47 Decisions, One Mistake: Vietnamese Refereeing Enters the Data Revolution

Minute 73, Anfield, February 2026. Sadio Mane received a through ball in a clearly offside position. Referee Mike Dean stood less than fifteen metres from where the ball was taken, screened by a player in the exact instant the pass left the passer's foot, and did not blow his whistle. The ball went in. Liverpool equalised at 1-1. The match ended to prolonged jeers from the away end, and for the rest of that week, the whole of England talked about exactly one decision.

I kept forty-seven decisions.

That night I rewound the entire broadcast, and built a manual spreadsheet across twelve criteria: the referee's standing position, his viewing angle, distance to the point of contact, reaction time, and impact on the flow of play. Forty-seven times Mike Dean had to make a ruling in ninety minutes, covering corners, throw-ins, fouls, cards, advantage and offside.

He was right forty-six times. Wrong once. And that single error decided the result.

I tell this story neither to defend nor to accuse anyone. I tell it because it is the reason I abandoned the habit of writing commentary by feel. A referee with a 97.9 per cent accuracy rate in a match can still be called a saboteur, simply because his one mistake landed in the most expensive place on the pitch. This profession is judged on a distorted denominator, and we have accepted that denominator for far too long.

A profession with no scoreboard

Refereeing is the only job on the pitch whose output cannot be measured in goals. No goals, no assists, no clearances. The only thing on offer is consistency. And for nearly a century, consistency was measured by feel.

Professional football accepted that for a very long time. From the early 2000s, clubs moved into data analysis. Expected goals, PPDA, line-breaking passes became daily language in the coaching room. Referees, meanwhile, still worked with their eyes and their memory. A referee runs 11 to 12 kilometres per match and makes roughly 200 to 250 decisions, of which 40 to 50 directly shape the game. He has no scoreboard to check himself against.

FIFA wrote VAR into the Laws from 2026 under a slogan spelled out in the protocol: minimum interference, maximum benefit. It sounds tidy. But nobody has defined benefit in numbers. How many seconds is too long? How many errors is too many? What percentage of accuracy counts as success? No threshold has been published. And without a threshold, every argument about VAR drifts towards emotion, because emotion is the only thing left to hold on to.

The FIFA VAR protocol limits intervention to four categories: goals, penalty kicks, direct red cards, and mistaken identity. Those four do not cover every error that can occur on a pitch, and the gap between the scope of intervention and the scope of error is where most controversy is born. A foul leading to a goal in the 30th minute can be reviewed. An identical foul in midfield in the 12th minute cannot. That asymmetry is not the referee's fault. It is the design.

Vietnam entered this story later, through a fairly narrow door. VAR was trialled in V.League from 2026 with FIFA technical support, then gradually expanded across subsequent seasons. The Vietnam Professional Football Joint Stock Company set a clear objective: reduce errors in decisive situations. The objective is right. The measurement, however, remains open.

The gap is there. Not whether VAR is good or bad. But what we are measuring VAR with.

Three layers of error

Technical error, perceptual error and systemic error are three different things, and collapsing them together is the most common mistake in any debate about refereeing.

A technical error is a mistake about the Law. A player standing beyond the last defender when the ball is played is offside. There is nothing further to discuss.

A perceptual error is correct in Law but wrong in reading the situation. A referee plays an advantage too early, costing the fouled team a genuine counter-attack. In Law, the decision is not wrong. In game management, it is. This type of error rarely appears in reports, because no clause in the Laws specifies how many seconds late an advantage must be played.

A systemic error occurs when the Law or the guidance is so vague that two referees watch the same incident and reach two different decisions, both defensible. The phrase negligible contact in the fouls section of the Laws is a classic example. Nobody can quantify negligible in Newtons, and therefore nobody can say with certainty whether a collision is a foul by reading the Law alone.

Back to the forty-seven decisions at Anfield. Mike Dean's single error was technical. But the cause did not belong to the Law. He was screened by a player in the instant the ball was played. That is a geometry failure, not a knowledge failure. And geometry can be fixed with positioning drills, with angle training, with process. Knowledge needs the Law. Geometry needs process.

In V.League, when a referee misses a handball in the box, the debate almost immediately splits into two poles: the referee is poor, or the referee is compromised. Almost nobody asks the simplest question: was his angle blocked? Where was he standing relative to the line of the ball? How many metres was he from the point of contact? What was the orientation of his body as the ball was struck?

That is precisely the question VAR answers. But it is also a question a fourth official in front of a monitor can answer, even without VAR. The difference between the two is not technology. It is whether anyone is willing to spend the time measuring.

The numbers I measured myself

In June 2026, a broadcaster invited me to work as a VAR analysis specialist for matches at the World Cup in Russia. The France versus Australia match on 16 June was the first time I saw a penalty awarded after a referee reviewed a monitor mid-match. Antoine Griezmann's conversion from twelve yards opened a week of argument.

Every commentator in my room said the same sentence: VAR kills the rhythm of the game. I did not argue. I started a stopwatch.

I timed every review across the tournament, then cross-referenced fourteen other VAR decisions. Average review time: 101 seconds. And the total increase in added time per match, compared with the previous season: 2 minutes 37 seconds.

Two minutes thirty-seven seconds.

How much time does a team lose on average to a goal celebration? How much to a slow substitution in the 85th minute? How much to a throw-in where the ball is wiped with a towel? The time cost of VAR is smaller than the time cost of the delaying tactics that existed in football long before VAR was invented.

But the feeling is entirely different. Traditional delays are scattered into dozens of small fragments, a few seconds each, spread across the match. VAR concentrates into a single block, at the exact moment the crowd is most tense. The perception of time is not proportional to real time. It is proportional to the intensity of the waiting. A hundred seconds of silent waiting lasts longer than three hundred seconds passing amid noise.

That was when I moved from writing event descriptions to writing systems analysis. And it was also when I realised I had to measure things myself rather than cite other people's numbers. The most important data about referees is rarely in official reports. It is in the notebook of someone rewinding a broadcast at two in the morning.

Eighty-nine matches, empty stands

In June 2026, football returned after lockdown. I agreed to contribute analysis to an independent study on the effect of crowds on referee decisions. I collected data from eighty-nine Premier League matches, before and after the pandemic.

The result: yellow cards fell 23 per cent. Penalties rose 31 per cent.

The two numbers moved in opposite directions, and that opposition is the interesting part. If the only cause were referees feeling less crowd pressure, both indicators should have moved the same way. Yellow cards falling while penalties rise suggests at least two separate mechanisms at work.

Mechanism one: without crowd noise, referees stop using cards to manage the game. A yellow card in professional football is not only a punishment. It is a cooling tool, a dialogue tool, a way for a referee to tell twenty-two players that he is still in control. Fewer spectators, less pressure to control, fewer cards.

Mechanism two: without crowds, players stop acting. A fall in the box in front of forty thousand people triggers a collective reaction, and that collective reaction weighs on the referee. In an empty stadium, a fall is just a fall. The rise in penalties may not be because referees became braver, but because genuine fouls previously masked by crowd noise are now seen more clearly.

The data cannot fully separate the two mechanisms. But it is enough to say one thing: if removing crowds cuts yellow cards by 23 per cent, then the image of the objective referee never existed, only referees under pressure at different levels.

An empty stadium does not lose its soul; it simply returns the soul to its rightful owner.

What does this mean in Vietnam? V.League owns one of the most intense stadium atmospheres in Southeast Asia. Hang Day, Thien Truong, Lach Tray, Hoa Xuan, each with its own kind of pressure, none alike. That is an asset for Vietnamese football. But it is also an input variable for every decision on the pitch. Vietnamese referees are neither more nor less fair than English referees. They operate under a different pressure regime, and their training programmes contain almost no pressure calibration component.

From player micro-data to referee micro-data

In November 2026, I sat in the stands at a World Cup, watching England play Iran. The world focused on Bukayo Saka's hat-trick. I took notes on Jude Bellingham.

He was nineteen. In that match, Bellingham touched the ball 78 times. But the more important number lay elsewhere: 41 of those 78 touches were one-touch, and he almost never held the ball for more than three seconds. I called a former scout to check what I had seen. He confirmed it.

Macro indicators cannot express that. Passes completed, pass accuracy, chances created, all correct, all ignoring processing time. And in modern football, processing time is what separates a good midfielder from an outstanding one. I wrote a series on Bellingham after that tournament, predicting he would become the leading central midfielder of his generation, before most major newspapers mentioned it.

The same logic applies to referees. We measure referees with macro indicators: yellow cards per match, penalties per season, VAR interventions. Those indicators show outcomes, not processes.

Try micro indicators instead. The time from the foul occurring to the whistle sounding, what is the average and what is the standard deviation? The average distance from the referee to the point of contact when he makes a decisive call? The reversal rate when the referee is viewing from a diagonal angle versus a straight one?

These are all measurable variables, and more importantly, they are trainable ones. A referee cannot train to become fairer, but he can train to cut his reaction time below 0.8 seconds, or to hold the optimal distance to the point of contact on aerial balls.

Vietnam has a pipeline problem in referee development. The number of FIFA-standard referees is very small. The path from V.League 2 to V.League 1 to the FIFA list is narrow and slow. Under those conditions, measuring micro indicators allows a shift from mass training to targeted training: instead of a single fitness camp for thirty referees, an individual plan for five with FIFA potential, focused precisely on each one's measured weakness.

The opportunity cost

Every technology decision carries an opportunity cost, and Vietnamese football rarely discusses it. VAR needs equipment, operators, training, time. In a league where many clubs still struggle with stadium lighting standards and pitch quality, that money is not unlimited.

I am not saying drop VAR. I am saying ask the question: with the same quantity of resources, which option produces the greatest improvement per unit spent?

Suppose the same money were used to move a group of referees into full-time professional status, pay them enough that they need no second job, invest in a video review room, and teach English to FIFA candidates. Within three years, that group could move closer to international standard. Whereas VAR, if simply installed without an accompanying measurement process, will create a new layer of transparency but not a new layer of capability.

VAR does not replace referee training; it only makes the quality of that training easier to see. A well-trained referee will use VAR as a confirmation tool. A poorly trained referee will use VAR as a lifebuoy, and every grab at the lifebuoy is a moment when authority leaves the pitch.

Authority changes places

The deepest thing VAR changes is not accuracy. It is where authority sits.

Before VAR, the on-field referee had the final word in every situation. After VAR, across the four reviewable categories, the final word can belong to someone in a closed room several hundred metres away, someone the crowd cannot see and does not know by name.

The referee's power does not come from the whistle, but from the ability to read the situation. When reading the situation is relocated to another room, the whistle is still there but its meaning has changed. In Vietnam, where referees are appointed by the federation and where public trust in officiating has long been fragile, this relocation can be read two ways. The first: this is a safety net, a second layer of checking to reduce error. The second: this is a statement that the on-field referee cannot be trusted.

Both readings have grounds. Which one prevails depends almost entirely on whether the federation publishes data on interventions. If every VAR reversal is logged, categorised and published periodically, the shift of authority becomes a reform. If not, it becomes a new suspicion.

What I once got wrong

I was once a VAR sceptic, and that is why I understand those who hate it.

In 2026, when I began this analysis, I argued that VAR would make referees more accurate. After nearly a decade of data, I have to narrow that claim. VAR makes errors more visible. It does not make referees more accurate. Those are two different things, and I merged them into one.

This is not a retreat. It is a correction, and analysis as a trade lives on corrections like this.

The second thing I want to say plainly: the biggest problem in Vietnamese football's officiating is not VAR, and it is not the competence of referees. The problem is that we still treat refereeing as a moral question rather than a technical one.

When a decision is disputed, the debate immediately shifts to integrity. Whether there was impropriety. Whether there was outside influence. Those questions are not meaningless, but they cannot be answered with data, and therefore they never end. Meanwhile, the answerable questions, angle, distance, reaction time and standing position, are almost never asked.

You cannot train to become more honest. You can train to stand in the right place. Choose the wrong question and nothing improves, no matter how many cameras there are.

Cameras find the error, but people find the cause.

What I want to see next season

The best referee is the one nobody mentions after the match.

In Vietnam, we are not yet in a position to produce many of them, and we have no way of knowing whether we are moving closer or further away. That is what I want to see change.

Not more matches with VAR. But a periodic published table, a few times each season, with four lines: the average number of seconds per review, the percentage of decisions overturned, the share of errors caused by angle rather than knowledge of the Laws, and which referees are appointed to the matches with the highest decision density.

When those four lines are printed, the debate changes subject. No longer who was right or wrong on a specific incident, but whether the system is progressing or standing still. When data enters the dressing room, emotion has to leave through the window.

And Vietnamese refereeing will have something it has never had: a way to measure itself.