When the Spreadsheet Is Empty: Vietnamese Football Analysis and the Trap of Missing Data

When the Spreadsheet Is Empty: Vietnamese Football Analysis and the Trap of Missing Data

On the night of January 5, 2026, Rajamangala Stadium in Bangkok was packed, and I sat nearly three thousand kilometres away, in front of a nearly empty spreadsheet. The second leg of the AFF Cup final between Thailand and Vietnam was about to begin. My file had twelve columns of data on the Vietnamese national team, and seven of them were blank. No expected goals, no pressing figures, no running-distance data. In Vietnam, most of those numbers have never been recorded systematically.

I sat still for a long time before typing the first word. In analysis, there is a very specific temptation: when data is missing, people start to invent. Not blatantly, but politely — filling the gaps with feeling, with memory of last week's match, with the belief that the stronger team will win. Then they stick the label "analysis" on it.

That moment taught me something it took me years to dare to write down: the hardest part of reading football through numbers is not finding the numbers. It is knowing when there are no numbers, and being honest about the gap. An empty spreadsheet is not a blank sheet on which you can draw anything. It is a warning.

A Vietnamese Man Reading Football from Shenzhen

I was born in Vietnam and now live in Shenzhen, working as a sports betting analyst. The job sounds alien to pure football fans, but it is really an old profession given a new name: reading the game more carefully than others, and being honest about your own degree of certainty.

I started in 2026 as an esports player and then a tournament organiser, before moving into media. The real turning point came in the summer of 2026, when I was twenty, still a sports journalism student, interning at a small tactical analysis site in Shenzhen. On that 2026 World Cup night, I looked at the ball with different eyes. During France against Argentina in the round of sixteen, I sat and calculated expected goals by hand for France's twelve shots. Kylian Mbappé generated 1.8 xG from just four runs behind the Argentine back line. I wrote an article with my own numbers, headlined with the claim that Mbappé was breaking the definition of a winger. My boss called it dull. A week later, a betting analyst shared the piece.

I realised something that remains my number one principle: data you calculate yourself is more persuasive than gut feeling, and more trustworthy than numbers you borrow without verifying them. From then on, I never wrote an analysis without building my own spreadsheet. It costs time, but it is my brand.

But from that same point, I began to notice a bigger problem: I was working in an industry where data in China and Europe is so dense you can count every single run, while in Vietnam even the most basic columns remain blank. Vietnamese fans watch the same match, but read it through a far poorer frame of information. And that affects more than perception. It affects money.

How Poor Is Vietnamese Football in Data

To picture it, compare two ecosystems. A Premier League match produces thousands of data points per game: player positions recorded by the fraction of a second, passes by pitch zone, the xG value of every shot, pressing metrics, ball-progression metrics. In the V.League, until recently, a match was usually recorded only with raw numbers: goals, shots, cards, corners. No xG. No PPDA. No passing maps.

This poverty has an infrastructure cause, not an intellectual one. Recording xG requires multi-angle cameras, positional tracking systems, and a team encoding every phase of play from video. In the big European leagues, those things are provided by specialist data companies and paid for by the league. In Vietnam, that cost exceeds the revenue of most clubs. The result is a long data gap, and that gap produces three consequences.

First, match analysis in Vietnam depends heavily on the human eye. People talk about "the flow of the game", "spirit", "character" — concepts that are real but unmeasurable. Second, transfer decisions in the V.League are mostly based on scattered video and an agent's word, not on valuation models. Third, and this is the most dangerous consequence, fans and part of the media fill the data gap with stories. When there are no numbers, people tell stories. And a story always sounds more plausible than a dry truth.

I do not say this to criticise Vietnamese football. I say it to frame the problem correctly. A football nation can win the region while remaining poor in data. But a football nation that wants to go beyond the region must become rich in data, because at a higher level the gap between teams is decided by details the eye cannot see.

The First xG Lesson of My Life

Back to the 2026 story, because it contains my entire method. When I calculated xG by hand for France–Argentina, I had no software. I rewatched the video, noted the position of every shot, estimated the angle, the distance to goal, the number of defenders blocking, then looked up the average conversion probability. Twelve shots, a few minutes each. The whole process took nearly four hours for one half of football.

The result showed France generated a much higher xG value than their actual first-half goals, meaning they had wasted many quality chances. But what caught my attention was not the total, but the distribution. Mbappé generated 1.8 xG from just four runs, meaning each run produced value close to half an expected goal on average. That is an abnormal figure. An ordinary striker needs more actions to generate the same amount of value.

I wrote the article arguing that Mbappé was redefining the winger role, not through dribbling, but through the ability to turn a single run into a chance. My boss read it and said it was dull because there was nothing new. But the "new" part lay elsewhere: before that, people judged Mbappé by speed, by explosive bursts on highlight reels. I judged him by the value each run produced. The same player, two ways of reading him, and the data-driven reading produced a completely different conclusion about how dangerous he was.

The lesson here is not "xG is a good metric". The lesson is: when you build a number yourself, you understand how it was created, and you know where it can go wrong. That is entirely different from opening a website, reading a ready-made xG figure, and believing it. A ready-made number does not tell you what its underlying model assumes about shot quality, goalkeeper position, or defender pressure. And when you do not know the assumptions, you cannot judge the reliability.

The Summer of 2026: When the Ball Stopped

The summer of 2026 was a quiet summer. The pandemic postponed almost every league in the world until June. I was twenty-three then, newly hired as a data analyst for a betting company. During ninety days without football, I had nothing to analyse, so I decided to build something else: a dataset on the rate of performance decline by age.

I collected data on three thousand two hundred players from 2026 to 2026, focusing on average running distance per match by age group and by position. The result was not shocking but very useful: wingers decline in average running distance by about twelve percent after the age of twenty-nine. Centre-backs decline more slowly. Target strikers decline along a different curve, tied more to duels than to speed.

From that quiet summer, I learned to hear football through numbers. When football returned, my company used this model to value the 2026 summer contracts. And I won a big bet by predicting that Willian, then thirty-two, would not be able to meet the intensity of the Premier League. This is the point I want to stress: that conclusion did not come from watching whether Willian played well or badly. It came from comparing his age against the running-distance decline curve of the entire group of wingers over five years.

That is the real power of background data: it lets you make a judgement about a player you have never watched for a single minute, simply by placing him in the right reference group. Fans look at Willian and see a technical player, a former Premier League champion. The model looks at Willian and sees a thirty-two-year-old in the position group with the fastest decline.

But I must also be honest about the flip side. That model can be wrong, and has been wrong with players who move to new roles. A winger who converts to central midfield no longer depends on sprinting distance, and his decline curve deforms. When I used the model while forgetting that variable, I made wrong predictions. I recorded every mistake in a separate notebook. That notebook, to this day, is my most valuable asset.

Euro 2026 and the Bet Against the Crowd

In July 2026, I was twenty-four, working at the betting company, tasked with analysing fifteen knockout matches of the Euros. This was the phase where I learned to go against the crowd with discipline.

The Italy–Austria round-of-sixteen match is the clearest example. The crowd overwhelmingly backed Italy to win, because Italy is a big team on an impressive run. But my data showed a different picture. Austria's PPDA was just 7.8, meaning they pressed very aggressively, allowing opponents few passes before applying pressure. Meanwhile, Italy's success rate for passes into the final third was only about twenty-one percent. That is a team struggling to get the ball into dangerous areas.

I recommended backing Austria with a one-goal handicap, and taking the Under on total goals at the 2.5 line. The match ended two-one to Italy, but only after extra time, and Austria held nearly forty-eight percent of possession against a big team. I won the handicap bet. My boss, who hated data, had to acknowledge the analysis, because I had given accurate figures for the deadlock the match would face.

What I learned was not "always bet against the crowd". What I learned was: going against the crowd without replacement data behind you is just recklessness dressed up as confidence. A contrarian view only has value when it is insured by a dataset the crowd has not seen. At such times, the line I often write in my analysis is: look at the number, not the name on the shirt. The name on the shirt drives emotion. The number drives decisions.

World Cup 2026: Data Knows How to Lie

In November 2026, I was twenty-five, managing a four-person analysis team. Saudi Arabia beat Argentina two-one, in a match almost no model in the world predicted correctly. This was the lesson that changed my entire workflow.

After the match, I reviewed all two thousand one hundred runs Saudi Arabia made across three pre-tournament friendlies. I found something abnormal: in those friendlies, Saudi Arabia played very deep, barely pressing, the shape sitting back. But at the World Cup, they pushed their line up unusually high, catching Argentina offside ten times in the first half alone. There are two explanations. One is that Saudi Arabia suddenly changed tactics. Two is that they deliberately played deep in friendlies to hide their real shape.

I believed the second explanation, and told my team: old data is useless if the opponent actively distorts it. Immediately, I rebuilt our noise-filtering process, discarding friendlies whose running density was more than twenty-five percent below average, because that is a sign a team is hiding its hand. From then on, every analysis of mine had to cite its data source, check reliability, and never use a single match to conclude anything about a team.

This is the point I want readers to take to heart: data is not automatically correct, and wrong data is more dangerous than no data, because it carries the appearance of objectivity to lull the reader. A wrong number printed in bold looks more credible than a truth spoken softly. That is why I question the source before I question the conclusion.

Why Vietnamese Football Analysis Easily Falls into the Trap

Now I connect these pieces to Vietnamese football. The three lessons above — building your own data, constructing background data, and checking the honesty of data — collide with a very specific context: a rising football nation with thin digital infrastructure.

When Vietnam won the 2026 AFF Cup, fans were flooded with emotion, and that was entirely right. But soon after, a wave of analysis poured onto forums, mostly based on feeling rather than data. People talked about the "character" of Nguyễn Xuân Son, the "stubbornness" of the defence, the "warrior spirit" of Đỗ Hùng Dũng. Those remarks are not wrong. But they cannot be verified, cannot be reused, and cannot be turned into decisions.

The deeper problem is this: when a football nation lacks background data, people tend to judge players by their biggest matches. Nguyễn Xuân Son scored in the final, so he is seen as a top striker. But the right question for an analyst is not which match he scored in, but how much value he generates per ninety minutes, and how stable that value is across matches. A player who scores in the final may have had low xG throughout the tournament and simply got lucky at the right moment. Another player with consistently high xG but no goal in the decisive match may be the better choice for the future. In Vietnam, we often lack enough data to distinguish those two cases.

The same goes for transfers. In the V.League, a contract is usually judged by reputation and by a few recent matches. No model measures whether a twenty-eight-year-old midfielder still has enough running to play in a high-pressing system. No model values a centre-back by tackles per ninety minutes and the quality of the opponent he faces. As a result, the domestic transfer market operates on collective intuition, and collective intuition is easily inflated by a few highlight moments.

Here I want to raise a consequence few notice. When data is missing, a player's value is not measured by ability, but by the story told about him. A player with an agent who tells good stories is valued higher than a player of equal ability who is rarely mentioned. This is something Vietnamese football can learn from how the big leagues operate: when data becomes public, player value is anchored to numbers, and stories lose their power to set prices.

The Contrarian Angle: Is Data Poverty an Absolute Disadvantage

Here I want to offer a contrarian view, because that is how I work. People usually assume that lacking data is a disadvantage. That is true in the long run. But in the short run, and in a specific market, lacking data creates a kind of advantage I call the early-mover advantage.

Think of it this way. When everyone has the same dataset, the information advantage disappears, and the market becomes so efficient that profit is hard to find. In leagues where data is abundant, thousands of analysts look at the same xG figure, and the difference comes only from slightly more sophisticated models. By contrast, in a data-poor market like Vietnamese football, the first person willing to build a spreadsheet gains a huge advantage, because almost no one competes at that data layer.

I have lived this. When I started calculating xG by hand in 2026, I was one of very few people in my environment doing so. Not because I was better, but because most people found it time-consuming and unnecessary. The very gap others ignored was where I found value. This is a paradox: a market's data poverty is not only a problem, but also an opportunity for the person who knows how to fill it.

But I must immediately place a guardrail around this view, or it becomes sophistry. The early-mover advantage exists only while the market is young and only if you genuinely do the work of building data. It does not exist for the lazy, for those who just wait for a ready-made number to read. And it disappears as the market matures. In China, where I live, the data-advantage era ended long ago. In Vietnam, that window is still open, but not for long.

One more thing must be said clearly. Data poverty is never a justification for fabricating numbers. On the contrary, precisely because data is thin, fabricating numbers is more dangerous, because no one is able to verify and catch the error. In a data-dense market, a wrong number is exposed within hours. In a data-thin market, a wrong number can survive for years and become a "fact" cited in a circle.

What This Article Could Get Wrong

I always end every analysis with a section on the assumptions that could be wrong, because that is discipline, not performative humility.

First assumption: I assume that missing data is the main problem of Vietnamese football analysis. This could be wrong if the real problem lies in coach education or league structure. Data may be a symptom, not the cause.

Second assumption: I assume that building your own data is always more trustworthy than using ready-made data. This is true in my case, but could be wrong for someone without video-coding skills. A self-calculated wrong number is worse than a ready-made correct one.

Third assumption: I assume my age model and xG model still hold. Football changes fast. A decline curve built from 2026–2026 data may be outdated after advances in recovery and nutrition. I must rebuild it periodically.

These three assumptions are a guardrail so you can read this article without turning it into an absolute belief.

A Signal for the Next Round

The ball stops rolling, but the stream of numbers keeps flowing forward. For Vietnamese football, the signal worth watching in the coming period is not a specific win, but whether someone begins to publish background data systematically. When a V.League club publishes running-distance and xG figures per match, that will be a bigger milestone than any single contract.

The crowd sleeps inside emotion; I stay awake with the spreadsheet. But I also know that my spreadsheet can be empty, and when it is empty, being honest about the gap matters more than filling it with a good story. Every match is a confession of probability. And my job is to sit and listen to that confession, even when it only knows how to stay silent.