The Mislabelled Drawer: When a Grammy Bulletin Landed in the Football File

2:14 a.m., Kuala Lumpur. I was sitting in front of the screen with my eleventh coffee of the day when the item surfaced in the aggregator under a small blue tag: "Football".

Beneath that tag was a story about a Mexican singer who had just received a Best New Artist nomination at the 27th Latin Grammy Awards. The ceremony takes place at the MGM Grand Garden Arena in Las Vegas on 12 November. The nominations list was published on 16 September and contains 11 names from different countries. The singer wrote on Instagram that life is beautiful, that you ride a bike around the city and one day you get nominated at the Grammys.

I sat still and looked at that item for about thirty seconds. No club. No player. No coach, no line-up, no transfer, not a single line about a ball. Nothing but a label pasted in the wrong place.

In eight years working alongside teams in Malaysia, I have learned something no school teaches: most mistakes in this trade do not happen where people look hard. They happen where people paste labels. A wrong label does not hurt anyone immediately. It quietly pushes a piece of information into the wrong drawer, and three weeks later somebody opens that drawer and believes they are reading about football.

I saw something similar on Selangor FC's worst days after the 2026 AFC Cup semi-final, when torn tactical slips lay scattered across the tunnel floor, and not one of them still carried its original meaning. A dressing room never lies — it is only that the listener is not patient enough.

A transfer window drowned in noise

The transfer window is when my profession gets loudest, and also when I have to say the least. In Malaysia, as in Vietnam, Thailand or Indonesia, every summer brings thousands of lines: this club is bidding, that club refused, this agent is negotiating, that player has been photographed at an airport. My morning digest can run to three hundred lines. Of those three hundred, perhaps four are true.

Outside readers often assume the problem is fake news. They imagine someone invents a story, and our job is to expose it. My experience says otherwise. Most of what readers absorb during a transfer window is not fake: it is true information placed in the wrong drawer, real information attached to the wrong motive, old information wearing the clothes of new information. People rarely invent. They label.

I have a habit my newsroom colleagues used to laugh at: before a big match I arrive two and a half hours early, not to interview anyone, but to stand in the corridor and watch the players walk past. I note the order in which they enter the meeting room. I notice who ties his laces first, who looks down, who laughs. Eight years of that taught me a team speaks through details no data sheet ever records. I look at the feet, but I listen to that kid's heart.

And precisely because I trust small details, I do not trust the way most football information is sorted today.

The classification layer: where truth bends without anyone bending it

Picture an aggregator. It takes in thousands of articles a day, from hundreds of sources, in dozens of languages. It does not read the way a person reads. It counts. It looks for familiar keywords — competition names, club names, player names, the word "transfer", the word "squad". When keyword density crosses a threshold, it applies a label.

That label is a decision. Not an observation.

In the case I described, that chain of decisions went wrong at a very small point. An article about a music nomination contains the word "award". A system had learned that "award" often appears alongside football — the Ballon d'Or, player-of-the-year awards, organiser awards. So it labelled. From there the item entered a pipeline, and at the end of that pipeline was my screen at 2:14 a.m.

The telling part is this: had I not looked closely, I would have missed it. Had I been an ordinary reader scrolling, I would have passed it by and subconsciously noted, "ah, something about football again today". The human brain does not check every label. The human brain trusts the frame.

This is where I want to pause a little longer, because it matters more than it appears.

Once I asked a seventeen-year-old boy at the Selangor training ground why he never read transfer aggregators. He answered roughly that he used to, until one day he realised he was mentally preparing for a departure that never happened. He said it lightly, but I have never forgotten it. That boy was Aiman Zakri, then without a place in the official squad list, riding a bus forty minutes each day to training. I followed him for three sessions, and that was the first time I understood that a wrong label can enter a person's emotional life.

The classification layer is not neutral. It is a machine with a point of view, and that point of view is written in its training data, in its makers' habits, in whatever it has seen many times before.

Why the transfer window is the cradle of wrong labels

Football has many subjects. Match results are the easiest to classify: there is a score, two teams, a date. Injuries are fairly clear. Tactics are vaguer, but there is still a line-up to anchor to. The transfer window, by contrast, is almost entirely vague.

A typical transfer item has no score, no starting eleven, no referee's report. It has a player, an interested club, a owning club, an agent, and a gap in the middle called "negotiations". That gap is where football's language becomes loosest. The same sentence — "talks are ongoing" — can mean the two sides met once, or can mean the agent called and nobody picked up.

Because the language is loose, keywords become useless. The aggregator sees "negotiations" and "club" and labels it "transfer progressing". It does not know that the writer carefully noted in paragraph seven that the club has not responded. The caveat sits in paragraph seven; the keyword sits in the headline. And the headline is the only part that travels.

The transfer market is only a mirror — look into it and you see a club's fear. A club afraid of relegation attracts the label "hunting a striker" more easily than a club winning titles. A player nearing the end of his contract carries the label "parting ways" about three months before he does. The label arrives first; the truth arrives later, or never.

The four layers of a piece of information, and which ones get dropped

In my trade, a decent piece of football information passes through four layers.

The first is observation. Someone sees something real: a player arriving at another club's training ground, a lawyer entering a club building, a player left out of a registration list.

The second is confirmation. Someone tries to verify what they saw through a second source and, more importantly, through a second motive.

The third is explanation. Someone can answer why this is happening now, and who benefits if it happens.

The fourth is transmission. Someone writes it in language an ordinary reader understands, preserving the three layers above, and adds nothing and removes nothing under the pressure of speed.

Most transfer items readers encounter contain only layers one and four. Observation and transmission. The confirmation layer is skipped because it takes time. The explanation layer is skipped because it demands knowledge and sometimes contradicts what readers want to hear.

An automated aggregator can only work at layers one and four. It can count words, apply labels, distribute. It cannot test motive. It cannot explain interest. So when a large platform labels and distributes at the scale of millions, it is distributing a version of football stripped of its two most important layers.

Readers do not know this. They believe they are reading copy that passed editorial checking, when in fact they are reading copy that passed classification. Checking and classifying are two different jobs. Checking asks: is this true. Classifying asks: where does this belong. Those two questions cannot substitute for each other.

The linguistics of a transfer item

There is an exercise I run with young editors: hand them a transfer paragraph and ask them to underline every word expressing certainty. The result is almost always the same. Of ten sentences, seven contain phrases like "reportedly", "could", "is considering", "has been linked", "likely". The remaining three describe general context nobody can verify.

A passage like that has a strange property: it is not wrong in any single sentence, yet it is wrong as a whole. Every verb is insured by a softening word. The writer has protected himself perfectly. But the reader, finishing it, remembers one thing only: that player is on his way out.

This is the mechanism I call "right in every sentence, wrong as a whole". And it is the most fertile ground for the classification layer, because that layer does not read softening words. It reads keywords. It sees "player", "club", "transfer" and labels. Words like "could" and "reportedly" are treated as noise — compressed, discarded.

The result is a paradox I meet almost weekly: the most careful article is the most misunderstood. The more cautious the writer, the more hedges he uses, the more the system flattens him into a hard label.

The agent as a label amplifier

No discussion of the transfer window is complete without agents. I know a few of them in Kuala Lumpur and Bangkok. Most work seriously. They have families, players to feed, contracts to sign. But their work, by its nature, is to generate labels.

A good agent understands he is not selling a player to a club, but selling a player to a market. That market is hundreds of shutters, thousands of eyes, and a continuous current called rumour. To lift his player above the surface, he must plant a label. The label might be "drawing interest from several clubs". It might be "has not renewed". It might be "has been in contact with a club abroad".

These labels are not technically false. They are merely incomplete. And the classification layer, hungry for keywords, swallows them, multiplies them, and spits them back as something that looks real.

One thing I always keep in mind reading such items: who benefits if this story spreads. If the answer is the agent, I read on but hold back. If the answer is a club wanting to sell, I read on but count again. If the answer is nobody but a few sites needing clicks, I close it.

Big clubs do not wait for luck — they prepare for the silence before the storm. The larger clubs in Southeast Asia, those with data-analysis departments and in-house agents, almost never reveal their business through rumour channels. The loudest deals are usually the ones the selling side needs to be loud. Noise is a negotiating tool.

Once I asked an executive at a Malaysian Super League club why his side did not publicly deny a false rumour. He laughed and said denying a rumour is the surest way to keep it alive for two more cycles. Silence is a calculated choice. And in that silence, the label keeps flowing.

The cost of a wrong label

If I point out one newspaper mislabelling a story, the story is small. If I point out a system mislabelling for millions of readers, the story is much larger. The issue is scale.

When a local Vietnamese outlet publishes a wrong transfer item, readers can cross-check three other outlets and spot the difference. When a platform labels an item wrongly, readers have nothing to cross-check against, because the label has already told them this item belongs here. The frame guarantees its own validity.

I think of young players. A wrong label kills no one, but it leaves very thin scars in the mind of a person still growing. A boy reads that he is "about to be sold" when nobody has asked him anything. Another reads that he is "negotiating with a club in Thailand" when he merely asked a friend for directions at an airport. Those labels do not come from malice. They come from a machine unable to distinguish "a name that was mentioned" from "a person who was bought".

During the 2026 pandemic, when competitions stopped and Selangor lost its main sponsor, I lived alongside the club as a board member for three months without football. I recorded training sessions without a ball, players jogging around a pitch full of weeds. I raised fifteen thousand ringgit to pay the wages of cleaners and stadium security guards, people who never appear in the papers. In those three months I understood that football is not the names on a news ticker. Football is a system of people breathing together, and that system only withstands pressure when the information inside it is treated with decency.

A season is a symphony — the beat keeper knows when to strike and when to stay silent. A team's beat keeper is not its most famous player. It is usually the one who understands best what is happening, and has the discipline not to carry it somewhere it should not be carried. The same logic applies to journalists. The beat keeper of information knows that a correct label is worth more than a fast one.

The Aiman Zakri case: the right person, nearly filed in the wrong place

Let me recount a time I nearly made a mistake, to show how the classification layer affects writers themselves.

In 2026 I spotted a seventeen-year-old at the Selangor training ground. Slight, playing deep, without a place in the squad list. I watched him for three sessions. I wrote a 1,500-word analysis of his positioning, and posted a short clip shot on my phone.

After it ran, several other sites picked it up and labelled it "young talent discovered". It sounded good. But the label dragged an effect I had not foreseen: within days, three different clubs made enquiries about him, and the Selangor coaching staff began worrying he would lose focus. Someone asked me whether writing that piece was a way of "driving up his price", when all I had wanted was to record what I saw.

My lesson was not to stop writing about young players. It was this: when you spot something still unformed, you must weigh carefully the label you attach to it, because that label will go out into the world before the player himself understands what is happening to him. I look at the feet, but I listen to that kid's heart. And the heart of a seventeen-year-old is not built to withstand a thousand labels pasted on by others in three days.

Had I filed him in the wrong drawer then — say, tagging him "player in negotiations abroad" when he had merely trained well on a morning nobody watched — the story might have gone another way. He might no longer be riding that forty-minute bus to training each morning with a head full of unanswerable questions.

A cross-border lens

There is another dimension outsiders often miss. Southeast Asian football runs to its own rhythm, and that rhythm is not always reflected accurately in the international feed.

In Europe a transfer story usually comes out of an almost fully professionalised agency system, with release clauses written clearly into contracts, and a media apparatus strong enough to verify. In Southeast Asia the same story may come from a conversation in a coffee shop, a hint from a second broker, or a social-media status.

When the international feed takes a story from this region and labels it to a European standard, it produces a particular error: an item with local reliability presented with international-grade authority. I once held a printout from an international outlet describing a Malaysian player, containing a detail about a signing ceremony that never happened. The detail was written with such confidence that anyone without local context could not doubt it. I once held newsprint in my hands; now I hold a phone, but the pulse of the match has not changed.

That is why I always tell young editors: when you take a story about a player from a football culture you do not live in, treat every label as a hypothesis to be tested, not as an established fact. The label never travels with its source. It only travels with the reader.

Born in Vietnam and working in Malaysia, I have come to see that the two football cultures touch the same ball with two different temperaments. Vietnamese fans tend to read transfer news with the emotion of attachment to a shirt, so a wrong label hurts them as if they themselves had been sold. Malaysian fans tend to read it with the calculating eye of a club's interest, so a wrong label angers them because they feel their intelligence has been insulted. Two different reactions, one cause: a label nobody checked.

The misunderstanding from outside

Outsiders look at the transfer window and say the problem is fake news. Fake news must be stamped out.

I think that view is right but insufficient, and because it is insufficient it leads to the wrong remedies.

If the problem is fake news, we build a fact-checking unit. But most of what flows through the pipeline is not fake. It is true information severed from context, old information reborn as new, incomplete information presented as complete. A fact-checker cannot catch those, because they do not lie at the level of the sentence. They lie at the level of the frame.

A second misunderstanding is more common: people think a label is neutral. That classification is just arrangement, like shelving books. But a bookshelf is not built by the person reading the books. It is built by the person deciding this volume belongs here, and that person has reasons of his own.

When a platform files a music bulletin in the football drawer, the damage is not that a reader is mildly annoyed at reading the wrong thing. The damage is that the drawer becomes a little less trustworthy, and readers learn that sometimes the football drawer contains things that are not football. Trust erodes that way, one thin layer at a time, with no single event to point to.

The interesting part is that most proposed solutions aim at the people who write: write more carefully, verify more, take more responsibility. All true, and I do them daily. But the labelling layer is never touched. This is where someone who has spent years in a newsroom feels helpless: you can write a hundred correct pieces, and the system can still bury them under a thousand wrong labels, because labels travel faster than articles.

One line I often think when I see a distorted headline: if readers cannot tell which of the four layers they are looking at, the responsibility belongs to whoever built the pipeline, not to whoever endures the output. Readers are under no obligation to do the work of a classification system. They already have another job: living their lives.

There is a third misunderstanding, subtler and rarely discussed. People believe technology will fix itself. That when a model is big enough, it will stop being wrong. But a model wrong at the level of the frame is wrong systematically, and every time it is wrong, it is confirmed by the very readership it lured into the wrong drawer. Page views become evidence for an error. That is the closed loop no keyword filter can break.

What I saw in the corridor

Back to 2:14 a.m.

I could delete that item. It does not matter. But I chose to pin it, screenshot it, save it into a folder I call "wrong labels". That folder already holds more than four hundred screenshots accumulated over four years. Among them are items about a basketball player labelled football because of the shared word "transfer". Items about a national marathon labelled football because of the shared word "event". Items about a music nomination, like tonight's.

I do not collect them to accuse anyone. I collect them the way a beat keeper collects beats. Every wrong label is a misplaced beat. A single misplaced beat does not break a piece of music. But a beat keeper understands that misplaced beats tend to gather.

I do not think I am doing data hygiene. I think I am keeping a drawer trustworthy. When a reader opens the football drawer, they should find football. It sounds trivial, but to achieve it at the scale of thousands of articles a day, you need more than a keyword filter. You need a checking layer that understands football is not defined by the words it uses but by the entities it names: a club, a player, a coach, a match, a table.

I have seen that in a place nobody thinks about. At the 2026 World Cup semi-final in Saint Petersburg, the only goal came from a corner that had been rehearsed for ten minutes before kick-off. After the final whistle I wandered the stadium grounds and picked up a note someone had dropped, sketching the opponent's defensive shape. That slip reminded me that what decides outcomes usually sits where no one is filming. Labels are the same. They are made very far from the pitch, and they shape the pitch in ways nobody sees directly.

Takeaway

What I believe will happen over the next few seasons: as the volume of football information passes the threshold where humans can sort it themselves, readers will begin judging platforms by the quality of the drawer rather than the quality of individual articles. Whoever keeps the drawer clean wins. Whoever lets impurities in will lose trust, and that trust is lost slowly, quietly, until there is nothing left to lose.

As for my trade, I will keep arriving two and a half hours early, standing in the corridor, watching players walk past and noting the order of their footsteps. I do that because it is the only data I can verify with my own eyes. What I write afterwards may be mislabelled by a system. I cannot control that. I can only control whether I show up in the corridor.

No algorithm stood beside me in that corridor at that moment. And I hope no classification layer will ever dare tell me which drawer the thing I just saw belongs in.