The "Football" Label and a File Without a Ball

On Tuesday morning, among hundreds of files landing in my "football" folder, one made my hand stop. I opened it the way I open every other file — waiting for a match, an expected-goals chart, a back line slowly cracking minute by minute. Instead there was a name. A twenty-one-year-old woman. A small town called Mãe do Rio, in the state of Pará, northern Brazil. And a line so short it felt cruel: she was shot dead in front of her father.

I read it slowly. The recorded date of death was September 16. Days earlier, she had left custody after a provisional release granted on September 9. The social media account with roughly twenty-four thousand followers had disappeared from the platform. At the scene, investigators noted the letters "TCP," known as the initials of a Brazilian criminal organisation. The Civil Police of Pará opened an investigation. She had never been convicted in the earlier case. And the authorities, as of now, had confirmed no one's responsibility for her death.

Then I looked at the corner of the file. There, in a cold typeface, sat a tag: football.

That was the moment I understood I was holding something that did not belong in this folder. A file about a person's death had been misnamed by some classification engine, somewhere. "Every anonymous match inside me is a verse still waiting to be sung." But this was not a match. And my profession, right now, obliges me to say so before anyone turns it into something else.

Every sports desk today is a vast data-routing system. Every minute, thousands of items pour in from hundreds of sources: club statements, players' posts, reporters' dispatches, federation notices, and the aggregated wire copy of international agencies. No one reads all of it with human eyes. The machine sorts first. A person checks afterwards.

The sorting machine works in a fairly simple way: it relies on keywords and surrounding context. It sees a name that looks like a player's, a number that looks like an athlete's age, a place name that looks like a venue, and it applies a label. Most of the time, that works smoothly. But the small remainder — the part I was holding this morning — is where the system fails.

I have covered football in Europe, the Middle East, East Asia and Southeast Asia. I have sat in packed press rooms and in stands so empty I could hear wind slipping between plastic seats. Years in this trade taught me something that training programmes rarely say out loud: a sports editor today does not only need to understand football. We need to recognise when something with no connection to football is wearing a football shirt to walk into our folder.

Why does this matter more than it appears? Because dirty data does not stay still. It spreads. A mislabelled item enters a search index, a topic model, an archive that another editor will query at three in the morning while writing about a match in South America. And if no one stops it, it becomes part of collective memory — the memory I still call the memory stand.

If this file were analysed with the proper toolkit of my trade, the result would be systematically empty. There is no line-up to read. No tactical diagram to draw. No possession figure, no pass count, no expected goals. Every column of the analysis table I normally use sits blank, and it sits blank for the right reason: this source contains not one unit of football knowledge. I could invent a back four out of the word "Pará," invent a rivalry out of the letters "TCP," invent a contract out of the figure of twenty-four thousand followers. But every invented word would be a betrayal of my own readers.

What I can analyse — and must analyse correctly — is the structure of the story and the outlines it traces.

First, the timeline. The young woman left custody after a provisional release on September 9. On September 16, she died. The gap between those two points is one week. In investigative circles, such a short interval between release and killing is usually treated as a priority line of enquiry — retaliation, the silencing of a witness, or the settling of some debt. I say "usually treated as" rather than "is," because I do not know. The source does not say either. And the gap between what I infer and what I know is exactly where every error in this craft begins.

Second, the social media accounts. They vanished from the platform before she died. No cause is given. It could have been a voluntary withdrawal from public view while legal proceedings were running. It could be something else. A careful writer records both possibilities and does not pre-select a conclusion.

Third, the two letters at the scene. "TCP" was noted, and immediately beside it, in the same passage, sits the statement that authorities have not confirmed anyone's responsibility. I have stressed this once, and I will stress it again: keeping those two sentences side by side is a professional duty, not a stylistic choice. Pulling them apart — letting a summary keep the letters and drop the denial — is how a news item becomes an accusation.

Fourth, the person behind every line. She is recorded as having been investigated in connection with an earlier case. In the same breath, the source states clearly that she was never convicted. I have watched crowds turn an investigated person into a guilty one with a few shares. "The day I was rejected for being a girl, I understood that a ball never reads paperwork." People are different. People read each other's paperwork very carefully, and they often read it wrong.

If this story were drawn as a line, the shape would be familiar: a period of near-anonymity with twenty-four thousand followers, a jolt that pushed her onto the front pages, a silence as the accounts vanished, then a violent event that carried the story beyond Brazil's borders. This is the shape of an attention cycle. I recognise it because I have seen that shape elsewhere — except that elsewhere it usually ends in a defeat, and never before in a death.

The attention this story generates is far larger than the volume of verified information behind it. We have an abbreviation, an association, an open investigation. We do not have a motive, an arrest, a conclusion. The distance between heat and factual foundation is the most fertile ground there is for false information. A sports editor, whether he wishes it or not, stands on that ground every day, because the system he uses is blending this kind of story with that kind of story.

What my industry rarely admits is that a wrong label is not a small technical glitch to fix when there is spare time. It is a diagnostic signal. When a file about the death of a twenty-one-year-old woman can slip into the "football" folder, the right question is not what happened to this particular file, but how many other files went down the wrong path before I could open them.

The honest answer is that I do not know. And that not-knowing is the problem. If the classifier runs on keywords, then a criminal term that shares a sound with a football term may be enough to trigger a false label. Such an error rarely travels alone. It usually travels in batches. Which means this morning's file may be a symptom of a wide-scale misclassification, not an isolated accident.

Sports media tends to have two reactions to matters like this. The first is to ignore it — to call it a technical department problem, not a writer's problem. The second is to absorb it — to seize the heat of the case and push it into sports articles for clicks, turning a family's tragedy into bait. Both are wrong, but the second is more dangerous, because it looks like work.

I once heard a colleague say that sports news and crime news are separated by a single word: "supporter." In some countries, organised supporter groups have at times been documented in connection with organised crime. That is real background knowledge, and it makes it easy to leap from a criminal case into a football story. But in the file I was holding, not a single line connects this case to football. No stand, no supporters' group, no club. The bridge between the two banks exists only in the mind of whoever wants to cross it.

And there is one more thing the craft must say out loud. How we report on the dead matters as much as how we report on the perpetrator. When we recount that a person was once under investigation, we must also recount that the person was never convicted. When we recount that a symbol was present at the scene, we must also recount that authorities have not confirmed it. Those two sentences do not weaken the story. They make it correct.

I left the file on my desk. I did not delete it. I flagged it, reported it to the classification team, and wrote these lines. "A contract is not signed in ink, but in the memory of an entire stadium." That memory is built from what we choose to record and what we choose to remove. If I removed a woman's death because it scored no goals, I would have struck a line from my industry's own ledger with my own hand. If I turned that death into a sports headline, I would strike that line in another way.

What I can do, and what I believe this industry will have to do more of in the years ahead, is keep the boundary clean. Read closely before labelling. Verify the person before naming them. And remember that behind every file is a human being — twenty-one years old, a father, a small town in Pará, a social platform that has gone dark.

In a world where machines grow faster than we do, slowing down by a single beat before classification may be the highest form of respect we have left.