Vietnamese Football: When the Analysis Model Returns an Empty File

On a Saturday evening I reopened the spreadsheet I had built for a round of V.League fixtures. Twelve metric columns, three data layers, a separate tab for pressing. All of it blank. No broken formula, no misformatted date. The feed simply did not exist.

I sat looking at that skeleton for a while. Eighteen thousand people in the stands had seen the match. Hundreds of thousands watched it on a screen. But when the final whistle went, almost no number had been recorded in any way that could be queried again the next morning.

I am not here to complain about technology. I am describing the state of a football culture I have followed long enough to know that this void is not rare. It is the default.

Data going missing is not the loss of data — it is a category of data. And when a football culture produces too much of this "empty data," the first thing eroded is not analytical capability. It is the ability to know what you are actually talking about.

To understand how a spreadsheet can come back empty after a professional match, you have to know which route Vietnamese football data travels.

In Europe's top leagues, that chain is fairly clear. A data provider signs with the league, builds a camera system, tags events minute by minute, then sells the feed to clubs, bookmakers and broadcasters. A shot in the 88th minute is logged with coordinates, the shooter, the type of situation, the creator.

In Vietnam, the chain is broken into pieces. The league is operated by VPF, but the data rarely belongs to VPF as an exploitable asset. Clubs are backed by corporations. Match information flows through newspapers, social media and fan groups rather than a unified technical pipeline.

Vietnamese football runs on a corporate-patronage model. A large conglomerate keeps a club alive, pays wages, pays bonuses, and sometimes pays amounts that never appear in a public report. When resources come from one person or a small group, information follows the same route: it does not spread into a system, it contracts into relationships.

The league narrows around Hanoi in terms of resources. Playing fields, academies, investment and media attention all lean toward one pole. This is not inherently bad, but it creates a structure in which data struggles to survive, because data needs repetition and neutrality, while relationships need distinctiveness and personalisation.

That is the context. Not a story of backwardness, but of a different structure. Misread that structure and every conclusion downstream is skewed.

In 2026 I tasted the opposite feeling. I ran an xG model on a match in another national league in the region, where event data was logged densely enough for the model to have something to read. The result matched the prediction, the piece spread, and I believed I had proven something.

By the 2026 World Cup that belief was paid for. A model built on PPDA and defensive-line height called a shock result correctly, and I enthusiastically urged people publicly to follow it. Then in the next round the model said a strong team would win. They lost. A lot of people lost money for listening to me. I spent three weeks rewriting the code, adding league variables, adding noise.

I tell this not to flagellate myself. I tell it to say that even with enough data, models still fail. So when the data does not exist, what exactly are we doing?

That is the central question of Vietnamese football. In the V.League, what we have is not a poor model. It is a model with no raw material.

A concrete example. xG needs shot coordinates, situation type, goalkeeper position, number of defenders. In a V.League match, what percentage of shots have all four of those data points logged and published? The honest answer is: we do not know, because nobody measures it. To know, you would need an independent recording mechanism — the thing the league has not built.

PPDA requires counting how many passes the opponent makes per defensive action. To count, you must classify defensive actions by phase and by zone. In many matches, even the accurate-pass count for each team is a contested figure between different sources.

Based on my experience watching matches across many seasons, the consequence does not stop at missing metrics. The consequence is that we are forced to judge by eye, by feel, by relationships — and those things are not wrong, but they do not accumulate into knowledge.

There is one exception: the academies. Hoang Anh Gia Lai brought its first cohort into the senior team around 2026, after years of training at the affiliated academy. PVF built its own centre. These are places where longitudinal data exists — height, weight, minutes, injuries — because an academy must track players over years, not over one match.

But that data stream stops fairly early. When a player reaches the first team, when the first contract is signed, when a transfer happens, the information scatters. Intra-V.League transfer fees are rarely published. Contract lengths are not fully disclosed. Wage structures even less so.

So what I call data here is not a number on a website. It is the ability to reconstruct an event in a way that two independent people can verify and reach the same conclusion. Vietnamese football, across most of its dimensions, does not yet provide that.

There is one kind of Vietnamese football data that is produced in great abundance: controversy data. Every round, a VAR situation, a penalty, a red card is enough to generate thousands of comments, articles and frame-by-frame analysis clips. Information here is not scarce. But almost all of it is reaction, not recording.

xG does not score goals, but it makes people argue more than the actual ball does. In Vietnam, we get the arguing part first, and the metrics arrive later — or never.

The national team is the clearest exception. When it plays in World Cup or Asian Cup qualifiers, event data arrives from AFC and FIFA systems. Suddenly we have passing metrics, pressing metrics, heat maps. But that is data about the national team, not about the domestic league. And it appears in cycles, not continuously.

That means a player can be measured very closely for ten matches a year and be entirely invisible in the other thirty. A coach can be judged through the eyes of spectators all season, then suddenly be re-evaluated by the whole international metric system over two weeks.

Overseas moves lay the paradox bare. Doan Van Hau went to SC Heerenveen on loan for the 2026-2026 season. Nguyen Quang Hai went to Pau FC in France for 2026-2026. Each time, the data volume about that player spiked, because the destination league already had recording systems, and then dropped again when they returned.

That gap creates a larger paradox. The higher you go, the more data there is. The closer you get to football's cradle, the less there is. Precisely the place that most needs data for long-term decisions — hiring young coaches, investing in academies, retaining players — is where data is thinnest.

Here I have to stop myself.

The first temptation is to conclude that data explains everything. It does not. Correlation is not causation. A team with higher xG is not guaranteed to win. An academy with more data is not guaranteed to develop better players. Sometimes dense data just means someone bothered to write things down, not that someone did things right.

The second temptation is the opposite conclusion: without data, nothing can be analysed, so skip it. That is a mat to lie down on. I have seen pieces arguing that Vietnamese football cannot be assessed because it lacks numbers, and using that line to avoid reaching any conclusion at all. That is indifference dressed as caution.

The third temptation is importing models. Take something built for the Premier League, pour V.League data into it, and trust the output. This trap is the most subtle, because it produces a number. But if the input was recorded to a different standard, by different eyes, for a different purpose, then the model is not measuring Vietnamese football. It is measuring the importer's assumptions.

All models are wrong, but a few are wrong in useful ways. The right model for Vietnam right now is not one that predicts scorelines. It is one that points to where the data is rotting.

Stay here a moment. If I had to name the single biggest blind spot for Vietnamese football over the next three years, it is not the defence and not the attack. It is this: decisions about people are made through relationships more than through records, because the records do not exist. And a football culture that decides about people without records will always trend toward reusing what is familiar instead of trying what is new.

That claim may be wrong. I leave it open rather than closed.

The signal I will track in the next round is not the scoreline.

It is something much smaller. Whether, after a round of fixtures, someone publishes each player's minutes in a way two different sources can confirm to the same figure. Whether a club states a player's contract length publicly. Whether a match produces a data map sufficient for one person in Hanoi and one in Saigon to see the same thing.

If those things start happening, every model behind them will change. If not, we keep living in a football culture that returns an empty file — and its most beautiful stories will still be written by feel, not by evidence.

Every spreadsheet is a meditation, except that when you finish you have lost money. I kept that blank sheet from Saturday evening. Not to complain. To remind myself that in Vietnamese football, the most telling thing is sometimes not what we know, but what we have never recorded.