When Nine Empty Columns Get Filled Before Lunch: Vietnam Football's Broken Data Supply Chain

One morning at a training centre, a 22-column report landed on my desk. Nine columns were empty. Nobody in the room knew where those nine columns had come from, which device produced them, who had pressed the stopwatch, or which phase of the match they described. By lunchtime, all nine columns had numbers in them.

The person who filled them was not a fraud. He was a young analyst who had clocked in at six in the morning, had been questioned by the coaching staff three times before noon, and had no authority to say the words "I don't know". The social cost of an empty cell is far higher than the professional cost of a wrong number. That single asymmetry explains why so many analytical reports across Southeast Asian football look fuller than they actually are.

Vietnamese football today has things nobody could have dreamed of twenty years ago: GPS vests on every academy player, event files for almost every V.League match, optical tracking systems in a handful of stadiums, data packages purchased from international vendors, and a generation of analysts under thirty who handle data better than many European peers of the same age. Collection capacity has grown exponentially. Verification capacity has barely moved.

The gap between those two curves is the subject of this piece.

The data supply chain and its three fracture points

A number on a report does not appear on its own. It passes through at least four stations: the recording device, the labeller, the cleaner, the interpreter. Loss can occur at any station, and loss at any station produces the same outcome — an empty cell filled with a guess.

The first station is the device. GPS vests sample at different frequencies depending on brand and configuration, and on whether satellite lock is sufficient. At some V.League grounds, heavy roof structures over the stands push positional error up noticeably. The same player, the same sprint, two different devices can produce maximum-speed figures that differ by several metres per second. Nobody is lying. The devices simply do not speak the same language.

The second station is the labeller. What counts as a "successful duel"? A long pass that an opposing defender heads out for a throw-in — is that a failed pass by the midfielder, or a successful clearance by the defender? Two analysts will produce two answers, and both are defensible. If labelling rules are not written down, any metric built from them cannot be compared across matchdays, let alone across seasons.

The third station is the cleaner. This is the least glamorous and most damaging station of all. Raw data always contains faults: a minute of lost signal, a substitute who entered before the system updated, a phase of play missed by a camera because of a blind angle. A strong cleaner marks clearly which cells are missing, which are interpolated, and which were simply not measurable. A weak cleaner quietly fills the gaps so the sheet looks continuous.

The fourth station is the interpreter, and that is where I stood for nearly a decade.

When a number is left blank, it gets filled by pressure, not by measurement

In 2026, at fifty-three, I accepted a role as data consultant for a V.League club. My first task was not to build a model but to write down definitions for twelve physical metrics, including high-intensity distance, pressing actions within five seconds of losing the ball, and the share of passes entering the final third. Those twelve metrics filled three pages, and those three pages mattered more than the entire software package the club had purchased.

On matchday 18, against the capital club, I saw young midfielder Nguyen Trong Huy cover just 8.2 km in ninety minutes, roughly fifteen per cent below the team average for his position. I recommended substituting him on the hour. The coaching staff ignored it. The team lost 1-3, and the third goal came from a phase in which he failed to track back in time.

After the match I presented a fourteen-page analysis with every phase timestamped and every number sourced. From that point the coaching staff began reading the report before reading the league table. The team finished fifth, four places above its pre-season projection.

But the story worth telling is not the finishing position. It is the nine empty columns in the report I received that season, and the fact that nobody ever asked me where they came from. When an organisation has no mechanism for handling missing values, those missing values find their own way out — as a plausible-looking number.

Data never lies. The people reading it do.

Lessons from the 2026 World Cup and the trap of the emotional variable

In June 2026, at fifty-four, I worked as data consultant for a sports broadcaster covering the World Cup in Russia. During the France-Belgium semi-final I sat in the operations room, feeding live figures to the commentator. On fifty-two minutes, with Belgium pressing, I supplied data showing that centre-back Jan Vertonghen had covered 7.9 km and that his average speed had dropped twenty-three per cent compared with the first half. I recommended highlighting the fatigue in Belgium's back line.

The commentator ignored it. He talked about fighting spirit. On fifty-eight minutes France scored, immediately after a slow step from that same Vertonghen, in a phase where Kevin De Bruyne had to drop deep to cover and vacated the space in front of him.

The broadcaster was criticised for missing the decisive development. Part of the blame fell on me, on the grounds that I leaned too heavily on numbers. I spent the following three weeks reviewing footage of all sixty-four matches, cross-checking every metric against what actually happened, and produced a two-hundred-page document on fatigue-index forecasting.

The 2026 World Cup taught us that emotion is the hardest data noise to filter.

But there is a second lesson fewer people mention. My data was right in direction and wrong in magnitude. I said Belgium's defence was deteriorating; I could not say it would collapse within six minutes. The distance between a measurable trend and a specific outcome is exactly where most analytics departments fool themselves. An indicator tells you the probability has shifted. It does not tell you which event will occur.

A good data practitioner is not the one who predicts correctly most often. A good data practitioner states the error margin of their own work before anyone has to ask.

Euro 2026: the report that arrived late

In 2026, at fifty-seven, I studied the effect of Euro 2026 — pushed into 2026 — on the physical condition of Southeast Asian players. I found that Vietnam's national team had six players who had already exceeded 2,800 club minutes before entering World Cup qualifying. The 2,800-minute mark is not a magic threshold. It is the point at which, in my tracking data, soft-tissue injury rates begin to rise at a visibly steeper slope.

I submitted a recommendation to reduce Nguyen Quang Hai's load against the UAE. It was ignored. Quang Hai suffered an ankle injury in the twenty-third minute, the team lost 0-1, and lost its advantage in the race for a deeper run.

I do not tell this story to claim I was right. I tell it to point out that workload data existed, was read, and still did not change a decision. A recommendation with no standing in the dressing room's power structure is not a recommendation. It is an archived document.

I later compiled my own dataset on forty Southeast Asian players who featured at Euro 2026 and the Tokyo Olympics. The result: 57.5 per cent of them declined by an average of eighteen per cent in performance within two months of the tournament. A German researcher subsequently used the report in an article on the post-tournament syndrome.

Euro 2026's injuries were not a curse. They were a report delivered late.

Every number is a confession, if we are patient enough to listen.

The transfer market: where data is bought with belief

There is a paradox in how V.League clubs value players. They use metrics to assess opponents, but instinct to price the players they are about to sign. A striker with seven goals in half a season will be valued above one with five goals but a higher expected-goals figure and twice as many attempts inside the box.

The difference is this: goals are what happened, while xG is what should have happened more often. The market pays for the first.

The transfer market is the only place where people pay for hope rather than output.

That is not commercially wrong, but it produces a very concrete technical consequence. When player prices are pegged to goals and highlight reels, academies are forced to produce players who generate moments rather than players who hold a team's structure. Ten years later we have a dazzling generation of attackers and a pool of holding midfielders so thin that the national team must either naturalise or push defenders upfield.

This is the consequence of a valuation model, not of a coaching philosophy.

Data is a mirror; a fool sees himself in it, a wise man sees the team.

When data is produced to serve an annual report

There is another kind of empty cell, more subtle, and it usually appears in documents about women's football.

In many women's competitions in the region, data is collected mainly to populate the appendix of a corporate social responsibility report: matches staged, attendance, broadcast hours, scholarships. Those numbers are real. But they do not answer the question any analytics department must answer: how does this team score, where does it lose the ball, and which line is thinnest.

When nobody asks that question, the women's game is commercialised at the media layer and left blank at the analytics layer. A women's player who scores fifteen goals in a season may not have a single shot map to prove her value to a foreign club. She is assessed by a scout's impression, while a male peer of the same age is assessed by a three-page data sheet.

This is not a moral issue. It is a data infrastructure issue, and it is fixable with money, people, and one straightforward governance decision: every women's match must be event-labelled exactly as a men's match is.

The contrarian angle: football is asking for the wrong thing

When V.League clubs talk about data, they usually ask for more: more cameras, more subscriptions, more staff, more reports. That demand is legitimate and necessary. But it does not address the root problem.

The root problem is that no number carries a provenance record. Nobody knows what it was measured with, under what conditions, by whom, and against which definition. A metric without provenance cannot be compared across matchdays, cannot enter a model, and cannot stand behind a transfer decision worth billions of dong.

Football should ask for less, and ask more strictly. Instead of column twenty-three, ask for one column stating the source of the other twenty-two.

I test myself before every report with one question: if the majority is right this time, do I have any data that would tell me I was wrong? If the answer is no, the report is not good enough to send, no matter how many pages it runs to.

By the same logic, look at the smaller clubs that have produced surprise results in recent V.League seasons. The public tells a story about spirit and local identity. The data tells a different story: these teams live on a very narrow core of thirteen to fifteen players covering more than seventy-five per cent of available minutes. A narrow core creates stability, and stability creates results. But a narrow core is also the most easily dismantled asset in the following transfer window.

A small club's success, read through data, is the opening act of another talent raid. This has happened in many leagues, and the V.League is not immune.

The trap is confusing correlation with causation. Did the team win because its narrow core was stable, or was it stable because it won early enough that the coaching staff never had to rotate? Those two hypotheses imply completely opposite transfer policies. Telling them apart requires minutes data by matchday, by points margin, and by injury status. Very few clubs store all three layers.

Being sixty-two has not slowed me down; it has told me which data is worth waiting for.

The next-cycle signal

If you run an analytics department in the V.League this season, do not start with a new model. Start with a missing-value protocol: every empty cell must be marked as "not measured", "not labelled", or "interpolated", and every report must state the missing rate of each metric.

The signal worth watching over the next two seasons is not which new metric appears. The signal is clubs beginning to publish their metric definitions before they publish their results. When that happens, a failed pass will mean the same thing in every analytics room, and only then can we start arguing about tactics in a shared language.

If the nine empty columns are still being filled before lunch, then even the most expensive model is merely decoration on a belief that was already held.

Just look at the numbers and you understand everything — but only when the number still carries the trace of the person who measured it.

---

This article belongs to the football data analysis section and draws on the author's direct observation as a club data consultant in the V.League and in international broadcast operations rooms between 2026 and 2026. Technical metrics are used according to the public definitions of the football event-data industry. The content serves no betting purpose whatsoever.