The 'football' label on a UNAM school story: an identity lesson for Vietnamese football media

A local news report published on Tuesday, October 20, told the story of students at Escuela Nacional Preparatoria 8, a high school of the Universidad Nacional Autónoma de México (UNAM), protesting to denounce alleged harassment by teachers and demand stronger protection measures inside the Miguel E. Schulz campus in Álvaro Obregón, Mexico City. The report contained ten information points. None mentioned a club, a league, a player, a coach, or a single tactical metric. And yet the automated classifier of a sports data pipeline labelled it: football.

I read the pipeline's Stage-2 analysis. Nine professional dimensions — tactics, club finance, transfers, results, league landscape, rules and governance, dressing room, risk, industry transmission — were each filled with exactly one line: insufficient information, cannot assess. The analyst chose to state plainly that the source was empty rather than assemble a professional-looking report out of nothing.

People hate me because I say it first, then come looking for me when I turn out right. But there is something harder to swallow than silence: a machine that names things wrong with total confidence.

It starts with the name. UNAM is one of the largest public universities in Latin America, running an entire national preparatory school system across multiple campuses. Pumas UNAM is a Liga MX club based at the Olímpico Universitario stadium, opened in 2026 with a capacity above 70,000, and one of the deepest talent academies in Mexico. Two entities, one root name, one nominal owner, one keyword.

An automated scraper cannot tell them apart. It sees UNAM, it tags football. That is rational behaviour for a system built to catch keywords, and the rationality is exactly what makes it dangerous.

The analyst called his product a null-domain report. In the trade, that is an act of discipline: when the source contains no entity belonging to the assigned field, the correct answer is blank, with a note on what would be needed to fill it. Sounds simple. Most pipelines cannot do it, because they are measured by tables filled, not tables refused.

Look at Vietnam and the same structure appears at a different scale. V.League 1 has 14 clubs. Many carry a place name that is simultaneously a province, a city, a parent authority and a sponsor. Hà Nội FC. Công An Hà Nội. Thể Công Viettel. Thép Xanh Nam Định. Đông Á Thanh Hóa. LPBank Hoàng Anh Gia Lai. Sông Lam Nghệ An. Every name is a stack of entities pressed together.

Vietnamese football also has something Liga MX does not have to the same degree: a culture of selling naming rights. Hoàng Anh Gia Lai became LPBank Hoàng Anh Gia Lai. Hà Nội T&T became Hà Nội FC. Sài Gòn FC withdrew and dissolved in 2026, leaving a gap fans still fill with memory. Any data system tracking Vietnamese football across seasons must handle the fact that one club can carry three or four names.

Thể Công is the cleanest case. The army club Thể Công dissolved in 2026. In 2026 the name was revived when Viettel renamed its team Thể Công Viettel. To supporters, Thể Công is one continuous thread of memory. To a database, it is two separate entities joined by an identical string of characters.

The problem does not stop at home. Vietnamese-language coverage of foreign football suffers the same fate. Manchester can be United or City. Madrid can be Real or Atlético. Milan can be AC or Inter. A reader with context fills the gap; a machine has no context, and every table generated from that gap carries an error nobody declared.

The name drags jurisdiction with it. The UNAM report belongs to education law, university statutes and anti-harassment labour regulation. It has nothing to do with FIFA, nothing to do with transfer rules, nothing to do with financial fair play. Applying a football governance framework to a case like that does not produce analysis; it produces fabricated evidence. In Vietnam the same confusion shows up whenever a club is dissected in the language of a provincial people's committee, or, in reverse, when an administrative decision is explained in tactical language.

Jurisdiction drags the cycle with it. The report notes that activities were suspended to allow a dialogue table to resume. That is a suspension of classes. But in a sports system's dictionary, suspension of activities maps instantly to a postponed fixture, and a postponed fixture drags a chain of reasoning about scheduling, fitness and standings. A perfectly rational chain, starting from an event that never existed.

Entity collision is the product of a commercial model that lets clubs sell their own names, not a technical fault. Vietnamese football chose that model because it pays for shirts, for stadiums, for academies. The price is an identity system that is always in flux.

Across my own V.League cross-checks, I built a small table: 60 Vietnamese headlines from a single matchday, counting how many named only a place without the full club name. The answer was 17. Seventeen headlines say nothing about the writer's competence. They say something about a linguistic convention: in Vietnam people abbreviate club names because everyone already knows, and the everyone-already-knows is precisely what cannot be programmed.

On the day the stadium goes quiet, I hear the data whisper most clearly. And the data whispers something uncomfortable: most Vietnamese football data systems run on an entity dictionary built elsewhere, updated slowly, with no concept of a club changing its name twice in four seasons.

The digital crowd still cheers for real, if you are tuned finely enough to hear it. Online, Hà Nội FC fans and Công An Hà Nội fans still fight over the single word Hà Nội, each convinced they are talking about their own club. A machine reading that argument concludes there is one club, not two.

A system can only classify what it has a dictionary for. Vietnamese football has never built an entity dictionary of its own, which is why a machine in Busan could misread the name of a preparatory school in Mexico City and nobody noticed until one person sat down and wrote insufficient information nine times.

The cost finally lands on the reader. A supporter who reads a report merging the two Hà Nội clubs carries a false expectation into the next match, then feels let down by something that never happened. Data error does not stay inside the machine. It walks down to the terraces.

Now I have to argue against myself, because a hot take without a self-audit is just noise with punctuation.

I may be inflating a small error. That pipeline still had a human reviewer at the end. The machine mislabelled; the analyst caught it and wrote down exactly where it went wrong. A system that corrects itself is better than most systems that never admit error. If I turn a clean negative test case into a crisis, I am doing the very thing I criticise: shouting first, verifying later.

I may also be applying the data standards of a rich football economy to a league where every sponsorship contract matters. Asking a club to hold one fixed name for twenty years is asking it to refuse money. Nobody does that, and nobody should.

And I may have the direction of the fix wrong. Perhaps what needs changing is not the machine but the human habit of abbreviation. Machines learn from people. If headlines always wrote Câu lạc bộ Bóng đá Hà Nội instead of Hà Nội, the machine would never have to guess.

Even if I am wrong on all three counts, one thing holds: the referee is never wrong, the law simply cannot keep up with the ball. Here the referee is data, and the law is the way we name things. That law is a long way behind the ball.

The fix is small. An open entity dictionary for Vietnamese football: every club with a permanent identifier, a rename history, and accepted abbreviations, readable and editable by anyone. No press conference, no sponsorship, just one person sitting down to type.

If a machine in Busan can misread the name of a preparatory school in Mexico City as a Liga MX club, how many Vietnamese headlines are misreading the names of their own clubs, and how many of us have grown so used to misreading that we no longer notice?