Vietnamese Football Is Analysing Itself With Empty Data

Eleven o'clock on a Saturday night. I open the four V-League matches I recorded during the week, rewind, and stop on the build-ups I marked in red pencil. On the desk there is a notebook, a laptop, and three open tabs. The first tab is a club's homepage, where the match-stats section has been empty since the season began. Another tab is a broadcaster's score page, where possession is listed as 53-47 and nobody says how it was measured. The last tab is a fan page with forty thousand followers, posting a heat map of average positions, an image I first saw online in 2026 in an article about the Premier League.

Four Vietnamese matches sit in my machine, and I have not a single piece of raw data to cross-check against. People call me a disruptor. I call it seeing one step ahead of the run. But tonight there is nothing to disrupt, because there is nothing to compare.

Most of what passes for analysis in Vietnamese football today is storytelling with numbers attached, not measurement. I write that after fifteen years in the trade, and I write it as a man who earns a living from invisible metrics.

What the V-League publishes consistently is goals, cards, minutes played, and occasionally attendance. No expected-goals model is public. No event data per touch. No agreed definition of a pressing metric, which means two people discussing the same match may be discussing two different things. The big academies, Hà Nội FC, Viettel, Hoàng Anh Gia Lai, Công An Hà Nội, all put GPS vests on their players, but that data lives in the analysis room and in contract negotiations, not in the newspapers.

Demand runs the other way. After Thường Châu 2026 and the AFF Cup title the same year, a generation of Vietnamese fans learned expected goals from Premier League broadcasts and then applied it to a league that publishes no expected goals. The result is an imported analytical culture: someone else's heat maps, someone else's vocabulary, conclusions about Vietnamese players. People slap the label gegenpressing on a match in which a team simply ran more than its opponent in the second half.

I still cling to the simplest public data, because it has never lied to me. In June 2026, with global football suspended, I rewatched every V-League 2026 match I had missed and wrote down each squad's average age. Quảng Nam's was 29.8. Hà Nội FC had six players born between 2026 and 2026, each of whom played more than fifteen hundred minutes. I wrote that Quảng Nam would be relegated and that Hà Nội FC would dominate the V-League for five years. At the end of 2026, Quảng Nam went down. In Quảng Nam I learned one thing: people hate you for being right a season ahead of them.

But that is public data, available to anyone. The difficulty lies a layer below, where no spreadsheet exists.

Four questions I put to any metric handed to me. How is it defined, and does that definition hold across matches. How large is the sample, and is it distorted by two fixtures against bottom-half teams. What was the quality of the opposition, because a pretty number against a side with nothing left to play for says little. And who did the counting.

In the V-League, each broadcaster computes possession its own way. I once compared two sources for the same match and found an eight-point gap. Neither was technically wrong; they were simply counting different things. A pressing metric built on two matches is noise, not a trend. A heat map does not tell you where a player received the ball, or in what phase.

So I count by hand. In my notebooks each match has four columns: the number of passes the opposing back line is allowed before my team's front line engages, the receiving positions of both full-backs in the first twenty minutes of each half, the number of balls recovered within ten seconds of a long clearance, and the rest-defence shape when possession is lost in the opponent's half.

Those four columns tell a story the scoreboard cannot. A team that wins 2-0 while allowing its opponent eighteen passes on every possession controls nothing; it merely met a side that did not know what to do with the ball. A team that loses 0-1 while repeatedly forcing long clearances and winning second balls around the box is on the right path. The scoreboard is the easiest thing to read and the easiest thing to be fooled by.

In a league that publishes no data, the man with a stopwatch counting by hand is the last honest instrument left. That is precisely why my job has become harder, not easier, in an era when everyone talks about data.

Take the back-three trend. Whenever a team loses two in a row and its back four is torn open, it lines up with three centre-backs the following week. The press calls it a bold change. I call it reputation-risk management. When a back four leaks, the coach is singled out. When a back three leaks, the blame is shared by the system, by the wing-backs who did not run enough, by the midfielders who did not drop in time. Responsibility is diluted.

The technical problem lies elsewhere. A back three only works when both wing-backs have the lungs to cover the flank and the ability to defend one against one. In the V-League, the number of full-backs who satisfy both conditions can be counted on one hand. The result is a team that moves from a back four being exposed on the flanks to a back three being exposed on the flanks, except the hole now sits in a position where nobody wants to own the mistake.

There is a line I wrote years ago and still stand by: Tiki-taka did not die. It merely exposed the limits of soulless disciples. That holds for every system imported here, the back three included. People take the tactical shell and leave behind the physical base, the pitch quality, and above all the data needed to test whether the shell fits the body. I call my own approach anti tiki-taka: no worship of possession, no worship of formations, only worship of evidence.

The money flows in a worrying direction too. Streaming platforms pour cash into broadcast rights, lose money to secure exclusivity, and repeat the exact mistakes of pay television twenty years ago. Almost none of it reaches data infrastructure. A league that sells its streaming rights at a high price while still refusing to publish per-match event data is selling emotion, not instruments. Fans pay to watch, and still have to trust someone else's word about what they just watched.

Where I could be wrong.

There is a version of this argument I do not want to belong to. The version that says Vietnamese football only needs to import Western metrics and everything will become clear. The tempo of this league is different. The heat, the humidity, the pitch quality are different. Metrics measured on English turf cannot measure a seven o'clock kick-off on a surface that has survived ten rounds. A coach who has stood on the same pitch for fifteen years may read a match better with his eyes than a spreadsheet can. And I have to admit it myself: the Quảng Nam call was one season, one sample, in a year when the format was scrambled by a pandemic. People celebrated me for the result, but I know my hit rate is only proven when I dare to publish a second, third, tenth prediction.

It is also possible that fans do not want data. They want stories. The night I wrote about England, the whole country was singing. Three weeks later they understood why I was not. But in Vietnam, the people arguing online do not want numbers in order to understand; they want numbers in order to win. If that is true, data is just a new weapon in an old fight, and I am helping to hand it out.

That is my biggest blind spot: I believe measurement makes argument better, while most argument only wants a winner.

Milestone. At the end of round ten of the 2026-2027 season, I will come back with two questions. First, at least two V-League clubs publish match data at the per-possession level, even if only internally and opened to the press. Second, the four teams at the top of the table at that point field a starting eleven with a lower average age than the rest of the league. If both hold, I keep counting by hand. If both fail, I will be the first to write that the disruptor was wrong, and I will print it.

The value of a prediction does not lie in the result. It lies in whether you dare to print it.