clause runs under a hundred words. Applying it requires pinning down a single instant: the moment the ball leaves the passer's foot.

At 50 frames per second, a broadcast standard for decades, the gap between consecutive frames is 20 milliseconds. A sprinting defender covers 15 to 20 centimetres in that time. Between two adjacent frames, a defender can cross the entire width of a foot.

That is why boundary offside calls always carried a degree of irreducible uncertainty. The referee was not wrong. The referee was reading a document with pages missing.

Semi-automated offside was built to fill exactly that gap. It does not merely draw straighter lines. It changes the question: instead of asking which frame shows the ball leaving the foot, the ball sensor fixes that instant, and the system projects every relevant player onto the same timestamp.

I once spent a month cross-checking VAR data from the first twelve matches of a major finals against the original law text, logging every decision into a spreadsheet of more than two thousand rows. It taught me something no lecture could: most offside controversies are not about the law. They are about the frame. People argue about a decision when the real argument is the quality of the evidence behind it.

Based on my experience watching matches, an offside call made from 50 fps footage and one made from a semi-automated system are different kinds of decisions, even though both land in the same column of the match report. The first is a judgement. The second is a measurement. Filing them under one statistical heading is a methodological error, and any league still doing it is feeding viewers an illusion of precision.

Law 12, handball, and a formula written in place of data

Where Law 11 leaves a gap in the frame, Law 12 leaves one in intent.

Handball is an offence when a player deliberately touches the ball with the hand or arm, when the arm makes the body unnaturally bigger, or when the hand or arm is above shoulder level in that situation. Separately, scoring or creating a goal with the hand or arm is always an offence, even unintentionally.

Two structural problems sit inside those sentences. The first requires determining intent, a state inside a player's head that no camera records. The second requires measuring a geometric quantity, the arm making the body unnaturally bigger, without offering any threshold.

Unnatural by how many degrees? Relative to which posture? Which player? At what running speed?

Lawmakers must choose between two things that cannot coexist: a clause flexible enough for the endless variety of handball incidents, and one specific enough that referees in two different leagues reach the same verdict. Choose flexibility and you get consistency of spirit but not of application. Choose specificity and you get a disallowed goal for a 47-degree silhouette instead of 45.

I paid for reading that law too quickly. In 2026, during a World Cup opening match on national television, I explained a handball incident and closed with a flat assertion that the touch was deliberate. Legally that was a misreading, because I skipped the whole analysis of distance, ball direction and arm posture before concluding on intent.

Social media reacted within minutes. I did not argue back. I downloaded the VAR data from the first twelve matches, built a spreadsheet of over two thousand rows, and spent a full month checking every decision against the source law. I abandoned declarative writing. I moved to a branching structure: if clause X applies, the conclusion is A; if clause Y applies, the conclusion is B. And I always state which branch I am using. That structure is not hesitation. It is the only way an analyst stays honest when the data cannot settle on one answer.

An empty file, forty-seven replays, and the value of counting

In 2026 I was a senior specialist for a new sports platform in Shenzhen. A colleague asked me to analyse a disallowed goal in an AFC Champions League knockout tie. The referee flagged offside. Nothing to discuss, was my first reaction.

I watched the footage forty-seven times. I measured the defender's running angle frame by frame against the moment the ball was played, and found the offside clause had been misapplied at the step of fixing the instant.

The piece took three weeks, ran over six thousand words, and carried fourteen annotated frames for each step of the reasoning. It was shared more than two hundred thousand times.

The lesson was not in the outcome. It was that when a file looks complete, people stop checking. The feeling of completeness is what blocks the truth.

Forty-seven replays is not obsession. It is a procedure, and procedures repeat, while instinct does not.

An empty stadium is the referee's finest laboratory.

I have verified this across years of watching matches without crowds. With no roar pressing down from four sides, referees tend to call what they actually saw rather than what the crowd is shouting that they saw. In that environment, contested decisions fall noticeably in the boundary-error group, the category shaped by social pressure, while the error group tied to legal interpretation, a purely technical category, stays broadly flat.

That is worth thinking about. A meaningful share of the refereeing controversies we see on television is not a problem of law. It is a problem of the stands.

The transfer market and the same problem, priced in money

A transfer rumour and a VAR incident are the same kind of document. Both are claims about an event not yet fully verified, issued by a party with an interest in the outcome. And both are received the same way: audiences react to the conclusion, not the evidence.

My reliability grading for transfer news is simple and identical to how I grade visual evidence.

The first tier is documented confirmation from a club or federation. Near-absolute, and also the dullest, which is why it rarely spreads.

The second tier is reporting from journalists with a long, checkable track record. Highly credible, but even the best reporter is describing one stage of a negotiation, not the final outcome. A story true in June can become false in July without anyone lying.

The third tier comes from agents or clubs with a clear motive in the leak. It should be read as a negotiating move, not as a news item.

The fourth tier aggregates the third and presents itself as the second.

Based on my experience following transfer windows, most of the chaos does not come from missing information but from every tier being delivered in the same tone. Once the fourth tier learns to write like the second, the whole grading system collapses and fans fall back on feeling.

One further point is underrated. In a deal's cost structure, the hardest part to police is not the transfer fee but the signing-on fee and agent payments. Those sit on a different line and therefore at the edge of core financial monitoring.

A free agent generates no transfer fee. He still receives a signing-on fee, and the new club still pays above market wage for the position. The real cost is no smaller. It is simply recorded elsewhere. This is why modern financial monitoring polices what is visible well and what is invisible poorly, and the invisible is always where the money flows most.

Corners, dead data and a moral test

A corner kick is a moral test for the creator.

I have said this for years, and it becomes truer each season. Every corner generates a vast data trail: player positions, ball trajectory, jump timing, run direction, distance to the nearest opponent, outcome. All of it is logged. Most of it is never used to improve a team. It is used to sell: to persuade an employer, an agent, a board that this coach has a better corner system than that one, on a sample of seventeen corners.

Seventeen corners is not a sample. It is a lottery dressed in spreadsheets. Anyone who has worked with sports data knows a threshold must be reached before any conclusion can be drawn. That threshold is hard to sell. Nobody pays for a report stating there is not enough data to conclude.

This is where I regularly disagree with football's data analysts. I do not object to data. I object to data being used to manufacture an appearance of certainty for conclusions that have no basis. I have sat in rooms where a tactical model was presented on six matches, and beside a coach explaining that a defeat came from metrics not being executed properly, when what I saw on the tape was three individual errors in three separate incidents.

The analyst is not lying. They are reading a file with more empty boxes than they admit.

How far one missing data point travels

Trace the path of a missing data point through football, because it travels further than anyone assumes.

In the academy, a youth coach notes a player too young to compete. Nobody enters that record into the central system.

At the club, the analysis department has no data on him, so he never appears in internal scouting reports.

In the transfer market, a sporting director at another club runs a filter and never sees the name.

In media, fans never hear of him, so no demand forms.

Commercially, a sponsor looking for a young face for the next campaign skips him.

In market data, his estimated value stays below his true value for an entire career, because the valuation system can only read what was recorded.

One blank box in an academy becomes, twenty years later, a pricing error no one can correct. This is the most expensive failure mode in modern football, and it is almost never discussed, because it produces no dramatic moment. No disallowed goal. No whistle. No one to blame. Only a file that was never written.

The contrarian case: the pressure to produce a verdict

Here I want to be blunt about what I consider the industry's biggest blind spot.

We criticise those who reach wrong conclusions. We almost never criticise those who reached a conclusion when the honest answer was insufficient evidence.

Football runs on a quiet, irresistible pressure to have an opinion. A pundit who says on air that there is not enough data to conclude is treated as unqualified. A club publishing a twenty-page report of nothing but insufficient data is treated as unprofessional. An article ending in an unanswerable question is treated as worthless.

So people fill the gap. With tone. With confidence. With a barrage of numbers from a small sample. With a formula that looks scientific. With calling a momentary judgement a principle.

I have done it. I filled a gap with certainty, live on air, in front of millions. And I learned that the cost of filling a gap is not lost credibility. The cost is lost information. The moment I said the handball was deliberate, I closed every other possibility. I was no longer analysing. I was only asserting.

This is the paradox of every analysis screen: the more certain the voice, the more the audience believes, and the less information is actually conveyed.

Refereeing trends and what must change

Watching refereeing evolve, I believe the biggest shift of the coming decade will not be new technology. It will be procedural transparency.

Some leagues now publish the audio between referee and VAR after each matchday. This is the most valuable change in decades, because it turns a black box into an auditable file. When everyone can hear the reasoning behind a decision, the argument shifts from whether the referee was right to whether the law is sensible. That is progress.

My proposal has three parts.

First, every VAR decision should carry an incident reference code so anyone can trace it back and check it against the relevant law text. No code, no oversight.

Second, leagues must record explicitly in the match report when no camera angle was sufficient to conclude. This matters most. When a decision is made without adequate evidence, the report should say so, rather than logging it as the output of a complete process. A recorded gap is a useful gap. A concealed gap will return.

Third, data publishers must be accountable for the integrity of what they publish. When a dataset is empty, the publisher must say it is empty, not leave the reader to discover it.

None of this requires new technology. It requires something far harder: the willingness to admit you do not yet know.

Closing

I still keep that file with the seven empty boxes, printed and folded into a notebook. When students ask me how to analyse a match, I show them that sheet before any dataset.

Seven empty boxes teach what thousands of pages of data cannot: a system is only as trustworthy as its ability to report its own failures. A referee is only as trustworthy as his ability to say he did not see. An analyst is only as trustworthy as his ability to stop before concluding.

In the coming years, as semi-automated offside becomes standard in more leagues, we will get more accurate verdicts. We will not get fewer arguments, because arguments do not come from error margins. They come from the feeling that the process behind a decision is something we are not allowed to see.

And when that process exposes its empty boxes, as it always will, the question for everyone sitting in front of a screen, referee, commentator or fan, is down to one: will you say you do not know, or will you fill the gap with an assertion?