Trang chủInternational FootballCategory Error: When the Machine Calls Politics Football
International Football

Category Error: When the Machine Calls Politics Football

Huỳnh Quân2026-10-01 01:08

I have just finished reading a football analysis. Thirty data points. Not a...

I have just finished reading a football analysis. Thirty data points. Not a single player. Not a single club. Not a single match. Not one tactical system.

What had been labelled "Football" was in fact a political news report: the Chief Minister of Khyber-Pakhtunkhwa, Sohail Afridi, the PTI party, a long march scheduled for October 4, and warnings from the federal government about the possible imposition of Governors' Rule or a state of emergency.

The classification machine called it football. Without hesitation. Without asking a question. It nodded, labelled it, and filed it in the right drawer.

The moment I realised that, I realised something else: the biggest problem in football in 2026 is not on the pitch. It is in the machine sitting behind your screen, in the way things are classified, labelled and distributed whenever a ball is involved.

A machine that misreads a political article as football will also misread a midfielder as a defender. It will misread a mid-table club as a title contender. It will misread a friendly as a final. And worst of all: it will teach millions of viewers to see things the wrong way too.

An empty stadium is when the truth walks out of the data, not out of the singing. But this time the stadium was not empty. The problem was that it was a completely different stadium — and nobody in the stands realised they were sitting in the wrong place.

Football in 2026 is no longer told by people. It is told by a chain of classifications, and that chain is misreading itself.

I am writing this piece not to tell you about a province in Pakistan. I am writing to tell you about the machine that has been feeding you football for years — and about the price you pay every time it calls something by the wrong name.

This is the story of a category error. And of how football forgot how to tell itself apart.


CONTEXT: WHICH MACHINE READ YOUR FOOTBALL

To understand why a provincial Pakistani political report slipped into a football analysis pipeline, I have to tell you how football is produced today.

Twenty years ago, when I was still sitting in the newsroom of The Sporting Seoul, a football article passed through three pairs of hands: a reporter wrote it, an editor read it, and a final gatekeeper decided which page it belonged to. People classified. People took responsibility. If someone labelled a political piece as football, that person was called into an office, scolded, and reminded that he was damaging the paper.

In 2026, that chain is longer, faster and thinner. An article is harvested automatically, parsed by an algorithm, labelled by a model, pushed into a data pipeline, and finally delivered to an editor who is too overloaded to look closely. People no longer classify at the front of the chain. People only clean up at the back.

And when people only clean up, the machine will clean up in their place. It does not read for meaning. It matches. It does not ask "is this football". It asks "does this article contain a keyword I have seen attached to football before".

The political report contained such a keyword. It was three letters: PTI.

In the context of that article, PTI is the name of a Pakistani political party. But inside a classification system fed on keywords, an identical three-letter string can mean something entirely different in another context. The machine does not distinguish context. It sees a familiar pattern, and it drags the article into the football drawer.

That is a small error. But a small error at the front of the chain becomes a large error at the back. A political article that enters a football pipeline will be analysed as football. It will be searched for tactics, for line-ups, for match data. And when nothing is found, the machine will not say "I was wrong". It will say "this article lacks tactical information".

I read that analysis. Thirty data points. It contained a single economic figure — GDP at 6.1% under Imran Khan versus 2.3% under the incumbent government — and someone tried to force it into the box marked "performance data". A macroeconomic number turned into an on-pitch performance indicator.

That was when I understood: the machine is not just misreading one article. It is misreading an entire sport in exactly the way it misread that article.

Based on my experience of watching matches, I have seen this happen thousands of times. A name is labelled before anyone looks at the role. A club is ranked before anyone looks at the structure. A number is thrown out before anyone asks what it measures.

When I was fired, I did not lose a profession — I lost faith in the people sitting in the stands. And losing faith in the stands is also losing faith in the way the stands are fed information. That day, in the newsroom of The Sporting Seoul, fifteen of us lost our jobs because revenue had fallen. There was no final month's salary. But what I remember most is not the money. What I remember most is the way a system can discard human beings because of a number on a balance sheet.

Eight years later, I see exactly that logic inside football's data pipelines. Discarding context for a number. Discarding people for a keyword pattern.

So I sat down to write this piece. Because if the machine has called politics football, then it has also called many other things by the wrong name — things you still believe.


THE CORE: FIVE LAYERS OF CATEGORY ERROR DISTORTING FOOTBALL

Category error is not a rare technical incident. It is a structural disease, and it spreads through five layers of modern football. I will move from the shallowest layer to the deepest, because the deepest layer is the one you see least.

LAYER ONE: ENTITY MISCLASSIFICATION — WHEN THE NAME FOOLS YOU

Let us start with the name. Football is more obsessed with names than any other sport. And precisely because it is obsessed with names, it is the most fertile ground for entity misclassification.

Think about shared names. How many clubs in the world carry the word "United"? Manchester United, Newcastle United, Leeds United, Sheffield United, West Ham United — and that is only England. Cross into Scotland and you find Dundee United. In Italy you find a string of clubs once carrying "Unione". In South America, "Unión" appears everywhere. The machine reads "United" and cannot tell a top English club from a lower-league club in another country.

Now think about abbreviations. PSG, PSV, PFC, BSC, FCSB, CSKA, AS, FC, AC, SC, CF. The two letters "FC" alone sit in front of thousands of clubs across the planet. A classification system built on abbreviations does not classify football. It classifies identical strings of characters.

And here is the most important thing about entity misclassification: it does not stop at the level of comedy. It leaks into news. It leaks into league tables. It leaks into serious analysis.

Imagine a model that reads the headline "United win 2-0 at the weekend". It does not know which United. It does not know which league. It does not know the context. It only knows that a team called United won, and it assigns that result to the United it knows best. And so a win by a third-tier club becomes form data for a top club.

That is exactly what happened with PTI. The machine read a familiar string and assigned it to a drawer it had seen before. Context was discarded. Meaning was discarded. Only the pattern match remained.

Entity misclassification is the cheapest error to make and the most expensive to fix. It begins with a single letter and ends with a wrong conclusion about an entire sport.

I have seen this at a larger scale. During my years working in Seoul, I watched Western news agencies repeatedly confuse Asian players who shared a surname. They confused a Japanese striker with a Korean defender simply because both carried a common family name. Then they wrote about one player's form using the other player's numbers.

One wrong name, one ruined story. That is the first law of category error.

LAYER TWO: CLUB MISCLASSIFICATION — THE STAR HIDES THE STRUCTURE

Now we go one layer deeper. If layer one is an error about the name, layer two is an error about the nature.

This is the layer I have lived inside for thirty-eight years. And it is the layer where the modern machine errs most, because the machine is built to prioritise whatever is easiest to measure: reputation, transfer value, goals scored, trophies won.

A club is classified by its stars. A club with a hundred-million-euro striker is placed among the title contenders. A club with no stars is placed among the average. The machine never asks: what is this club's operating system? Where are its tactical gaps? How do the parts fit together?

Germany were eliminated in the group stage — I did not guess, I read the squad structure while they only saw stars. In 2026, I sat in a studio in Seoul, among colleagues laughing at me because I had declared that the reigning World Cup champions would leave the tournament in the group stage. They looked at the names. I looked at the gaps between the lines.

I looked at the speed of Hirving Lozano and the slowness of Jérôme Boateng. I looked at a midfield that could no longer shield a defence that had lost its balance. I looked at a team classified by its past rather than its present.

Germany lost 0-1 to Mexico through a Lozano goal. They beat Sweden 2-1 in a match they nearly lost. And they went out after losing 0-2 to South Korea in the final group game. The whole world called me a prophet. I did not prophesy. I simply classified along the right axis.

The machine's problem is that it classifies clubs along the star axis. That axis is wrong. Football does not operate through stars. It operates through structure, through gaps, through the way the pieces tolerate one another when one piece is pulled out of position.

Think about the classic misclassified clubs. A club that wins a title through a collective pressing system but is undervalued only because it has no name on the list of the most expensive players. A club that spends heavily, gathers enough stars, but has no system to bind them together, and is overvalued for no reason.

The machine cannot tell those two cases apart. It only sees the total squad value. And so it ranks the second higher than the first.

A club is not the sum of its stars. A club is the structure that lets its stars either flourish or disappear.

I call this wrong classification axis the "axis of light". The machine always chooses the axis of light because light is easy to measure. What shines is easy to count. What operates quietly is hard to count. And so the machine miscounts.

This is why most football predictions you read turn out wrong. Not because the writers are poor. Because their classification axis is poor.

LAYER THREE: PLAYER MISCLASSIFICATION — WHEN THE HIGHLIGHT KILLS THE ROLE

From clubs we descend to players. And here misclassification becomes cruellest, because it directly affects the careers of flesh-and-blood human beings.

The way the machine classifies players is simple: it counts whatever appears in the highlight reel. Goals. Assists. Beautiful dribbles. Long-range shots. Spectacular saves.

The machine calls that "performance". But it is not performance. It is a set of moments curated to sell advertising.

A defensive midfielder doing the dirty work — cutting passing lanes, covering, breaking the opponent's rhythm — does not appear in the highlight reel. The machine does not count him. And so it concludes he does not matter. But when he is absent, the whole system collapses, and nobody understands why.

That is player misclassification. It classifies by moment, not by role. And when you classify by moment, you will always see a team made of disconnected individuals rather than a functioning machine.

Think about a full-back. In one system he is an attacking wing-back, providing width and crosses. In another system he is a purely defensive player who never crosses the halfway line. The same player. Two entirely different roles. But the machine classifies him by the same name, and so it compares him with himself in a role he has never played.

This is why debates about the world's best players become meaningless. People are comparing different roles with a single yardstick. And that single yardstick is always goals.

Classifying a player by goals is like classifying a chef by the number of knives in his kitchen. It measures the most countable thing and ignores the most important one.

I have watched this destroy so many careers. A player labelled a "failure" because he did not score enough, when his real role was to create space for others to score. A player labelled a "star" because he scored inside a system designed to feed him, then failing when he moved to a system that no longer fed him.

The machine sees no difference. It only sees the number. And the number always lies when you do not ask what it measures.

LAYER FOUR: METRIC MISCLASSIFICATION — WHEN THE NUMBER MEASURES THE WRONG THING

Now we enter the layer I care about most, because it is the layer I spent six months of my life studying: the metrics layer.

In March 2026, COVID-19 shut down every league. I lost my commentary work because there were no matches to commentate. Instead of waiting, I used six months to analyse 105 Bundesliga matches from the 2026/16 to 2026/20 seasons. I compared home records before and after football returned to empty stadiums in May 2026.

The result: the home-win rate fell from 43% to 37%. Average goals rose from 2.8 to 3.1. My ten-thousand-word piece on crowdless football was subsequently shared widely among European coaches and analysts.

But here is what I learned from that research: most of the football metrics you read every day were designed for a world that no longer exists. They measure things that had meaning when crowds were present, and they lose that meaning when crowds vanish. And when crowds return, we still use the old metrics to explain a new world.

That is metric misclassification. We classify a match using yardsticks built for a different context.

Take the simplest example: possession. For decades, people believed that the team with more possession was the better team. The machine counts the percentage and ranks the higher side as the stronger one. But possession does not measure quality. It measures ownership. A team can have 70% possession and lose 0-3 because it holds the ball in harmless areas, while the opponent lets them hold it and waits for the counter-attack.

The machine does not see that. It sees 70% and concludes the other team played better. And so it writes a wrong analysis, based on a wrong metric, about a match it never watched.

Then there is xG — expected goals. This is a metric I both respect and distrust. I respect it because it tries to measure the quality of chances rather than merely counting outcomes. But I distrust it because it is so often misused.

xG measures the quality of a shot based on position, angle, type of pass and a few other factors. It does not measure intent. It does not measure psychological pressure. It does not measure whether a defender was in the right place to make the shot harder. And it does not measure whether the team actively created that situation or merely got lucky to find it.

When you use xG to say "this team deserved to win", you are classifying a match with a metric that does not measure will. You are saying football is a pure game of probability, with no people, no decisions, no moments.

And that is the deepest category error of the metrics layer: turning football into mathematics and forgetting that mathematics only describes the shell of the match, never its soul.

I bet on data before anyone called it data. Now they call it professional instinct. But I never let data think for me. Data is a tool for seeing more clearly, not an answer that spares you from looking.

LAYER FIVE: NARRATIVE MISCLASSIFICATION — WHEN POLITICS WEARS FOOTBALL'S CLOTHES

And this is the final layer, the deepest one, the layer that has brought us back to the Pakistani article from the start of this story.

For years I have watched a phenomenon I call "the invasion of narrative". Football is no longer only football. It has become a frame for telling every other story. Politics wearing football's clothes. Economics wearing football's clothes. Geopolitics wearing football's clothes.

A match between two countries becomes a diplomatic war. A transfer becomes a symbol of national power. A sacking becomes an internal political crisis.

And at some point the classification machine can no longer tell football from politics. It reads a long march by a political party and calls it football, because both share the same keywords: strategy, team, attack, defence, victory, defeat.

That is not a random error. It is the inevitable consequence of letting football be invaded by narratives that do not belong to it.

Look at how football is covered today. Half of all articles are not about what happens on the pitch. They are about what happens in meeting rooms, in bank accounts, in power relationships. They are about the politics of football, and sometimes about politics told in the language of football.

And when political narrative invades football, the classification machine — already taught to look for familiar patterns — can no longer tell the boundary.

That is why I was not surprised to see a provincial Pakistani political article slip into a football analysis pipeline. I was only surprised it took so long.

When football opens its doors to every other narrative, it loses the ability to recognise itself in the mirror. And the machine that reads your football will be the first thing to go blind.


THE CONTRARIAN ANGLE: WHERE I MIGHT BE WRONG

At this point I have to do what safe writers never do. I have to mark out where I might be wrong.

People call me a contrarian. I call them people afraid to look in the mirror. But looking in the mirror also means looking at yourself, even when the view is unflattering.

So where might I be wrong?

Possibility one: perhaps the machine is not wrong at all, and I am the dinosaur. Perhaps classifying by name, by star, by number is simply a phase of football, and that phase works better than I think. Perhaps what I call "category error" is just how modern football optimises attention, and attention is what really matters.

I have to admit this: football is an entertainment industry, and entertainment runs on attention. If a machine misreads a political article as football but that makes millions of people click, then economically it was right. I cannot deny that.

But being economically right does not mean football is right. And if we judge football only by the attention it generates, then we have admitted that football is no longer a sport. It is an emotion-producing machine.

Possibility two: perhaps sometimes the star really is the structure. There are players whose presence is itself a system. They are not a piece inside the structure; they are the structure. If so, then the machine's preference for stars is not an error — it is a truth I deny because I love structure too much.

I think this is true in some cases. But the number of such players is very small. And the classification machine cannot tell a star who is genuinely the axis of a system from a star who is merely an expensive name. It treats both the same. That is an error.

Possibility three, and this is the one that troubles me most: perhaps none of this matters to the audience. Perhaps viewers do not care whether the classification is right or wrong. They care about emotion, about story, about the feeling of belonging to a club. And if so, what I have written here is just a debate among professionals, not a problem for football.

But I do not believe that. I believe viewers care about the truth, even when they do not know they care. I believe an audience led astray by wrong classification will gradually lose faith in the very sport it loves. And when faith disappears, attention disappears with it.

You can buy players, you can buy coaches, but you cannot buy a ball that knows how to lie. Likewise, you can buy a classification machine, but you cannot buy it an eye that knows how to see.

This is where I might be wrong. But even if I am wrong about one of those three possibilities, I am still right about one thing: category error exists, and it is spreading. The question is not whether it exists. The question is whether we will fix it, or let it teach us how to see wrongly.


WHAT I AM BETTING ON: A VERIFIABLE PREDICTION

I am not writing this for you to nod and forget. I am writing it to stake my reputation on a verifiable prediction.

My prediction is accountancy, not fortune-telling. So here is the concrete number.

Within the next twenty-four months, at least one major football event — a match in a top league, a big transfer, or a governing-body decision — will be reported completely wrongly because an automated classification system misread an entity, a role or a metric. And that error will leak into at least one serious analysis before it is caught.

I am also betting on something else, harder to measure but more important: the number of football analyses built on the star axis will not fall. It will rise, because the machine is increasingly fed on whatever is easiest to measure. And so the gap between what is told and what actually happens on the pitch will keep widening.

What I learned after being fired: the truth does not sign a contract with anyone, it finds its own way on air. But in a world where the airwaves have been taken

Cầu thủ liên quan