When a Fashion Obituary Was Labelled as Football

In a small apartment in Shanghai, my screen lit up before the city had woken. A data line ran through the internal news aggregation system: “Bob Mackie, costume designer, dies at 87.” Right beside it, the classification label read clearly: football. I sat still for a few seconds. Bob Mackie never wore a club shirt. He never stood on a touchline, never signed a contract, never walked into a post-match press conference. He was the man who designed costumes for Cher, Carol Burnett, Tina Turner, Diana Ross. He won nine Emmy Awards, received more than thirty nominations, was nominated for three Academy Awards, and was inducted into the Television Academy Hall of Fame. And yet some machine decided that the story of his life belonged to football. In that moment, I understood that the problem was not Bob Mackie. The problem was us. Twenty-eight years of watching this industry taught me one thing: every mistake in this profession starts somewhere very small. A misplaced comma in a contract. A name misprinted on a scoreboard. A misaligned label at the data layer, where nobody looks. And from those small places, an entire large story gets built, and millions of people believe it. Bob Mackie died, and the news was announced through his own Instagram account. That is the most credible origin an obituary can have. But when the report passed through the automated classification pipeline, it was assigned to a field it had never touched. This story is not alone. It is an alarm bell for an entire football-media ecosystem running on automated data, an ecosystem that I, and many colleagues in Vietnam, China and Europe, live inside every day. I remember afternoons at a club training ground where I spent years following the team. The assistant coach held a printed sheet full of metrics, but he did not read a single line. He just watched the players run. He told me: “The data tells me how many kilometres he ran. My eyes tell me whether he still believes in himself.” That sentence has followed me ever since. It explains why I never write an article made only of numbers. And it also explains why the story about Bob Mackie made me stop. A data system with no human eye checking it is a system running on thin ice. It can be right a thousand times in a row. But on the thousand-and-first time, it can be wrong in a way nobody expects, like labelling the obituary of a costume designer as “football”. I once heard a friend who works in data engineering in Shanghai say: in any information pipeline, the most dangerous error is not the one that crashes the system. The most dangerous error is the one where the system keeps running, keeps returning results, keeps looking correct. A silent error. An error with no alarm. An error only a human can catch, and only if a human bothers to look. The record about Bob Mackie is one of those silent errors. No fire alarm rang. No exception was thrown. The data line simply drifted on, carrying a wrong label, ready to blend into the enormous stream of information that every football fan absorbs each day. When I analysed this record, eight of the nine football analysis dimensions had to be left blank. Tactics, club finance, the transfer market, league positioning, rules and governance, the dressing room, the risk profile, the industry transmission chain, all had no input data. Not because information was missing. Because the information belonged to an entirely different field. This is what I want you to picture clearly: when a record is mislabelled, it does not just ruin one article. It ruins a process. It forces the analyst to say “insufficient basis” on almost every front, while the real problem lies somewhere else entirely. If this wrong label slipped into a football database used to compute metrics, to rank, to forecast, the consequences would not stop at a single article. It would seep into the numbers experts use to assess players. It would seep into the models analysts use to price value. And at some point, it would seep into the faith of the fans themselves. The second problem this record exposes is subtler. Among the biographical facts about Bob Mackie, most carry no source of verification. The Emmy count, the Oscar nominations, the film credits, all are externally verifiable, but the record itself does not mark them as verified. Only the death event is tied to a clear primary source: his own Instagram account. This asymmetry is striking. A record can have accurate facts at the data level, yet still lack a structure of trust at the presentation level. In my profession, that means: an article can be accurate in its figures and still untrustworthy in its structure. And credibility, in the end, is not about how many times you are right. It is about how far a reader can check you. I learned this lesson painfully at a World Cup. In the tunnel after a come-from-behind defeat, I sat beside a captain with red-rimmed eyes. His team had just lost after leading, three goals conceded in the final fourteen minutes. His passing accuracy in that match reached ninety-two percent. But ninety-two percent says nothing about the way he sat there. I rewrote that story, not with metrics, but with breath. A colleague called my writing weak. But within hours, thousands of fans shared it. They shared it because they needed someone who understood their pain, not another statistics table. That lesson applies directly to today's story. When I saw the “football” label on the obituary of a costume designer, I did not see a mere technical error. I saw a system that had forgotten that behind every data line is a human being. Bob Mackie is not a player. But he is an eighty-seven-year-old man who spent a lifetime making famous people beautiful, and when he died, his story deserved to be told in the right place. Being labelled “football” is a silent insult: it says the system could not be bothered to understand who he was. I stand in the crowd and understand that tears are also a language of love. But data, if not illuminated by a human being, will never read that language. It only knows how to label and push on. In Vietnam, football news platforms are also shifting rapidly toward automated aggregation. Every day, thousands of reports are generated, aggregated, classified and pushed to readers. I do not oppose that change. I understand why it is needed: fans want information fast, complete and free. But speed and reliability are not opposites. They only become opposites when we are lazy. And that laziness, in the long run, will destroy the very thing we are trying to build: the reader's trust. On a trip between Shanghai and Hanoi earlier this year, I sat beside a young editor. He told me that every day he has to handle hundreds of automated reports, and there is never enough time to read them all. He said: “I just glance at the headline, and if there's a familiar player's name, I push it up.” I understand him. But I also understand that it is precisely that moment, the “glancing at the headline” moment, where silent errors are born. What troubles me most in analysing this record is speed. The news of Bob Mackie's death was announced on a Monday. It is the kind of story with high news value, to be handled immediately, impossible to defer. And that very pressure to process immediately is fertile ground for silent errors. In the sports industry, we live inside a news cycle measured in minutes. A match ends, three minutes later there is an article. A transfer deal breaks, ten minutes later there are hundreds of lines. That speed gives no one time to pause and ask: does this data line really belong here? I once witnessed a January morning in Shanghai, when a fake transfer story spread in just twenty minutes. A social media account posted a photo of a player signing papers. The photo was shared thousands of times. By the time someone noticed the photo was three years old, from a completely different club, the fake news had gone everywhere. Speed had beaten verification. And the fans' trust was eroded a little more. That is why I always stay in the dressing room longer than my colleagues. Not to hunt exclusives, but to listen. To verify across sources. Before I write, I ask the player, I ask the coach, I ask the team doctor. Three sources. Three angles. One truth. I learned to verify emotion and fact alike with a triangle of sources, because I once wrote sensational pieces and saw them wound other people. A team has more than victories; it has its own heartbeat, and I am the one who records it. But to record the heartbeat correctly, I must be sure I am placing the stethoscope on the right chest. The Bob Mackie story is a reminder that in the data age, the sports writer no longer faces only the pitch. They face the pipeline. They face the algorithm. They face classification decisions made by machines that do not know a player exists, let alone distinguish a player from a costume designer. And as I looked deeper, I realised this is not a story about a bad algorithm. It is a story about people who built a system where no one is responsible for checking the output. A misclassification is not the scariest thing. The scariest thing is that no one discovers it until someone, like me, happens to pause and ask: why is Bob Mackie here? On empty-stadium days, I hear the team's breathing clearly, and it is still beating. But on days when data pours in torrents, that breathing can be drowned by the noise of thousands of lines nobody reads carefully. There is a counterintuitive reading of this story. Most people will blame artificial intelligence, the algorithm, the laziness of machines. I believe that reading misses the point. The algorithm does not understand football, does not understand fashion, understands nothing, and it was born that way. The real fault lies in our faith: we have implicitly agreed that a label assigned by a machine is as trustworthy as one assigned by a human. We saved time at exactly the place where saving time is not allowed. From another angle, this incident exposes a paradox of modern football media. We demand ever more detailed data, running metrics, heat maps, predictive models, yet we check less and less whether that data is correctly labelled. We crave depth at the top while leaving the foundation to rot at the bottom. The taller an analytical building, the stronger its foundation must be. The classification label is that foundation. And a third paradox: this error may never be fixed if no one looks. Silent errors do not raise alarms. If a wrong record enters a database and no one detects it, it will sit there, quietly skewing the conclusions built on top of it. One day, an expert will use a number, believe it, and form a judgment about a player, while that number was born from a record that was never checked. I do not write this article to conclude. I write to open a question for those who run football media: in your data pipeline, who is responsible for the labels? And if the answer is “no one”, then you are running a system where even Bob Mackie can become a player.