The Mislabel Ledger: How a Mexico City Divorce Filing Entered a Football Dataset

I pulled the ledger. The record wore a label — "Football." Inside, the text read: the procedure for initiating an uncontested mutual-agreement divorce online in Mexico City (CDMX). The two do not sit together. Yet they sat — inside a data pipeline, inside a domain tag, as a silent error. In 2026, when Neymar's €222m release clause was triggered, I pulled the ledger the same way — wage documents, the FFP loopholes, Barcelona's amortization schedule. Back then every source had an address and a date. This time there is no address. Only a label, and a mismatch between label and content. This piece is the ledger of that mismatch.

Context: How a Label Becomes a Ledger

A modern content pipeline behaves much like a transfer window. An article enters, a domain label attaches, and it lands on the relevant analyst's desk. Stage-1 is ingestion — the article is read, segmented, tagged with a category. Stage-2 is deep analysis, where that tag is used to break the material across nine dimensions: tactics, club finance, results, league landscape, rules and governance, dressing room, risk, media narrative, and industry transmission. At every step, the assumption is that the label is correct.

The problem: a label is never proof in itself. It is a claim — a claim without a source. Much of my career has been spent measuring the gap between claim and proof. A club can say it is not selling its star; but line up the clause date, the wage band, and the registration window, and it becomes clear the sale was sealed three weeks earlier. In exactly the same way, a pipeline can call an article "Football"; but line up the twelve information points inside it, and it becomes clear it is nothing of the kind.

I will not speak in blockchain's technical vocabulary — that is not my field, and pretending otherwise would embarrass me. But I recognise its core promise, because it matches the founding principle of my own work: every entry should carry a verifiable source, and that source should not be alterable later. A ledger is valuable only when every row has a true history behind it. A pipeline that attaches a football label to an article containing no club, player, competition, or governing body has a corrupted block. And one corrupted block puts the whole chain in question.

Core Analysis: What the Content Says, What the Label Says

The Article That Is Not Football

Take the article's gist: a public-service explainer on how to begin an uncontested divorce online in CDMX. Note the key terms — the Judicial Power of Mexico City (Poder Judicial de la CDMX), the Virtual Office of Parts (OPV), FIREL / e.Firma / Firma Judicial electronic signatures, and PDF filing requirements. Not one of these is a football term. Not one club, player, league, or federation appears.

I state this plainly because the real story hides here. The point is not that the article is weak or wrong. The point is that the article is a perfectly sound civic explainer in its own domain — a public-service guide, objective in stance, informational in purpose. But it wears a label with which it has no relationship. Landing on a football analyst's desk, it is a category error — filed under a class to which it does not belong.

Auditing the Twelve Information Points

The twelve points from Stage-1 were examined one by one. The first three describe how to start an online divorce — entering the platform, identity verification, the initial application. Points five through twelve describe OPV, electronic signatures, and the technical requirements for PDF filing. Nowhere is there a single football fact. Football analysis can stand only with at least one formation, playing style, personnel usage, or match context. None exists here.

This is the first step of my professional habit — when information is insufficient, do not speculate; state clearly: "insufficient information, cannot assess." Many analysts err here. They force a weak source into a football mould and pass it off as analysis. I do not call that analysis; I call it fabrication. And my 37 years tell me that once fabricated analysis enters a ledger, it does far greater damage later.

The Silent "Not Applicable" Across Nine Dimensions

Stage-2 analysed nine dimensions. Tactical and technical analysis? Not applicable — no formation, no xG or PPDA, no squad. Club finance and transfer market? Not applicable — no club, no fee, no wage, so no valuation or FFP analysis is possible. Results and public-opinion cycle? Not applicable — a sample of zero matches. League landscape and team positioning? Not applicable — no league, no competition, no table. Rules and governance? Here the fracture is clearest.

The governance dimension operates in football through FIFA, UEFA, national associations, or league rules. But the governance framework here is Mexican civil and judicial procedure — unrelated to football governance. Measuring it with football rules means overstepping one's remit. Likewise, management and dressing-room analysis is not applicable, risk profile is not applicable, media narrative is not applicable, and industry-transmission analysis is not applicable. Almost every one of the nine dimensions returned the same answer: insufficient information.

To read this as failure would be wrong. This is correct behaviour. A pipeline is healthy precisely when it knows when to say "I do not know." The analyst who force-fills every empty cell does not fill the cell — he destroys it.

How a Label Goes Wrong: The Keyword-Collision Hypothesis

A technical question arises: how did the error happen? The most plausible explanation is a keyword collision at the classification stage. This is a hypothesis, not proof; I will not pass a hypothesis off as proof. Still, the pattern is telling. The error is not an isolated-field error — the title, the summary, and all twelve points point the same way: legal-procedural content. The mismatch is internally consistent. This suggests the fault is not inside the content but at the labeling stage.

That distinction matters. If the title had concerned football while the points concerned law, one might assume the article itself was ambiguous. But here every layer agrees. So the hypothesis is this: a mis-trigger at Stage-1 ingestion — likely a word or phrase that matched in a way that confused the classifier. It is a labeling defect, not a content defect.

Contagion of the Corrupted Block: Downstream Risk

Now the real danger. A single mislabel, left alone, does limited harm. But if this mislabeled record accumulates inside football datasets, it will contaminate the football analysis itself. Suppose a model counts this article as a football sample. Suppose a conclusion is then drawn from that sample — say, a domain average or a trend. Inside that conclusion now hides a divorce filing.

In blockchain terms, this is a corrupted block that later damages the reputation of the whole chain at verification. In my terms, it is a bad entry that, when caught in audit, casts suspicion on every entry before it. This is why I always say that the greater task is not correcting a bad record but installing the gate that keeps it out. But before that, there is one more thing — the most uncomfortable part of this case.

"Source: None" — The Biggest Gap

Note that Stage-1 lists the article's source as "Not specified." The author too — "Not specified." And beside each of the twelve points, the source cell is empty — "Source: None." Anyone looking only at the label error will miss this. But to me, this is the biggest story.

It means the problem is not only the wrong label. The problem is that this article has no verifiable birth certificate. Which outlet published it, who wrote it, when — none of it is confirmed. In other words, even if the label were corrected tomorrow, this article still could not stand as an acceptable source. A label can be corrected; a source cannot be restored.

Contrarian Angle: The Blind Spot Is Not the Label

While everyone stares at the wrong label, my question is different. I say: the wrong label is a symptom, not the disease. The disease is the absence of sourcing. If a pipeline ingests an article with no author identity, no publication date, no cited source, then attaching a wrong label is only a matter of time. Football today, something else tomorrow. In a ledger where an entry has no history behind it, there is no way to prevent a bad entry.

Here I want to caution blockchain enthusiasts. In blockchain's name, many speak only of immutability — once written, it cannot change. But immutability alone is not enough. If a false fact is written immutably into the ledger, that is not protection; it is a sentence. The real value is provenance — who wrote it, when, from which document, and whether that document can be verified. To me, this article's greatest failure is not the label; its greatest failure is that it carries nothing resembling a source hash.

And I must add one more thing, or I would be pointing a finger at my own profession without pointing at myself. Those who say "football data is always clean" are wrong. Over years of poring over matches and ledgers, I have seen that sports-data pipelines are riddled with the same errors. Some believe a heatmap reveals a player's role; in truth a heatmap is often tea-leaf reading — the real role hides inside the system, not in the picture. It is the same here: the label is a picture, the content is the reality. Whoever looks only at the picture will pass off a divorce filing as football and never notice.

The Next Domino

A Mexico City divorce filing will never become football — that is not this piece's point. The point is that once a bad entry enters any ledger, it can be corrected; but note that before correction, we must know how many other bad entries sit there silently. Audit the neighbouring records. Review the classifier's trigger rules. And where the source cell is empty, restore the source before correcting the label. Because one question still hangs: if a divorce filing can enter a football dataset, then what else lies hidden in that dataset that we have not yet seen?