When the Match Data Table Returns Zero: The Story of Information Quality in Vietnamese Football

The match-tracking programme I opened last Saturday had two windows. One window ran the event stream: minute, player, action type, outcome. The other ran the positional stream: every player's coordinates on the pitch, sampled ten times per second. At minute 87, the event stream was still full. The positional stream returned zero. No coordinate line. No label. Just an empty frame sitting motionless on the screen at the exact moment the match most needed explaining. A crowd can remember a goal forever. I remember the third pass before it, where the real decision was made. But that night, the third pass did not exist in any of my data files. It existed in memory, in the video, in the roar of the stands, and everywhere else except the place it was supposed to be. I recorded the moment the feed broke: 21:40:12. Then, out of professional reflex, I opened three independent sources to cross-check. All three were empty in exactly that interval. A silence lasting four minutes, then the feed restored itself, returning coordinates as though nothing had ever happened. But those four erased minutes contained precisely the decisive passage of the entire match. And nobody in that night's press room knew that the thing they were commenting on had already vanished from the system. I stayed another forty minutes in my office in Da Nang, watching the empty frame blink. In thirty years on the job I have seen every kind of wrong data: missing data, skewed data, data typed in a hurry by people who did not understand what they were typing. But a feed that returns a perfect zero, at exactly the right time, then self-heals without leaving a trace, is the most dangerous kind of failure. It does not shout. It simply stays silent, and lets the reader believe everything is fine. That is why I am writing this. Not to report a technical incident, but to speak about what is happening quietly beneath the surface of the entire Vietnamese football analytics industry: a chain of information that very few people verify, yet which governs almost every major decision, from tactics to transfers to the way we remember a match. When the press room laughs at xG, I know I am reading the right book, the one they have never opened. But today I have to admit something more uncomfortable: sometimes that book has blank pages, and we only discover it after we have already quoted it in five different articles. Vietnamese football entered the data era about a decade later than Europe. By the time the Premier League was accustomed to every match generating millions of positional data points, most V.League matches through the mid-2010s were still recorded with manual stat sheets: shots, passes and possession counted by eye. Those numbers were not wrong out of malice. They were simply not enough to answer any serious tactical question. I remember 2026, when I covered a match between two top-four V.League sides, asking a head coach about his team's expected goals after a 1-0 win. He looked at me as though I had asked about the weather on Mars. The whole press room laughed. One male colleague announced loudly that women know nothing about football and just invent numbers. I did not argue. That night I stayed behind, retyped every passage of the game from the video, reconstructed the whole match minute by minute, and published an analysis longer than three thousand words. The conclusion was simple and uncomfortable: the winning team won through luck, not through dominance. Their expected goals figure was only 0.4; the opponent created better chances but did not score. The piece was shared more than two thousand times that week. Nobody apologised, but at least nobody laughed again. That was the first lesson, and it still holds today: in Vietnam, the person who puts data forward first usually has to answer for that data before anyone is willing to answer for listening. So I learned one rule I never break. Before offering any judgment, I cite the raw numbers. And before citing the raw numbers, I cross-check them against at least two independent sources. It sounds simple. But with the thin, fragmented data infrastructure of the V.League, cross-checking two sources often reveals that both are copying the same origin. This is called single-point dependency. One faulty counter, and three newspapers publish the same wrong number on the same day. Vietnamese football's data structure today can be divided into four layers. The lowest is raw event data: shots, passes, duels, cards, collected by service providers or by the clubs' own coaching staffs. The second is positional data, present only in the few matches equipped with a sufficiently dense camera system. The third is derived metrics computed from the two layers below: expected goals, passes allowed per defensive action, success rate of passes into the final third. The fourth is the layer of interpretation, where a number becomes a story, and also where everything is easiest to distort. Most readers only ever touch the fourth layer. They do not see the empty first layer, the second covering only forty percent of matches, the third interpolating numbers it has no data to compute. They see a tidy headline with a handsome figure. And they believe it. In the season the pandemic forced matches behind closed doors, I analysed a sample of one hundred and fifty-six V.League matches and found the home win rate fell from forty-six percent to thirty-eight percent. It was a shift never previously recorded at that scale. An empty stadium does not remove the truth. It only strips away the mist that forty thousand shouts once created. When the shouting is gone, prediction models trained on historical data begin to return systematically skewed forecasts, because they were trained on a world that no longer exists. That story taught me something I carry into every article since: when the context changes, old data loses its value, and the data analyst must be the first to doubt his own archive. An analyst at a Hanoi club shared that piece and told me they applied the idea of a correction coefficient for away matches. I do not know exactly how they used it. But I know at least one place was reading the number correctly. Back to the empty frame at minute 87. That incident is not rare. It is only rarely seen, because very few people habitually check the continuity of a feed before using it. A feed that breaks for four minutes and then heals will not raise an error message. It will produce a data file that looks entirely normal, missing exactly the four most important minutes. When I rebuilt the match from video and compared it with the file, I found three types of deviation. First, the assist leading to the goal was attributed to a different player than the one who made the decisive pass. Second, the passage leading to the goal was logged as starting from a goal kick, when in fact it started from a midfield turnover. Third, and most seriously, the entire four-pass sequence before the goal was collapsed into a single event, erasing all information about the structure of the move. These three deviations are not mere technical faults. They change how we understand the cause of the goal. If the passage starts from a goal kick, it is the product of a rehearsed pattern. If it starts from a midfield turnover, it is the product of pressing and transition. Two entirely different tactical stories, leading to two entirely different conclusions about a team's quality. Every transfer contract is a multi-variable equation. Most journalists only look at the coefficient before the equals sign. They look at the fee, at last season's goals, at age, and conclude. But the coefficient before the equals sign is only the most visible part of the equation. The real unknowns lie behind it: in which system did the player perform, under what pressure, alongside which teammates, at what point in the season. Over years of watching the domestic transfer market, I see a recurring pattern. A player scores heavily at a mid-table club and moves to a big club for a record fee. The next season he scores less and is called a disappointment. But when I analyse his positional data at both clubs, I see that the quality of the chances he receives has fallen sharply, while his touches inside the box remain almost unchanged. He did not get worse. He simply played in a system that generated fewer chances for his position. That is why I believe the transfer race among big clubs, in Vietnam as in Europe, is largely an arms race of branding. A club buys a player to prove it can buy. The genuinely valuable signing, measured by performance per unit spent, usually sits at the smaller clubs, where nobody is watching, where recruitment is based on system fit rather than headlines. There is a paradox here I have never fully explained. The more data becomes public, the more alike transfer decisions become. When everyone reads the same metrics, they reach the same conclusion, and they bid for the same player. Data, which should make the market more efficient, is making it more uniform. And in a uniform market, the advantage lies not in having data, but in having different data. That is why I spend most of my time building indices nobody has published. Not to be unique, but because the value of an analyst lies not in reading the metric sheet everyone can read, but in seeing what lives in the gaps between the numbers. And the largest, most dangerous gap is the one nobody knows exists. Across my whole workflow, I install a step I call the gap check. Before computing any metric, I check whether the raw data is complete. I count the minutes with positional data, the logged events, the players appearing in the feed. If any figure falls below the expected threshold, I stop and do not compute. This rule sounds obvious. But in practice, most analytical workflows I have observed skip it, because a metric computed from missing data looks exactly like a metric computed from complete data. Nothing on the screen tells you the number is lying. A number can lie, but a model validated across ten thousand matches has no reason to pretend. The difference between the two lies not in the complexity of the algorithm, but in the discipline of verification. A model is trustworthy only when its builder has spent more time trying to destroy it than to display it. That is the hardest part of this trade, and the least discussed. People like to talk about correct predictions. Nobody talks about the times a model returned meaningless output because the input was skewed. Nobody talks about the night a data column changed format, turning the next day's entire output into garbage. When I began this work, I thought the enemy of data analysis was public ignorance. I was wrong. The real enemy is the ignorance of the analysts themselves, and their willingness to trust numbers they have never checked. There is a great temptation anyone in this craft long enough has felt: to turn data into a religion. When you have spent thousands of hours rebuilding matches in numbers, you begin to believe the numbers are the truth and everything else is noise. But a number is a map, not the territory. A map can be very detailed and still lead you off a cliff, if it was drawn for a different land. I force myself to remember that whenever I read a beautiful metric sheet. I ask: in what conditions was this data collected? Who labelled it? What motive did the labeller have? And if all of it vanished, what would I still know about this match? The last question is the most important. And the answer usually unsettles me. In the match with the four-minute gap, if I relied only on the file, I would tell an entirely different story about the goal. I would say it came from a quick counter, because that is how the file was structured after healing. But the video shows another story: patient pressing, a lateral pass to stretch the opponent, then a vertical pass to break through. Three actions, three intentions, three moments. The file merged them into one. This difference looks small. But it is the difference between understanding a team and describing a team. At the macro level, this story repeats across the whole Vietnamese football system. Youth academies at major development centres such as Hoang Anh Gia Lai, PVF, and Viettel have produced many quality generations. But the data on their development, from age twelve to first-team debut, is almost never systematically preserved. We know a player succeeded. We do not know why. That is an enormous data gap, and it is not merely academic. When a club has no data on a player's development path, it cannot value its own asset correctly. When it cannot value it correctly, it cannot sell at the right price, and it cannot reinvest properly. This causal chain stretches across seasons, and the price is an entire academy system working below its true value. I once spoke with a scout at a mid-table club. He told me he assesses young players by eye, by feel, by fifteen years of experience. I asked where he stores those assessments. He said: in my head. I did not laugh. I understand why it happens. But I also understand that when he leaves, fifteen years of knowledge leave the club with him. That is the definition of a failed data system: a system where knowledge exists but cannot be transmitted. Back to the concept I raised above, the gap check. In the software industry it is called input validation. Before processing, check. In Vietnamese football it barely exists. Stat sheets are published with empty cells recorded as zero rather than recorded as no data. And the difference between those two things is the whole story. Zero is a fact. No data is a blank. When a stat sheet says a player made zero tackles, it tells you he made no tackles. When it says zero tackles because the system failed to record him for twenty minutes, it lies to you about a number that does not exist. This is the error I fear most. Because it does not produce a wrong number. It produces a correct number for a different question. And the reader, with no way to know, believes that number with total certainty. In recent years the volume of Vietnamese football writing containing derived metrics has risen sharply. That is good. But what I am unsure of is whether that rise has come with a rise in verification discipline. From my own match-watching experience, I believe the answer is no. I have read many analyses citing numbers that, to my knowledge, cannot be computed from available data. Not because the author deliberately cheats. But because they received those numbers from some source, and that source received them from another, and nobody in the chain ever checked the origin. This chain has a name: broken telephone. Each link passes on a slightly distorted version, until the original is unrecognisable. In football, broken telephone runs faster than anywhere else, because there is an incentive to transmit information faster than to verify it. Headline first, truth later. In such an environment, verification discipline is not a moral choice. It is a competitive disadvantage, and very few are willing to accept it. I am. I have never wanted to be the fastest. I only want to be the one who is right. That brings me to the central question of the data gap: if a gap can exist without anyone knowing, how do we know what we read is true? The honest answer is that we do not know. We can only reduce the probability of being wrong. And the only way to reduce it is to build the habit of doubting ourselves. At forty-six, after thirty years in the trade, I realise the thing I am protecting is not a method. It is an attitude. An attitude hostile to easy certainty, and friendly to discomforting uncertainty. I predicted Croatia would reach the 2026 World Cup final when nobody believed it. I analysed every team's qualifying campaign and found Croatia held one of the highest pressing metrics in Europe, with passes allowed per defensive action at just 8.2, and a success rate of passes into the final third among the top three. Many colleagues called me a keyboard prophet. When Croatia did reach the final and lost 2-4 to France, I received a few apologies, and an offer to work as a pundit for a major network. I declined. I wanted to stay where I could dig deep into data rather than speak in soundbites in front of a camera. Croatia did not reach the final through luck. Croatia reached the final because I counted how many times they ran twelve kilometres more than their opponents. I counted every metre. Not because I like counting, but because counting is the only way to turn a myth into an audit. But even in that story I must admit a limit. I can count distance run. I cannot count fear. I can count successful presses. I cannot count the fatigue accumulating in a player's legs at minute eighty. Data always has a border, and beyond that border lies everything that cannot be measured. This is where I face the greatest limit of my own method. For years I have written and said that data is a jury that cannot be bribed. I believe it. But I failed to see something I now see more clearly: a jury can only judge on what is placed before it. If a piece of evidence is withheld, whether by accident or design, a verdict will still be delivered, and it will be wrong. And the worst part is that the wrong verdict will be announced in the tone of absolute certainty, because the jury does not know evidence was withheld. So the turning point in my thinking in recent years is not learning to trust data more. It is learning to trust data conditionally. That is also why I write the column I call the data perspective, where I test my own hypotheses before publishing. I do not write to prove I am right. I write to find where I might be wrong. It sounds backwards in an industry where everyone wants to be declared right. But after thirty years I realise the long-term value of an analyst lies not in how often he is right, but in how often he discovers he was wrong before anyone else does. Back to the empty frame at minute 87. What I learned from it is not a new technique. It is a new humility. I spent seven years replacing the shouts on the pitch with numbers that cannot be argued with. But that night I realised there was a stretch of time when those very numbers shouted and then vanished. And I nearly believed them. This makes me re-examine how the entire sports data industry is developing. We are building ever more complex models on ever more fragile foundations. We are measuring more and more things, without checking whether the things we measure are actually present in the data. We talk about artificial intelligence and machine learning while still failing to solve the most basic problem: an empty cell recorded as zero. That is the paradox of our age. More data demands more discipline. Faster demands slower. For Vietnamese football, I think this is a crucial moment. Clubs are beginning to invest in data analysis. Academies are beginning to keep player records. Governing bodies are beginning to talk about the importance of data in developing professional football. This is a chance to do it right from the start, rather than build a tower on a foundation not yet poured. But that chance will pass if we do not look straight at the gaps. Not to decorate them, but to fill them. And the first step in filling a gap is admitting it exists. There is a question I ask myself whenever I begin a new analysis: if I had to defend this conclusion in a court of law, do I have enough evidence? If the answer is no, I do not write. If the answer is yes, I still check once more, because the real court is never the court of readers, but the court of time. Time is the only judge that cannot be bribed, and it is never in a hurry. It will wait until every gap is exposed, every wrong number found, every flashy prediction forgotten. And when time has ruled, people finally see who checked their data before publishing it. I do not write to be remembered. I write so that, on the day time rules, what I leave behind still stands on numbers that were checked. For Vietnamese football, I believe the signal for the next cycle has appeared. Not in ever-denser stat sheets, but in ever-harder questions clubs are starting to dare to ask: where did this data come from, who collected it, and how did we verify it? When those questions become habit rather than exception, Vietnamese football will enter another phase, not of more numbers, but of numbers more worthy of trust. As for last Saturday night, I saved that gap into a separate file. I did not delete it. I keep it, as a reminder that in this trade the most dangerous thing is not a wrong number. It is a gap disguised as a number. And my job, every day, is to tell the two apart.