When a Traffic Ruling Landed in the Football Section

The data file landed on my desk at 22:40, labelled "Football". I opened it the way I open a replay of a controversial incident. Eighteen information points. Not a single player's name. Not a single formation. Not a single minute of play. Instead: the SCJN, Mexico's Supreme Court of Justice of the Nation, photo-enforcement tickets in Mexico City, the joint economic liability of vehicle owners, and the voting record of justices Hugo Aguilar Ortiz and Giovanni Figueroa Mejía. The final recommendation to drivers: respect the speed cameras. In years of working with match data, I have seen plenty of strange files. This is the first time a ruling on traffic cameras has worn the costume of football news. My first reflex, as always when facing a difficult decision, was to stop. Before blowing the whistle, I review myself. CONTEXT: A CONSTITUTIONAL COURT IN THE SPORTS SECTION In substance, the file belongs to a constitutional court. The SCJN Pleno rejected the claim that Mexico City's photo-ticket system is unconstitutional. The court's president argued that the camera system helps resolve corruption problems in traffic enforcement. Some justices dissented. The system stands. There is no football in it. But there is a subject anyone who analyses football through data must care about: surveillance technology, the verifiability of visual evidence, and the question of who is accountable when a system delivers a wrong ruling. I spent most of my career in that room. The VAR room at San Siro, November 2026. Higuaín scored to make it 2-0 for Juventus against Milan. The feed showed him 0.2 metres offside. I hesitated. I was afraid of being wrong. I did not recommend a review. Milan lost 0-2, and afterwards the referee supervisor criticised me in front of the whole team. The pain was not in the scoreline. It was in having the evidence in my hands and not daring to use it. I spent a month reviewing 47 similar incidents, then built a 37-point checklist to standardise decisions. Since then, every conclusion of mine goes through a process, not through a feeling. That is why the file labelled "Football" bothered me more than a defeat. ANALYSIS: THE ERROR IS STRUCTURAL, NOT LEXICAL Look at how the error forms. An automated system collects text, extracts entities, then assigns topic labels. The keyword "camera" appears. The keyword "surveillance" appears. The phrase "review again" appears. In the vocabulary of a football classifier, all three carry heavy weight, because they attach to VAR, to goal-line technology, to replay reviews. This is the class of error I call context-driven false positive. The algorithm is not wrong about vocabulary. It is wrong about narrative structure. A text about a constitutional court and a text about VAR share a set of nouns, but they operate on two different logics: one concerns the relationship between state and citizen, the other the relationship between the laws of the game and its players. Modern football analytics runs on automated ingestion. Statistical platforms scrape hundreds of thousands of articles a day to update line-ups, injuries, form and transfer news. If the mislabelling rate crosses two percent, prediction models begin to drift. They do not collapse. They leak: a few indices get pulled toward events that never happened on a pitch. I have seen it at a smaller scale. Our data room once ingested a batch of friendlies as competitive matches, and one Serie A club's counter-attacking defensive index was pushed above its true level. It took three weeks to catch. For three weeks, every report sent to the coaching staff stood on a leaning foundation. With the Mexico file, direct damage is zero. No club is affected. But the signal is clear. If a constitutional ruling reaches the football section, then a criminal verdict, a tax notice, a customs regulation can reach it too. At the scale of millions of documents, leakage stops being an accident. It becomes a trend. I do not trust my eyes, I trust the slow-motion replay. And the slow-motion replay here shows something simple: every ruling deserves a review, including the ruling of the data. THE CONTRARIAN ANGLE: THE CHECKER, NOT THE ALGORITHM The natural reaction is to blame the algorithm. I do not. The algorithm only does what it was taught: find patterns. The people who taught it to find patterns, and the people who check its output, are the missing link. In a VAR room, no incident passes without human confirmation. There is a referee, a VAR assistant, an assistant referee. Three layers. Not because three people see better than one, but because the system needs a stopping point before a decision becomes an outcome. Sports data mostly has no such stopping point. It flows from scraper to model, from model to report, from report to a transfer decision or a tactical adjustment. Nobody presses confirm. Nobody signs off. One detail made me pause longer. The SCJN president said cameras resolve corruption problems. I have heard the same sentence in football: VAR resolves refereeing problems. Years after VAR arrived, the number of controversies has not fallen. It has merely relocated. Technology does not kill football, it kills blind faith — the faith that a device can replace a judgement. Traffic cameras and VAR cameras share the same limit. They record the truth, but they do not decide it. The person who decides is still in the room, still hesitating, still capable of being afraid of being wrong. The difference between a good system and a bad one lies in whether that hesitation is acknowledged. TAKEAWAY: A CHECKLIST FOR DATA What I want to take from this file is not a warning about artificial intelligence. It is a suggestion for how football should build its own verification layer: a 37-point checklist, applied to data instead of to incidents. Football is a game of errors, but the winner is the one who knows which errors are worth making. A mislabelled file is not worth making, because it costs nothing to fix and a great deal to ignore. And one question I leave with the people who run statistical systems: if you cannot point to the person who confirmed a figure before it entered your report, are you analysing football, or are you analysing a mislabelled file?