An Empty File, A Full Confidence: The Zero-Information Case in Football Reporting

The file was on my desk. The cover sheet was complete — name, club, position, age group — and at the bottom, in green pen, one word: "recommended." Every field between was blank. Birth year blank, minutes played this season blank, under-17 league appearances blank, and the column where a second independent source is supposed to confirm the whole thing, blank. Whoever built the file assumed the empty boxes would be filled later. Whoever signed it in green assumed the empty boxes belonged to someone else. Both were right, and that is exactly why the file closed.

I put it in the drawer with the thin files. I read the thin file first. The thick file usually belongs to the agent — photographs, cuttings, a coach's phone number, and a great many adjectives. The thin file holds only what someone counted and wrote down.

Another file of the same species landed on the desk days later: a nine-dimension analysis of a football article. Its cover was populated too — domain label "football," an article-type field, the full template. The values? Title not available. Source not available. Source quality not assessed. Article type unclassified. The information-point list empty. Time sensitivity never assessed. Nine dimensions, nine identical verdicts: insufficient information.

It is the same document in a different binding. Headers filled, body empty. The difference is that someone had to pick up a pen to sign the scouting dossier, while the dashboard coloured the analysis green by itself. The run registered as successful.

It is worth understanding the mechanics. Structural fields — article type, domain label — come from one stage of the pipeline. Value fields come from another. When the page body never arrives, because of a paywall, a JavaScript-rendered template, or a blocked fetch, one stage returns empty while the other returns normal. The output then looks like an article that genuinely contained nothing. An article containing nothing is close to impossible. A piece with a headline, a sport label and a type classification will normally surrender three to five hard points. When the structure is populated and every value is null, the fault is in the extraction, not in the source.

That distinction is fundamental and it gets skipped every day. A source that was never recorded cannot be graded low-confidence; it is zero, and anything multiplied by zero stays zero. The report did the only defensible thing — it refused to treat a missing source as a weak source and returned the whole matter as unassessable.

I recognise this because I have lived it from the other side. In 2026, in Hyderabad, I was on the press bench for the AFC Challenge Cup final: India 4–1 Tajikistan, a first Asian Cup qualification in twenty-four years. The match report went in forty minutes. Then I went back through the tournament's team sheets. The final XI averaged 26.4 years, and not one outfield player under twenty-three had started a single match. My follow-up argued the win was a ceiling, not a floor. The sports editor spiked it as negative. In 2026 I filed a column nobody wanted. I kept the carbon copy.

That episode changed my habits. A birth-year column began sitting beside every squad list, and ages were confirmed by ringing clubs rather than copying federation releases. Editors started routing golden-generation claims to me for verification. I learned to write the corrective piece first and the celebratory piece second.

Earlier still, in the summer of 2026, I had bought a hardbound register after a Federation Cup semi-final at Salt Lake Stadium. East Bengal 4–1 Mohun Bagan, a Bhaichung Bhutia hat-trick, an official attendance of 131,781. 131,781 went through the gate. The ledger kept what the reports lost. On the flight back I decided that after 2026 every attendance, every birth year and every transfer figure would clear two independent sources before it reached copy. Two decades later that habit became the skeleton of my youth-player database.

So when a file announces that it holds no information, I do not dismiss it as weak. I record it exactly as it arrived. And the nine-dimension report did the right things. It marked the compliance status of an unnamed club as unassessable rather than assuming the club clean. It refused to assign an agent motive to a transfer rumour that did not exist. And most importantly, it declined to model sanctions for a club it could not name. Building a breach scenario around an anonymous club is not analysis; it is an accusation with a table around it.

There is a cost asymmetry the industry keeps forgetting. A lost news cycle refills in two or three days; a fabricated name stays in the file forever. Football copy decays inside twenty-four to seventy-two hours. A wrong birth year, or a name attached to the wrong club, gets quietly copied into every subsequent report for decades. In a scouting dossier the price is higher, because there the file is not a record — it steers the next ten years of a seventeen-year-old's career.

The blank field is dangerous precisely because it applies pressure. Behind every empty box stands someone who needs it filled: the agent, the club's media officer, the fan account, and now the automated content generators that fill an empty payload with enthusiasm. The moment a system admits it holds no information is the moment the competition to invent some begins. Every wonderkid is a rumour until the under-17 ledger says otherwise.

Here is the second trap, and I want to name it plainly. An empty payload does not usually mean a cover-up. It usually means a broken pipe — a paywall, an unrendered page, a blocked fetch. Constructing a conspiracy out of that is as dishonest as filling a blank box with a name. If the same file returns tomorrow with three hard facts, and those facts vindicate the chatter, print it plainly. Contrarianism as a reflex is the same laziness I have spent three decades avoiding, wearing better clothes.

The fix therefore sits at ingestion, not at analysis. Any article with a null title and fewer than three information points should fail upstream with an explicit log entry. A green light on a dashboard should not be allowed to mean that a file contains something. A file left in a dark drawer slowly starts to feel like the truth. No shortcut through the archive: you walk the aisles or you guess.

The question is not what that analysis lost. The question is how many files sit in the register right now with a green pen at the bottom and nothing in the middle. The next time a dossier arrives, the first question should be who wrote the birth year down, and where the second source is. A file that cannot survive that question is not recommended.