No Football Inside the Football File: An Audit of One Wrong Label and a Ledger of 736 Names

On Monday I opened the intake queue. The third file carried a domain label at the top — Football. Eighteen years of opening this queue have trained the hand: team first, then player, then competition. That morning none of the three were present. The file contained a French actor who was also a theatre director and a playwright — Arnaud Denis, aged 43. His death took place in Belgium, by euthanasia. Before it came hernia surgery, then a polypropylene implant, then a long physical collapse. The theatre ledger holds an award, Molière nominations, and a 2026 role as Jean Moulin. It also holds a date I flagged immediately with red ink: 22 September 2026. There is no football — no club, no player, no match, no transfer, no governing body.

The label is wrong. In my trade a wrong label is not merely a wrong word; it is a wrong birth year, and it poisons the entire growth curve.

The story of how the ledger began matters here, because the test I am applying came from it. In 2026, having finished a civil-engineering degree, I moved into sports journalism, and one habit formed then and never left: source before claim, date before conclusion. In October 2026 I crossed from Rajshahi to Kolkata for the FIFA U-17 World Cup — twelve matches in nine days, notebook first, phone second. By December my ledger held 214 South Asian players born between 2026 and 2026, each with a club, a position and the exact match in which I first saw him.

At Russia 2026 I was not a spectator but an auditor. I logged the pre-20 international record of all 736 players across 32 squads and re-checked every entry against federation archives rather than highlight reels. That audit returned 61 percent who had appeared in a youth tournament before turning 20. In August I filed a 4,000-word report read mostly by coaches, not readers, which suited me.

On 9 February 2026 Bangladesh won the ICC U-19 World Cup in Potchefstroom; I opened a longitudinal file on all 15 squad members that same night. Six weeks later the stadiums emptied. In that gap I audited Rajshahi Division age-group records and found 40 percent lacked primary birth documentation. I verified 1,180 players by hand. In July 2026 I counted Pedri's load and wrote a two-page memo — 52 Barcelona appearances, six Euro matches, 629 tournament minutes — and he missed roughly 30 matches through hamstring injury the following season. The memo was filed and ignored.

I raise this history for one reason. A ledger's real strength lies not in the writing but in the intake filter. Once a wrong file gets inside, every sentence in it begins speaking the language of conclusions while wearing the clothes of analysis. So today's task is not football analysis. Today's task is returning that file to the label it deserves, and opening up the arithmetic of why.

My audit rule has three tiers: what the document shows, what the document implies, and what remains unproven. That division is not laziness; it is protection. Twenty years of habit teaches pattern-matching, and pattern-matching readily mistakes itself for proof.

Tier one — what the document shows. Name: Arnaud Denis. Age: 43. Profession: actor, director, playwright. Training: the National Superior Conservatory of Dramatic Art in Paris, one of the most institutional stages of French theatre education. Checkable career items: the Brigadier Prize, Molière nominations, a 2026 role as Jean Moulin. Manner of death: euthanasia in Belgium, and it was not sudden — a lawyer and a close friend were present during the procedure, and French media confirmed the event. The preceding medical line: hernia surgery, polypropylene implant, then severe disability. Grief across the theatre world, and a collective speaking out about implant adverse effects.

Stop at this tier and the piece is an obituary, plausibly tagged society or media, with zero relation to football analysis. This is not a minor clerical slip. A classification error is the most expensive error in record-keeping, because it errs once and then forces every downstream step to err with it.

Tier two — what it implies. The causal chain points at a specific product: a prosthetic implant. Two plausible storylines follow — a regulatory discussion on medical-device safety, and advocacy-driven follow-up coverage. Both are legitimate journalistic expectations. But implication and proof are different objects, and my job is not to seat implication in proof's chair.

Tier three — what remains unproven. The causal chain rests largely on Denis's own account and the testimony of those close to him, not on independent medical confirmation. The document names no separate medical record. It therefore stays flagged as an attributed claim, not an established fact.

The second unproven point is the date. 22 September 2026 does not fit the ordinary chronology of a retrospective obituary. It may be a transcription error, a misplaced figure, or a publication that has not yet occurred. Until that is settled, not one sentence from this file may be quoted downstream.

The third issue is source ratio. Across 51 information points, a large number carry the tag Source: None. That is not a small statistic; it is the signature of a defect. In my 2026 Rajshahi audit, 40 percent of records lacked primary birth documentation — the same flaw returning at the level of sourcing. A document that will not name its own source is not testimony; it is memory, and memory never becomes an archive.

My intake check has four columns: team, player, competition, date. If the first is empty the file goes to the doubtful tray. If the second is empty the file stops, because football analysis without a player is impossible. Here all four are empty. I could have treated that as routine error and moved on, but the Kolkata ledger taught me otherwise: the real problem is not the empty cell, it is presenting an empty cell as a full one.

I have written many times that xG is already being abused, because the metric cannot explain in-game decisions, player form or refereeing standards, yet it gets deployed as explanation anyway. The same thing happens with labels. When a file opens under the Football label, the downstream structure already expects tactics, formations, positional roles, transfer fees, financial compliance. The material is absent. Two paths open: reject the file and correct the label, or manufacture material to fill the template. The second path is cheaper, faster and entirely fake.

In youth football that fake has a familiar name: adding names to make the ledger look full. From outside everything looks right — totals complete, report smooth, signatures collected. Inside, the curve is false. In journalism the same act is called forcing the frame. The difference is only this: in football a false birth year surfaces in a federation archive, and here the only place it surfaces is the source line.

From my years of watching matches I will offer one observation. Many matches I have watched from empty stands — age-group qualifiers where nobody but the scorer is meant to be present. The empty stadium is the most honest witness in youth football, because instead of attendance it lets only the documents speak. The same logic holds here. This article has no gallery, no crowd appeal, but it has documents — and the documents are the only witness.

Another lesson came from the Kolkata ledger. By 2026 it was clear that words like footballer, midfielder and promising have no place in a youth description unless a birth year sits beside them. I do not scout highlights; I excavate birth years. That same severity now applies to subject classification: where footballer is written, I want the documentary equivalent of a birth year — club, position, competition, match ID. Without any one of those, the word is only a word.

I will admit a weakness here. My background is British and my training was shaped by the English football model, and that is a permanent trap — making one's own house the benchmark. The temptation existed with this file too: forcing material into an analytical frame. Measuring French theatre practice against the British tradition would be equally unfair. In every piece the local game's own logic goes first; comparison arrives later, as one variable, never as the yardstick.

Now the part where everyone stops and I keep suspecting. Everyone will say the fault is the label, and fixing the label ends the matter. I say the label error is the easy failure, and being easy it is the lesser danger, because it surfaces at intake — in the first three lines. The real danger is a correctly labelled file in which more than half the facts carry no source. That file enters the system silently, dressed as analysis, and ends up speaking in the language of decisions. The first needs a correction. The second has no cure unless the source ratio was counted from the start.

The second uncomfortable point is template compulsion. We take procedural pride in saying one structure fits every subject. In practice the structure is not a detective but a mould, and a mould's instinct is to fill empty cells. An analysis that cannot admit a shortage of material eventually learns to manufacture material.

Third, a protective note. The subject is euthanasia and a medical device — both highly sensitive. Building a football case study out of this would be wrong, and turning it into an advocacy piece would be equally wrong. Denis's death is testimony; making policy out of testimony is not the journalist's job, though the policy question can be placed on record without delivering a speech.

Finally, I set an exit criterion in advance, because attachment to an archive makes people defend their older pages. I will close this file when one of two things occurs: either a primary medical record arrives that corroborates or demolishes the implant chain, or a registration document explains the date anomaly beyond doubt. Until then both versions stay in the ledger — side by side, in two colours.

I leave the question with the reader. How many files in your system today are lying about their own name? And how many of them still have someone counting the sources?