The Referee's Eye: Three Layers of Sight on the Mat and the Trap of Video Review

THE REFEREE'S EYE: THREE LAYERS OF SIGHT ON THE MAT AND THE TRAP OF VIDEO REVIEW

October 2026, Jamsil Students' Gymnasium, Seoul. Round three of the light middleweight boxing final. Roy Jones Jr., nineteen years old, American, landing almost at will while Park Si-hun, the home fighter, mostly covered up. Post-fight statistics recorded eighty-six landed punches for Jones, thirty-two for Park. Five judges sat a few metres from the ring, each holding one slip of paper. When Park's arm was raised, the score was 3-2.

Fifteen thousand spectators were in Jamsil that night. None of them saw the thing that decided the fight. The thing that decided the fight lived on five handwritten cards, with no images attached and no written reasoning.

The judges involved were suspended afterwards. The investigation concluded that the scoring had gone wrong. The gold medal stayed in Seoul. No mechanism could reverse it, because in 2026 no procedure existed that allowed a camera angle to change a decision already issued.

I was not in Jamsil. I was born in Australia thirteen years later, and I had to read the entire file of that case in a library in Seoul when I was twenty. What I took from it was not outrage. It was a gap: if there had been cameras in the right four corners, would the story have been different?

The short answer is yes. But not in the way we assume.

THREE LAYERS OF EYES

On any mat, three layers of eyes deliver judgment at once. The first is the referee's eye: the person inside the ring, empty-handed, wearing an earpiece, deciding in a tenth of a second. The second is the judges' eye: the person outside the ring, with time, with a chair, with a card. The third is the camera's eye: lenses at multiple angles, slow motion, rewind, and — most importantly — no independent authority at all.

In football, that third layer was named in 2026 and its name is VAR. In combat sports it goes by other names: instant replay, video review, instant video review. It arrived later, more unevenly, and through very narrow doors.

The referee's eye has power but lacks time. The judges' eye has time but lacks raw data. The camera's eye has raw data but almost no independent authority. Those three layers do not naturally complement one another. They are connected by a list, and that list is written by human beings.

In professional boxing, the scoring system is the 10-point must: the winner of a round takes 10, the loser 9, occasionally 8 if a point is deducted or a fighter is knocked down more than once. Three judges sit on three different sides of the ring for a technical reason: each sees a different plane of the same event. The referee does not score the fight, but the referee deducts points, issues warnings, and can stop the contest. The referee does not write the score, but can bend it.

In MMA, the system is 10-9 must with three judges and a non-scoring referee. In South Korea, where I work, promotions such as Road FC and Black Combat operate almost entirely on the Unified Rules of the Association of Boxing Commissions and Combative Sports (ABC). Those rules were revised with effect from 1 January 2026, and that revision is one of the most consequential changes almost nobody noticed: the priority order among scoring criteria was inverted.

Before 2026, the three pillars of MMA scoring were effective striking, effective grappling, and cage control, ranked roughly equally and left to the judge's discretion. After 2026, the text is explicit: effective striking and effective grappling come first; cage control is a secondary criterion, counted only when the first two are even. The definition of a 10-8 round was tightened as well.

I read that revision four times in a week. What struck me was not the content but something else: a document that reorders the priority of criteria cannot be verified by the eye. It can only be verified by asking each judge whether he has read it.

In taekwondo, the camera layer was institutionalised far earlier and far deeper. The electronic body protector system was introduced to remove the human element from confirming points, and the instant video replay system became a permanent feature of competition rules. Under World Taekwondo competition rules, a coach is granted one replay request per bout. If the request succeeds, the request is retained. If it fails, it is gone.

One request. One bout. That is the entire authority granted to the third layer on behalf of a coach.

THE LIST OF REVIEWABLE THINGS IS THE REAL RULEBOOK

Here is the idea I want readers to carry away: in every combat sport, the document defining what may be reviewed is the most powerful document of all — more powerful than the document defining what counts as winning.

The reason is concrete. A scoring rulebook tells you what a winning round looks like. A review list tells you which category of error is permitted to be corrected. The second is stronger, because it determines which facts become recorded truth and which facts are left at the margins.

In boxing, one of the most common review situations has nothing to do with the score. It concerns the origin of a cut. If a cut came from a clash of heads, the fight may end one way. If it came from a punch, it ends another way. Video becomes a forensic tool, not a sporting one.

In MMA, the Nevada State Athletic Commission approved a rule permitting referees to use video replay from 2026. But that authority is limited: a referee may only review to establish certain specific situations, chiefly whether a foul occurred, whether an injury resulted from a foul, and whether the bout should continue. A referee may not review to correct a scorecard. Nobody may review to correct a scorecard, unless a special mechanism exists at commission level.

So the third layer of eyes is permitted to see everything, but permitted to act on a very short list.

I once spent four months counting. In 2026, when arenas closed because of the pandemic, I went back to the notebook I had started at seventeen — forty-seven pages of nothing but referee decisions reversed after review. I coded 1,247 officiating decisions from a football World Cup and three seasons of the Korean top flight, then extended the set to public data from combat sports events.

What I found was not the error rate. What I found was the correction rate. And the two numbers are very far apart.

THE PACQUIAO–BRADLEY PARADOX: REVIEWED, NOT REVERSED

In June 2026, Manny Pacquiao lost to Timothy Bradley by split decision in Las Vegas. Not long afterwards, the World Boxing Organization convened a panel of five independent judges to review the video. All five scored the fight for Pacquiao.

The result stood.

This is the point most fans misunderstand about the mechanism. The WBO can convene a review panel, but it cannot overturn a result, because that authority belongs to the state athletic commission where the bout took place — and that commission did not act. Five expert eyes looked at the same tape and reached a conclusion different from the one already entered into the record. The tape was right. The record was still right in its own way.

I retell this because it is the whole problem in miniature. We assume that when more evidence appears, the truth corrects itself. In combat sports, evidence and jurisdiction sit behind two different locks. One key does not open the other.

Data never commits a foul; the writer is the one who gets carded.

PARADOX TWO: SEEING MORE CLEARLY, ANGERING MORE DEEPLY

There is an implicit assumption in every debate about officiating technology: if spectators see more clearly, they will accept decisions more easily. That assumption is wrong, and wrong in a very specific direction.

Before video review, the crowd in the arena and the audience at home saw two different things. The audience at home saw the slow-motion angle, but that angle had no authority. The crowd saw the event with the naked eye, but had no angle. The two groups disagreed, and the disagreement was diffuse.

Once video review entered the rulebook, everyone saw the same thing. The same frame, the same slow motion, the same instant replayed twelve times on the big screen. The diffusion vanished. In its place came a more concentrated form of disagreement, and worse: that disagreement had nowhere left to doubt.

When everyone in the building sees a collision, and the final decision is still that nothing happened, the reaction is not doubt. The reaction is fury. Doubt can be soothed with explanation. Sight cannot.

I do not watch the strike; I watch the camera angle that watches the strike.

That is why I always check one thing before assessing a controversy: whether the frame the audience is permitted to see matches the frame the official is permitted to act on. Very often, the two frames diverge. And that gap is where the anger is manufactured.

THE SILENT SEASON: 2026 AND A VERY LARGE EAR

The 2026 season had no spectators, but it had a very large ear.

In combat sports, that is literally true. When the UFC ran events at the Apex in Las Vegas from the middle of 2026, and when ONE Championship staged closed-door events in Bangkok and Singapore, something appeared that had never before been data: clean audio.

Commentators spoke more quietly. Impacts were more audible. The breathing of fighters became a continuous stream of information. And most importantly, the referee's voice became a record that could be checked.

For years I could only guess what a referee said to fighters during a clinch break. From 2026, I could hear it. And what I heard changed how I read decisions.

A referee speaking in short commands is a referee in control of the rhythm. A referee speaking at length, explaining, is a referee who has lost the rhythm and is trying to recover it with words. A referee speaking to the judges outside the ring is a referee transferring responsibility. Those three modes correspond to three kinds of decisions, and they can only be heard when the crowd is absent.

I did not derive this from theory. I derived it from listening back to recordings for three straight weeks, logging every referee utterance against a timestamp, then matching them to final outcomes. It was one of the rare cases where a variable that had entirely disappeared — the crowd — turned out to be the source of a fuller dataset.

THE COMPENSATION EFFECT: THE 89 PERCENT AND WHAT IT DOES NOT SAY

During coding, I found a pattern I call the compensation effect. After one side suffers a wrong decision, officials tend to compensate that side with a favourable decision within the next two contests. In my dataset, the rate was 89 percent.

I wrote a thirty-page analysis of that effect, finished the charts over four months, and never sent it. The reason was simple and very professional: I was afraid I was missing a variable. I was afraid a pattern that elegant would turn out to be a by-product of some officials working in competitions with a higher density of contested decisions.

In combat sports, the compensation effect is even harder to prove, because the number of decisions per official per year is tiny. An MMA referee may work only twelve to twenty bouts a year. A sample of twelve data points proves nothing. That is the mathematical ceiling on every officiating argument in combat sports: the sample is too small to separate bias from randomness.

I say this because I have been on the other side. I once believed more data would settle everything.

THE BLIND SPOT OF THE MODEL: WHAT THE 2026 WORLD CUP TAUGHT ME TO FEAR ABOUT SPREADSHEETS

In 2026 I worked as a data assistant for a Korean news agency in Doha. I used my compensation model to predict that referees would limit cards in the group stage to protect the flow of the tournament — a hypothesis drawn directly from the fact that major competitions need to protect their television product.

The model was right in 26 of 36 matches. Then it collapsed entirely in a group-stage match between the Netherlands and Ecuador, when the referee issued eight yellow cards and awarded two penalties. My spreadsheet had no column for the variable I had missed: the social pressure when a regional host is eliminated early, and how that pressure travels down to the person holding the whistle.

After the tournament I spent three weeks interviewing two former FIFA referees. They showed me something I could not put into a table: body language. How an official stands after hearing the crowd. How that person touches the whistle before deciding. How that person looks toward the technical area.

I tell this story because it explains why I am cautious about every number in an article about officiating, including my own. In boxing, a referee may work twenty bouts a year and each bout has three judges. Those three judges are administratively independent of one another but not informationally independent: they watch one event, absorb one atmosphere, hear one crowd.

THE COUNTER-INTUITIVE TRAP: THE CAMERA DOES NOT REMOVE CONTROVERSY, IT RELOCATES IT

This is what I believe after re-reading the whole file from Jamsil 2026 to now.

Video review does not reduce controversy in combat sports. It moves controversy from the level of judgment to the level of procedure.

Before video, the argument was: what did the judge see? That is an argument about perception, and it can be soothed by saying that people viewing from different angles see different things. The answer is unsatisfying, but it is coherent.

With video, the argument shifts: why is the thing everyone can see not on the list of things that may be corrected? That question has no satisfying answer, because the real answer is: because that is how the rules were written.

The law is the only thing that never enters stoppage time.

In football this was once called the problem of the line. In combat sports it is the problem of the list. In both places the mechanism is identical: once you give technology the right to see, you do not give it the right to speak. You keep the right to speak. And the right to speak remains with human beings, carrying everything human beings carry — seniority, relationships, pressure from commissions, pressure from broadcast contracts, and the fear of being judged to have overturned a decision.

One notable experiment is rarely discussed: open scoring. Some boxing organisations have announced scorecards after round four and round eight in certain title fights, on the argument that judges would be less influenced by the final round and fighters would know where they stood. Major commissions such as Nevada and New York have still declined to adopt it widely. The reasoning against it is worth considering: if scores are announced mid-fight, a judge in round nine knows he needs a round to hold a particular score, and that may influence his own decision.

In other words, transparency can generate a new kind of bias. I do not take this as an argument against transparency. I take it as evidence that every procedural change creates a new blind spot, and that a good rule-maker is one who predicts the blind spot before it appears.

The referee reads the fight faster than anyone; I only write one beat slower.

And precisely because I write one beat slower, I see something the first layer of eyes cannot. In every major officiating controversy I have reconstructed from primary records, the deciding factor was almost never the collision. It was the authority structure built around that collision long before it happened.

At Jamsil in 2026, that structure was entirely closed: five cards, no images, no written reasoning, no appeals channel. At Rio 2026, the structure had more cameras but was still controlled by a single organisation — and the consequence was a string of judges suspended after contested boxing bouts. By 2026, the International Olympic Committee suspended the recognition of the international boxing federation and took over the running of the boxing tournament at Tokyo 2026. A sport had its operational control seized by the Olympic movement itself, and the stated reason was not technical. It was governance.

That is the clearest signal that the officiating problem in combat sports was never a problem of the eye. It is a problem of who is granted the authority to define that eye as correct.

WHAT SHOULD HAPPEN NEXT

I do not believe in giving cameras more power. I believe in opening the books.

Specifically, there are three things every combat sports organisation can do immediately, without changing its core rulebook.

First, publish round-by-round scorecards after the bout, indicating which criterion was applied in which round. This turns a decision from a number into a chain of reasoning that can be tested. A judge scoring a round 10-9 for effective striking is one data point. The same judge scoring 10-9 for cage control is another, and the difference between them is the thing worth arguing about.

Second, publish the review log: time of request, duration of review, number of rewinds, and reason for denial. Such a log turns an invisible act into a checkable record. I have spent years reading football refereeing reports, and what I learned is simple: a decision without a log cannot be debated, only criticised.

Third, separate the decision-maker from the person under commercial pressure. This is the hardest part and the least discussed. In many promotions, the referee works under contract to the very organisation selling tickets to the event they are officiating. I do not raise this to imply anything about anyone. I raise it because it is a structure, and structures produce behaviour.

If I could choose only one of the three, I would choose the second. Not because it matters most, but because it is the only one that can start this week, in any promotion, without a single conference approving it.

I have spent most of my writing life reading officiating decisions as a text rather than watching them as an event. That habit makes me slow. It makes me miss the moments audiences remember. But it gives me something else: I am never surprised by a controversy, because the controversy was never in the moment. It was in the line written long before.

Discipline is not punishment; discipline is a way of reading a fight.

And if there is one question I would leave with the people writing the rules for combat sports in Asia, it is this: if we have already agreed to pay for twelve more cameras around the ring, why have we still not agreed to spend ten more minutes explaining to the audience what those twelve cameras are permitted to see?