Five Columns, One Truth: What 116 Matches of Referee Logs Actually Say About VAR

Villa Park, 2026. No crowd in the ground. Aston Villa's Tyrone Mings was booked for dissent, and the pitch-side microphones caught every word. The match report the next morning recorded 'dissent' and moved on. The broadcast feed recorded something else — the words, the pitch of the voice, and the pause before the card came out. The press box heard a decision. The microphone heard a process.

That evening I opened my five-column sheet and changed the fifth column. It used to read 'correct decision' — meaning my opinion. After that it split in two: what the Law says, and what the microphone captured. The day that change happened, it became clear that football's disciplinary argument has been fought in the wrong place for eight years.

Eight years on, VAR's real controversy sits not on the pitch but on the threshold of the review room — and that threshold has never been entered in a public document.

This piece comes out of that log. Fifty-two matches at the 2026 FIFA U-17 World Cup in India, sixty-four at Russia 2026, forty-seven audio-logged incidents from the behind-closed-doors Premier League season in 2026, and sixty-three disciplinary incidents from the ISL 2026-21 bubble in Goa. One hundred and sixteen matches, forty-three penalties, twenty official reviews, one hundred and ten audio-tagged incidents. Every row carries a minute stamp. No gaps allowed.

The law first, the argument second

Before entering this argument, it is worth establishing what the statute actually says, because referee criticism in football is usually conducted without having read it.

Law 5 states that the referee's decisions are final. That sentence is the root of every dispute and also its terminus. Law 12 defines the disciplinary cards — caution, dismissal, and dissent. Law 11 is offside, whose language has been rewritten three times in six years. VAR has no existence outside these three laws.

The VAR protocol, written formally into the Laws of the Game by IFAB in March 2026 and first used at scale in Russia that summer, permits intervention in four categories: goal or no goal, penalty or no penalty, direct red card, and mistaken identity. No one reviews a second caution. That single line is the least-read passage in the entire debate.

The condition for intervention is two words: clear and obvious error. The Law does not ask whether the decision was wrong. It asks whether the error was clear enough to require correction. The first question is factual; the second is a threshold question. In my entire log, the argument was never about the first. It was always about the second.

To understand that, I had to go back to 2026. India hosted the U-17 World Cup that year, I was a 21-year-old journalism student studying in London, and I talked my way into a media pass and started a live referee audit across fifty-two matches. Every card, every penalty, every video review went into a spreadsheet. When I wrote that referees had awarded fourteen penalties in the group stage, a Kolkata daily quoted the tally. I spent the next month cross-checking each decision against the Laws of the Game.

What that month taught me is still the frame I report through: in football, the sense of a mistake arrives before the Law does, and is spoken before the Law is. The crowd sees contact; the Law sees intent, speed and position. The distance between the two is the actual story.

How the five columns work

My template is simple, and I built it in 2026 for a twelve-page Disciplinary Consistency Report.

| Column | What gets recorded | |---|---| | 1 | Minute, to the second | | 2 | Description of the incident | | 3 | On-field decision | | 4 | VAR outcome (intervention / none / silence) | | 5 | Correct decision under the Law |

In 2026 I watched all sixty-four Russia World Cup matches from a rented room in London and logged twenty-nine penalties and twenty official reviews. That dataset is what earned me a three-month junior analyst contract with a London sports data agency.

In 2026, two things were added. Project Restart opened the microphones in empty grounds, and I built a forty-seven-incident audio database. Then my editor handed me the ISL 2026-21 season in Goa, where I logged sixty-three disciplinary incidents from closed-doors feeds and standardised the tagging system. One hundred and ten incidents, three tags on each: minute, Law, outcome.

That standardisation was not a hobby. Without it, comparing two leagues is impossible. With it, the comparison became possible for the first time — and the result was not what broadcasters describe.

What the tape says

The first finding is in the numbers. Across the twenty official reviews at Russia 2026, logged minute by minute, the centre of the dispute was never whether a foul had occurred. It was always whether the on-field call was clear and obviously wrong. That is where VAR relocated the argument.

VAR did not reduce controversy. It changed the controversy's address — from the pitch to the review room, and from the decision to the definition of the rule.

The consequence is one no league admits publicly. While the argument was 'right or wrong', a highlight reel could settle it. Once the argument becomes 'was the threshold crossed', the highlight reel is useless, because a threshold is a phrase, not an image. In my log, columns four and five repeatedly diverged — meaning VAR arrived at a defensible outcome through a defensible process, and column five still did not agree. Put plainly for a reader who will never open the protocol: in a system where 'correct' and 'unappealable' collapse into the same word, review stops being possible.

The second finding comes from the one category VAR cannot touch. A second caution is not reviewable; a direct red card is. Which means a match can be decided by the one card that never enters a camera's field of view. Across the hundred-and-ten-incident database, this was the least auditable category by a distance. I call it the loophole, and it is not a technology failure — it is a design decision. In a statute with one door locked and a window left open, everyone leaves through the window.

The third finding came from the empty grounds.

The stadium was empty, but the audio log was full.

The Mings case is an example, not a rule. Across forty-seven incidents the pattern was that punishment was often decided by which microphone was closest, not by what was said. That is a procedural anomaly, and it never reaches print, because the newspaper writes 'dissent' while the Law writes 'using offensive, insulting or abusive language'. Between those two phrases sit hundreds of words that exist only on tape.

When I applied the same tagging system in Goa, the picture sharpened. The distribution of the sixty-three ISL incidents differed visibly from the forty-seven Premier League ones — the same Law, a different density. You can read that two ways. The first reading: Indian referees are softer. The second, better supported by the data: discipline is being used as a control instrument in both places, only in one league 'control' means keeping the game moving and in the other it means managing time.

Two different products are being manufactured from a single rulebook, and we are calling it consistency. Recall what consistency means: the same decision in the same situation. But if the threshold itself differs, there is no unit left to compare. What remains is calibration — and no one at IFAB publishes a calibration figure.

The 2026-21 revisions on handball and offside widened the problem rather than narrowing it. The language of 'unnatural silhouette' attempted a neutral threshold, but a player who has never jumped with a side-skip can now be judged unnatural. Semi-automated offside in 2026 sped up the decision; it did not change the underlying question of whether a leading shoulder is natural. Technology sharpened the image, not the ambiguity in the rule.

Three replays agree

Now the part where I have to argue with my own table.

The weakest column in my five is the fourth — 'VAR outcome'. It is weak because it does not register anything the public can feel. It registers twenty, forty, or seventy seconds of conversation inside the review room: what was found, on which camera, which Law applies, and whether that crosses the threshold.

Three replays agree. The fourth is where the story lives.

There is an uncomfortable consequence. Cricket chose a direct route — UltraEdge, ball tracking, the boundary line — and answers when the technology speaks. Football chose a linguistic route: 'clear and obvious'. No technology can tell you from any frame whether an error was clear, because clarity is a human judgement. That is why VAR stands eight years later in roughly the same place it began, only with better cameras.

And here I have to guard against a habit of my own profession. Sitting in London makes English refereeing feel like the neutral benchmark. I apply the same test to it. PGMOL issues public apologies, the Premier League's ownership tests and points deductions are argued in public, and every case runs into the same wall: the outcome is announced, the process is not. A federation's silence and a league's apology are both post-hoc. Neither contains the thing an audit needs — calibration data.

There is a further trap I catch myself in often: treating one leaked audio clip as conclusive. I have the Mings clip, and it is beautiful evidence, but if only that one out of forty-seven holds, I have to write that forty-six failed, not that the system proved itself. That is what five columns are for — so that no single column can tell the story alone.

The question returns: who holds the stopwatch

Auditing fifty-two matches at the 2026 U-17 World Cup, I thought the problem was the referees. After sixty-four matches in 2026, I thought it was the technology. After a hundred and ten closed-doors incidents in 2026, it was clear the problem belongs to neither a person nor a device. It belongs to publication.

A rule is only as strong as the person holding the stopwatch — and in football the stopwatch is still in a room with the door shut.

Three concrete steps build the way forward, and one league has to move first on each. Publish the review-room audio the same day the match ends, as some competitions have begun to do. Publish calibration data per competition: how many reviews, how many crossed the threshold, how many overturned the on-field call, how many should have. And take 'clear and obvious' out of the realm of language and make it measurable — a published error budget, outside which a referee is accountable and inside which a referee is protected.

Until those three exist, every big decision will be written the way it was written in 2026 — in two words, without the tape. And the next time that square box appears on screen, you will be asked what you think. What you think is not the question. The question is what was said in that room, and when the record becomes public.