Mislabeled Data: When Vietnamese Football Must Relearn How to Trust Numbers

That night, I sat before my screen in Chengdu, watching the newsroom's automated analysis system finish processing an article. It had tagged it “football.” But the content inside was a Christmas film from Paramount Pictures, starring Johnny Depp, directed by Ti West, scheduled for release on November 13, 2026. Not a single team. Not a single player. Not a single minute of stoppage time. Only a wrong label, born in silence, ready to flow into every analysis that followed.

I sat before that screen for a long time. At 69, I have watched football travel from handwritten pages to enormous data tables. But one thing never changed: the human being is still the final reader. If the final reader trusts the label instead of the content, then an entire newsroom can talk about football while saying nothing about football at all.

That wrong label made me think of the heat maps in the analysis rooms of V.League clubs. Those blue-and-red charts look so scientific. They are projected onto screens, printed out, placed into pre-match reports. But have we ever asked ourselves: where does the input data for those colored charts come from, and who labeled it? A wrong label at the input stage can pass through dozens of processing steps without anyone noticing, until it becomes a conclusion that people trust.

Context: When Numbers Took the Throne

Over the past two decades, world football has undergone a quiet revolution. From coaches' notebooks, data became an industry. Companies like Opta and StatsBomb collect millions of events every season. Every pass, every duel, every meter run is recorded, labeled, and sold to clubs at no small price.

Vietnam has not stood outside that current. V.League clubs have begun hiring analysts. The national team holds data sessions before every major match. Youth academies teach players to read number tables from a very young age. We learn to trust numbers, because numbers seem more objective than feeling, fairer than memory, and more durable than rumor.

But the wrong label that night reminded me of one thing: numbers do not create themselves. They are made by people, by labeling rules, by classification systems that can err. And when a system mislabels, that error does not stop at one article. It flows downward, like an underground stream, into every conclusion that follows.

I remember a time, a few years ago, sitting with an analyst at a V.League club. He opened his laptop, pointed at a data table and said: “This player runs the most on the team.” I looked at the table, then looked at the pitch. The player did run a lot, true. But most of those meters were runs back toward his own goal, covering for his teammates. The number was right. But the story was wrong.

That is the first lesson I want to retell. Data does not lie. But it does not tell the story by itself either. People tell it. And the teller can tell it wrong, even with real data in hand.

The Trap of the Label

A number only means something when placed in the right spot. A label only has value when stuck to the right object. When a system tags “football” onto a film article, it does not merely commit a technical error. It exposes a gap in how we build sporting knowledge.

Imagine that flow. A film article is tagged football. It flows into a database. A language model reads that database, learns from it, and begins to believe that football is connected to Paramount Pictures. Then a young journalist, under time pressure, consults that model and rewrites. The original error, small as a speck of dust, has become part of shared truth.

In football, this mechanism is even more dangerous. Because football is a sport of decisions made in an instant. A coach who trusts a wrong metric can pick the wrong player, set up the wrong formation, and lose an entire season. A scout who trusts a contaminated data table can bring in an unsuitable player, on a mismatched salary, and the price paid is not only money but the trust of the dressing room. The error is not in the number. The error is in blind faith in the number.

I have followed Vietnamese football for five decades. I remember matches where no one recorded anything at all. People remembered through memory, retold through words, passed on over roadside iced-tea stalls. The margin of error was huge, but also very human. Today we have data accurate to the second, yet we can still err at the highest level: the level of meaning. We can count everything, but how much do we understand?

A concrete example. In a match of the Vietnam national team, one player had a very high passing accuracy metric. The data tables praised him. But when I rewatched the footage, I saw that most of those passes were backward, sideways, in safe zones, under low pressure. The number was right. But it did not say what the fans wanted to know: who is the one willing to take responsibility? Who is the one willing to receive the ball in a difficult situation, turn, and face the opposing defender?

This is where I think of what I call the “invisible referee.” In the digital era, labeling algorithms, classification models, and metric tables play the role of referees who never appear on the pitch. They do not blow whistles, do not show cards, do not get booed by fans. But they decide what we see and what we ignore. And when that invisible referee mislabels, no one can appeal. There is no VAR for data.

I have spoken about this many times in press rooms. Younger colleagues often laugh, thinking I am too suspicious. But I do not doubt data. I doubt the process that creates data. I doubt the intermediate steps no one checks, the labels stuck automatically that no one rereads, the models trained on contaminated foundations that no one knows about.

The Transfer Market and the Trap of the Number

Nowhere is that trap clearer than in the transfer market. There, data becomes money. A club decides to spend millions of dollars on a player based on metric tables that no one in the leadership truly understands how to create.

I have witnessed such deals. A player with beautiful attacking metrics on paper is brought in at a high price, then fails in a new league. The reason is simple: his metrics were generated in a completely different tactical context, where he was free, served by teammates, left alone by opponents. When the context changes, the number no longer means anything. But the label “good attacker” remains, stuck fast to his name.

This is where Vietnamese football must be especially careful. As our clubs begin buying players based on international data, we are importing not only players but entire labeling systems we do not control. A player labeled “promising” in Europe may not fit the rhythm and style of the V.League. A beautiful number in one league can be a trap in another.

I remember an old story. A foreign coach came to Vietnam, bringing with him a massive database. He trusted it absolutely. In the first training session, he believed a young player did not have the metrics to play. But that player, by the intuition of Vietnamese assistants, had something data cannot measure: the ability to read the game. By season's end, that player became a key figure. The coach later told me a sentence I will never forget: “Data told me where that player runs. But it did not tell me what that player thinks.”

The Counterintuitive Angle: Heat Maps Are the New Astrology

I will say plainly what many in the industry do not want to hear: heat maps have become a kind of new astrology. They look scientific, they have colors, they make people feel knowledgeable. But most of the heat maps I see in analysis rooms only repeat something the naked eye already saw: this player is present in many places. That is not a discovery. It is a confirmation presented as a discovery.

The problem is not the heat map. The problem is that we have forgotten a basic question: how was this data created, and by whom? The wrong label that night was a reminder. If a system can call a film football, then it can also call a player who runs a lot a good player. It can also call a safe pass a smart pass. It can also call a failure a success, simply by changing how it labels.

I am not against data. I have used data throughout my career. But I believe in one principle: data is a servant, not a master. The human eye, the human ear, and above all a healthy skepticism remain irreplaceable.

At 69, I have learned that the world still runs faster than I do, but longing always stands still. There are things in football that no data table can measure: the moment a young player first hears the stands call his name. The fear in a goalkeeper's eyes before a penalty. The joy of a fan when their team scores in the final minute. Data can count those moments. But it cannot feel them. And football, in the end, is a sport felt before it is measured.

Here is the irony: we build labeling systems to save time, yet spend hours fixing wrong labels. We create models to reduce error, yet create new errors at a deeper, harder-to-detect level. We want machines to think for us, but forget to teach them to think right.

Once, a young colleague asked me whether my generation was resisting progress. I answered no. Progress is not about using data or not using data. Progress is about using data honestly. A 21st-century sports journalist cannot only know how to write. He must know how to read data, know how to question the origin of data, and know when to trust his own eye more than a metric table.

In the dressing room of a club, I once heard a coach tell his players: “Do not read the number table of the last match. Read the match again.” That is a philosophy worth pondering. Data is a summary. The match is the original. And no one should trust the summary without ever having read the original.

What Must Be Remembered

The wrong label that night was not an isolated incident. It is a symptom. When data becomes the backbone of modern football, the quality of the data becomes the quality of the sport itself. One mislabeled article may harm no one. But thousands of mislabeled articles, in silence, can shape how a generation understands football.

I recall my “Green Balcony” project, when I called 50 elderly fans and recorded 300 minutes of their memories of afternoons spent waiting for the ball to roll before old loudspeakers. No algorithm labeled those stories. We had to sit and listen, to ask, to write down. Slowly. Time-consuming. But right. And precisely because of that, it had value.

The pitch never betrays anyone, only people often forget that it also knows how to embrace. Data is the same. It does not betray us. We betray ourselves when we trust it without verification, when we label without reading the content, when we let invisible referees decide in place of our own eyes.

For Vietnamese football, this lesson is all the more important. We are at the stage of building our data foundation. What we label today will shape what we understand tomorrow. If we learn to label carefully, to verify before trusting, to keep the human eye beside the machine eye, our football will grow on solid ground. If not, we will build a beautiful house on sand, and only realize it when the first storm arrives.

Youth is not only speed, but the way one chooses to run forward. With data, it is the same. What matters is not how much data we have, but what we choose to believe within it. A wrong label, discovered in time, is only a lesson. A wrong label, blindly trusted, becomes a system. And a wrong system is a thousand times harder to fix than a wrong article.

I write slowly, because football is not in a hurry - it only waits for those patient enough to understand. Perhaps, in the data era, that patience is something data cannot buy. And perhaps that is the gift my generation can leave to the young holding the pen: not a clean database, but a habit of healthy skepticism, an eye that looks straight at the content before trusting the label. Because in the end, the final reader is still a human being. And human beings deserve to read verified truths, not labels stuck in haste on a silent night.

Do we have the courage to ask, each time we look at a number: how was it labeled, and by whom?