When the Data Goes Silent: The Discipline of a Null Result

The third monitor in the KBS analysis room turned grey in the 78th minute of the semifinal. The signal from the VAR booth cut out. No picture, no audio, no camera angle. In the room, everyone turned to me, the person assigned to run the disciplinary data model for the 2026 World Cup. The programme director asked, urgently: "Give me a conclusion, any conclusion, we go on air in four minutes."

I looked at the empty data sheet and answered: "No data, no verdict."

Four minutes later, the programme went on air with someone else's commentary. That night I was seen as rigid, as the man who ruined the broadcast rhythm. But I kept my decision. A disciplinary data model, when its input is empty, can only produce one honest output: a null result, a verdict that says there is not enough evidence to deliver a verdict. In my profession, staying silent before a data gap is not cowardice. It is discipline.

Six years later, I still tell this story whenever someone asks why a league disciplinary reporter spends so much time talking about what he does not know, instead of what he knows.

Context: An industry that does not allow silence

In 2026, Korean sports media entered a boom phase. Newsrooms raced to open data sections, hire analysts, and buy international statistical feeds. Inside that machinery, a league disciplinary reporter like me occupied a strange position. I did not sit in the stands, I did not stand on the touchline, and I did not sit in the press room. I sat in a small room with three monitors, reading match reports, breaking down every foul, and trying to answer one question: why did the referee show the card at that exact moment.

This job taught me something no school ever taught: most hot conclusions are written before the data has finished loading. Newsrooms need a headline within thirty minutes of the final whistle. Broadcasts need one clean declarative sentence. Nobody pays for a long-winded answer like "we do not have enough data to conclude". And precisely because of that, football analysis easily falls into a trap: producing conclusions that sound plausible but have no basis, then repeating them until the public believes they are true.

To understand a league, read the disciplinary record instead of the table. The table tells you which team won. The disciplinary record tells you why the match went that way. A team that keeps picking up cards from the 70th minute onward is often not playing dirty; it is running out of legs and forced into tactical fouls to cut off counter-attacks. Another team that collects yellow cards evenly on both flanks is usually playing a system that requires its wide midfielders to duel constantly. Those traces only appear when you sit down with raw data, not with summaries already edited for brevity.

The problem is this: if the raw data has not arrived, you must choose between two paths. One is to invent a conclusion that sounds expert. The other is to say plainly that you have nothing to say. This industry, for many years, chose the first path far too often.

The substrate of every verdict: information points

In my process, before any analysis is written, there must be something called information points. These are raw, uncontestable units of fact: when a match took place, which referee officiated, how many fouls there were, how many cards, in which minute. Every analysis of tactics, finance, form or management must grow from that substrate. No foundation, no building.

A null result is an honest conclusion: when the input contains no information points, every conclusion is fabrication, however beautifully it is presented. This is the principle I placed on my desk in 2026 and have never once taken down.

The irony is that this seemingly dry principle is what kept my model alive across multiple seasons. When you force a model to deliver a verdict while the data is insufficient, you do not make it stronger. You only make it more confident, and misplaced confidence is the most dangerous disease of the analytical profession.

2026: 1,847 fouls and a referee with his own taste

In 2026, I began building an analytical model from 1,847 fouls across 228 matches in K League 1. It was a vast block of raw data, and I had to standardise every single incident by hand, because at the time there was no standard data source for this league. Every night I sat breaking down match reports, recording who fouled, in which area, in which minute, and how the referee handled it.

The result startled me. Referee Kim Jong-hyeok showed cards to wide midfielders 2.4 times more often than the league average. This was a suspicious number, because in theory a good referee should treat every position on the pitch evenly. But data does not care about theory. It revealed a clear behavioural pattern: challenges in wide areas, where wingers constantly have to use pace to beat their man, were judged far more strictly by this referee than similar challenges in central areas.

I could have chosen to write a piece criticising the referee. But that would have been an emotional conclusion dressed in data clothing. I chose to follow the data to its end: whether that behavioural pattern could predict card decisions in the second half of the season. My model correctly predicted 73.6% of card decisions in the second half of the season. That number forced the editorial board to give me a dedicated column, instead of assigning me ordinary match reports.

But the bigger lesson lay elsewhere. My system does not expose the mistakes of players; it exposes the choreography of injustice. When you count enough incidents, you begin to see patterns the naked eye misses. One referee is not biased toward any team, but is biased toward a certain type of collision. Another referee is more sensitive to a player's reaction than to the danger of the challenge. Those things are not in the laws, but they shape matches more than any tactical diagram.

Every red card is a sentence written many phases earlier. The card the crowd sees in the 85th minute was in fact "signed" in the 20th minute, when that player committed his first foul and was entered into the referee's notebook. That is why I always tell young editors: do not rewatch the phase that led to the red card. Rewatch the whole match that led to it.

2026: The World Cup, VAR, and a number 3.2 times higher

In 2026, my model was used by KBS as the foundation for its VAR analysis during the World Cup in Russia. I reviewed all 64 matches and logged every VAR intervention. This was the first World Cup to apply video assistant referee technology across all matches, so every decision was a valuable piece of data.

The most notable finding: VAR usage rose 3.2 times in the semifinals compared with the group stage, concentrated on handball situations inside the penalty area. At first glance this seems reasonable; big matches have more decisive moments. But when I split the data by incident type, the picture became more complex. VAR interventions for ordinary challenges barely changed between rounds. Only the handball category spiked.

That says something about psychology: the more important the match, the lower the sensitivity threshold of both the referee and the VAR team toward handball situations. Nobody orders them to do this. But the pressure of a World Cup semifinal creates a kind of tension that makes people look harder at details that can change the course of the game. Data only records the consequence; the psychology behind the number is what I want to understand.

In 2026, I learned to trust the model before trusting emotion. Before that, whenever I watched a controversial phase, my first reflex was to feel. Afterwards, my first reflex was to open the data sheet and check whether this situation belonged to the category the model predicted would be prone to controversy. That approach did not make me cold. It simply gave my emotions something to stand on.

My detailed analysis was widely shared in Asian referee research circles and opened access to official data from the AFC. Since then, I have applied a standardised VAR analysis framework to every article: cross-checking three data sources before publication. If those three sources do not match, it does not mean I pick the one that seems more plausible. It means I do not yet have enough grounds to write.

2026: Empty stadiums and the referee's tolerance threshold

In 2026, the pandemic forced the K League to play in empty stadiums. This was a rare research opportunity, because for the first time in modern history we could observe professional football without the noise of a crowd. I used the data access I had built since 2026 and analysed 171 matches.

The result: yellow cards fell 18.5% compared with the 2026 season. This was one of the findings that forced me to rewrite many of my own assumptions. My initial hypothesis was that players would play more aggressively without a crowd, since they would not be reminded by the stands. The data showed the opposite, or at least a different story: it was the referees who changed their behaviour. Without the jeering of the stands, referees' tolerance for fouls rose. They showed fewer cards, because they were no longer pushed by the crowd to react harshly.

The stadium was empty, but discipline still sat in the stands. I used that line in the analysis that year to say that even when the stands have no people, the unspoken standards of the refereeing profession are still present. But they are present in a different, gentler way, and that was enough to change the statistics of an entire season.

The article was published on a prominent sports outlet and sparked a debate that lasted two weeks. Some objected, arguing that I had oversimplified a complex phenomenon. I took that criticism seriously. But I did not retract the conclusion, because the data from 171 matches is a solid enough foundation to defend a claim, provided I state its limits clearly. Data is never sent off. It can be misread, over-interpreted, but it is not removed from the pitch just because the crowd dislikes it.

The null result: when the right verdict is no verdict

Back to the KBS room in 2026. When the VAR signal cut out, I faced a situation every analyst encounters: zero input. No camera angle, no data, no information point at all. In that situation, there are three options.

The first is to invent a conclusion based on memory of similar situations. This sounds plausible, but it is a substitution: memory of another match is not data about this match. The second is to switch to emotional commentary, talking about the "instinct" of an insider. This is usually welcomed by viewers, but it breaks my professional principle. The third is to stay silent and wait for data. That is the option I chose.

I call it the null result. In science, a null result is not a failure. It is a valid finding: it says that with the available data, no meaningful conclusion can be drawn. In football, a null result is treated as failure, because this industry runs on emotion and on the demand for an immediate answer.

I do not blame anyone; I simply trace the marks they leave on the pitch. When there are no marks, inventing a mark is an act against the profession. And this is what worries me most about the future of football analysis: we are building an industry that rewards confidence rather than accuracy. An analyst who dares to say "I do not know yet" is seen as weak. An analyst who makes a bold but wrong prediction is remembered as someone who dared to think big.

That distortion has real consequences. It creates a generation of analysis in which every match already has a ready-made story, even when there is nothing to tell. It accustoms the public to hearing firm declarations about things the speaker has no data on. And when such an analytical culture spreads, it weakens the honest analysts themselves, those who take the trouble to say that the evidence is not enough.

Looking toward Vietnam: the same disease, a different medicine

I was born in Vietnam and work in South Korea, so I have had the chance to observe two football cultures from the inside. What is interesting is that both suffer from the same disease of data culture, but with different symptoms.

In South Korea, data infrastructure is strong. Leagues have official statistical feeds, clubs have their own analysis departments, broadcasters have data rooms. The problem lies elsewhere: the pressure of real time is too great. Because everything is measured instantly, people expect conclusions to be instant too. More data does not make people more patient. Sometimes it does the opposite.

In Vietnam, data infrastructure is thinner. Domestic leagues do not yet have statistical systems as detailed as the K League's, and a large share of analysis still relies on the direct observation of practitioners. But precisely because of that, the culture here has a latent advantage: people are less pressured to produce numbers instantly. If directed well, that slowness could become a foundation for building a culture of disciplined data, instead of racing after flashy numbers nobody verifies.

What I always remind myself when comparing the two football cultures is to re-check my assumptions. The K League model cannot be transplanted directly to the V.League, because match density, refereeing styles, and even how players respond to crowd pressure are all different. A number that is correct in Seoul can become a wrong conclusion in Hanoi. A cross-border analyst must remember that data does not translate itself. It needs to be re-read in local context.

I still remember a working session with a group of young reporters in Hanoi. They asked me for the secret to building a disciplinary model. I answered that the first secret is not the algorithm, but accepting that most of the data you collect will not lead to any conclusion at all. You have to learn to endure that emptiness without rushing to fill it with guesses.

The dark side of digitisation: when data becomes a commodity

There is one aspect of this story I rarely see discussed fully. Football data does not only serve fans and professionals. It is also the raw material for another market: the betting market. And this is the darkest side effect of the digitisation of sport.

Every foul logged, every card trend modelled, every refereeing behavioural pattern discovered, all of it has commercial value. A model that accurately predicts cards does not only help viewers understand a match; it can also help a betting company price an odds line. I have received partnership offers from parties I do not wish to name. I always declined, but I know not everyone in the profession does.

This raises an ethical problem that I think the whole industry needs to confront. When you build a disciplinary data system, you are creating a double-edged tool. It can help make match officiating more transparent, shining light on patterns referees do not recognise in themselves. But it can also turn those patterns into a product to be exploited for profit, and worse, to be manipulated.

The data shield can also become a trap. Once you have built a personal brand on data, you easily fall into the temptation of using it to dismiss every opposing view, even correct ones. I have reminded myself many times that data is the starting point of a debate, not its end. A number can prove a trend, but it cannot tell the whole story of why that trend exists. If I use data to silence dialogue, I have betrayed my own method.

I have also dissected my own wrong verdicts. In 2026, my model wrongly predicted a series of card decisions in the early part of the season, because I had not accounted for the fact that some referees were rotated up from lower divisions and brought different standards of handling. That was a lesson in how a model is never complete, and how the modeller must take responsibility for its limits rather than hide them.

The gap between the number and the person

There is one thing the data sheets never display: the mental state of the people involved. When I analysed 1,847 fouls in 2026, I could count every incident precisely. But I could not count the fatigue of a player in the 80th minute after three matches in seven days. I could not count the fear of being dropped felt by a young midfielder. I could not count the pressure on a referee who knows every decision will be reviewed from twelve different angles.

That is why I always ask "why" even when the number is already clear. A player who fouls in the 85th minute is not merely a data point. He is a human being facing a limit the statistics cannot measure. If I only write about the number, I will miss the most important part of the story. If I only write about emotion, I will miss the part that can be verified. My job is to hold both sides at once, even when they seem to contradict each other.

In recent years, I have seen more and more Vietnamese football analysis adopting the international data style. That is a good signal. But I worry that some pieces imitate the form while ignoring the discipline. They cite attractive numbers without verifying the source. They build firm conclusions without stating the limits of the data. An analysis with data but without data discipline is more dangerous than an analysis with no data, because it creates an illusion of precision.

I think of one specific image. An assistant referee stands on the touchline, eyes fixed on the play, arm raising the flag. He does not have thirty seconds to think, nor twelve camera angles to consult. He has only one viewpoint and one moment. If he is wrong, the whole stadium will remember. We analysts have more time, more data, and less risk. Yet we are often more hasty than he is. Something is wrong with that contrast.

What I keep after three seasons of data

After years of working with disciplinary data, I have drawn one lesson I want to pass on to the next generation of analysts. The value of an analyst lies not in the number of conclusions they deliver, but in the quality of the conclusions they refuse to deliver. Every time you say "I do not have enough data", you are protecting the credibility of the times you say "I have enough data".

The null result is not a poor product. It is part of the process. It is a sign that a system is being honest with itself. A model that never returns a null result is a model that is lying, or one designed to always have an answer, even when that answer is fabricated.

I believe the future of football analysis will be shaped by those willing to preserve the gap. When every match is recorded from dozens of angles, when every phase is data-labelled, the scarcest thing is no longer information. The scarcest thing is the discipline to use that information. Anyone can cite a number. Very few dare to say that the number is not enough to conclude.

In my analysis room today, the third monitor is still there. Occasionally it turns grey again. Each time, I remember that night at KBS, the director's urgent question, and the answer I still consider the rightest decision of my career. No data, no verdict. That is not avoidance. It is a promise to my own profession.

Because in the end, fans do not need another person who always seems to know everything. They need someone honest enough to say that sometimes, the rightest answer is that there is no answer yet. And if the next generation of analysts learns that before learning algorithms, our football will grow up a great deal.