Nine Dimensions, Zero Evidence: The Football Report That Passes Every Check and Says Nothing

A document arrives. Nine dimensions: tactical and technical, club finance and transfer market, results and public-opinion cycle, league landscape, rules and governance, management and dressing room, risk profile, media narrative, industry transmission. Each carries a table. The risk matrix has six categories, each with four columns — level, likelihood, impact, mitigation. There is a transmission-path diagram with upstream, midstream and downstream boxes all visibly occupied. There is a professional glossary: PPDA, FFP, PSR, panic premium, tapping-up, solidarity mechanism. There is a disclaimer. By format alone it is a mature professional product, and that is the first deception.

Now read inside the boxes. "Insufficient information." Next line: "Insufficient information." Five of the six risk-matrix rows repeat the same sentence, and the sixth carries a real warning — methodological risk, high. Information points: zero. Named entities: zero. Article title: missing. Source: missing. Time sensitivity: not assessed.

On its final page the document declares itself a null result. That honesty is rare and I respect it. But there is a rule in my ledger I have not broken in forty-seven years: a report with no name in it is not a report, it is a form. And this is a form with nine dimensions.

The pipeline has two halves and we call them by one name

Football analysis is now a pipeline. The first half pulls facts out of raw text — title, publication source, information points, entities, date. The second half builds conclusions on those facts. Call the first deconstruction and the second analysis. When the first returns empty, the second has two routes: lie, and populate the empty cells with plausible answers, or write down that it has nothing. This time it took the correct route.

Taking the correct route does not make the product safe. This document can still be downloaded, turned into a slide, attached as an annex to a committee. Think how an appointments committee reads it. It sees nine dimensions. It sees a risk matrix. It sees a glossary. It reads "high risk" and assumes someone assessed the sport. Page count is not evidence, but it looks like it.

That is why I started hand-coding in a Kochi gantry in 2026. Six weeks, six venues, a nine-person technical group, one woman. Not fifty matches — fifty-two, minute by minute, roughly four thousand seven hundred minutes of footage into a hand-built press-to-space map. A sixty-page note on Steve Cooper's England: Rhian Brewster's eight goals, Phil Foden's receiving angles, and the second-ball pattern that produced Jeakson Singh's 82nd-minute header against Colombia on 9 October 2026 — India's first goal at any FIFA tournament.

Proprietary position feeds existed then. I coded by hand because position data arrives with its assumptions attached. I coded fifty-two matches by hand before I trusted the algorithm. And when the Bundesliga returned on 16 May 2026 — Borussia Dortmund 4-0 Schalke, stands empty — I laid the same method across six hundred and twelve matches: three hundred and six with crowds, three hundred and six without. Home win rates fell from 43.3 per cent to 33.6 per cent, away-team yellow cards dropped by roughly 0.6 per game, and defensive blocks sat an average 4.2 metres deeper. Nobody funded that study. The crowd is not noise; it is a variable I had to remove.

Completeness is a formatting property, not an epistemic one

A nine-dimension document has nine chances to look rigorous and no obligation to be right. A template is a container; a container is not content. Football makes this mistake every week — reading a player's 92 per cent pass accuracy and calling him a midfield controller when eighty-one per cent of those passes travelled under five metres and backwards. The number was present. The number's meaning was not. The document was complete. The document's claim was not.

In my ledger a match counted as coded only when it had four things: minute, scoreline, date, attendance. Not four filled cells — four facts. That distinction is basic and it should be applied to every report. Thirty filled boxes with no fact inside is thirty empty boxes.

Somewhere around the forty-first minute the game stops being about football and starts being about load budget. I keep a ledger because memory lies about the seventieth minute. But a ledger can lie too, if the rush to fill cells outruns the discipline of recording facts. That is the first lesson of this null result, and I suspect half the people who read it will read it another way.

The cascade of N/A

Zero does not propagate linearly. It falls like a waterfall. Information points zero, so entity list zero, so named players zero, so tactical assessment impossible, so financial assessment impossible, so compliance level impossible. The document is honest about this descent, and that is its strength. But the honesty carries an irony: recording the descent so systematically makes the document longer than a real one would have been.

Take the transfer-rumour credibility dimension. It needs a source tier — which journalist, which outlet, what the agent's interest is. The source field is empty, so the dimension cannot run. It still occupies half a page. A reader of that half page does not conclude "this dimension is switched off"; the reader concludes "work was done here and the finding is that information is absent." Those are different sentences and the gap between them is the whole problem.

The fingerprint that says the extraction never ran

There is one real finding in the entire document, and it sits in the diagnostic, not the analysis. The particular pattern of empty fields — domain label populated, everything else blank — is a fingerprint. It separates "extraction never ran" from "extraction ran and found nothing." The first is a pipeline fault, the second an article fault. Their treatments are completely different and treating them as one is expensive.

I have seen that same fingerprint in Indian football data repeatedly. From 2026 to 2026 the ISL was a closed franchise structure with no relegation. The label was present — "football league." The substance was missing: the thing that makes the bottom of a table behave in a particular way, the obligation to survive. A model trained on European football read the lower half and saw a crisis that did not exist, or, in reverse, could not see the intensity where something genuinely was at stake. The pattern showed up at match twenty-eight. I believed it at fifty-two.

The promotions of Punjab FC in 2026-24 and Mohammedan Sporting in 2026-25 put real sporting stakes back into that dataset. Now something is genuinely lost at the bottom of the table, and the numbers there behave differently than they did. That is not the method's victory — it is the structure's. The method merely reads it. The day the structure changes, the old model starts lying, however full its cells are.

The same applies to an imported model's visa. Thirty-nine-year-old marquee signings, a closed franchise structure and a decade of missing pyramid data: a model calibrated on European standards is not a neutral instrument there, it is imported assumption. I build my own baseline first, then test the model against it. The model was right on seven of nine blocks and wrong on two; my hand-count tells me which two.

A filled template is more dangerous than an empty one

The instinct on seeing this mess is to fix the pipeline: improve the extraction, swap the scraper, repair the parsing. My objection is that the only thing that behaved correctly in this entire episode was the pipeline, which told the truth about having nothing.

The dangerous repair is this one: improve extraction so the templates fill. Because then you get the same document, same nine dimensions, with a number in place of every "insufficient information" — a number nobody audited. I have seen filled nulls. Sixty per cent pass accuracy, fifty-five per cent possession, twenty-seven press events. One hundred per cent field coverage. And when you ask what the denominator is, the answer is: the vendor's feed. No feed provenance, no audit of its assumptions, no date on its sample.

An empty document at least tells the truth. A filled one gets published, screenshotted, cited, and then turns into a scouting decision. I have seen fast averages padded by sample bias, and an estimated objective accuracy of 81.2 per cent — but I did not take that figure from someone else's report, I recoded it across my own fifty-two matches.

It also needs saying that sixty-three years and forty-seven years of experience is not an argument. A claim stands on the ledger, not on the years. I have been in press boxes since the 2026 Asian Games, and if that sentence looks like it is carrying my argument, delete it — I have fifty-two matches, dates, attendance figures and minutes. That is my passport.

Before you open the next document

This document type breeds in the media. Its first cousin is the federation press release with a date, an official's name, a quote and no denominator. The FIFA ranking sits in the 100–130 band; the release still says "closing on Asia's elite." It is the same null result in another costume — entity filled, framework filled, facts absent. "New dawn" is a press officer's draft, not a finding.

The next time a long nine-dimension analysis opens in front of you, do this before reading the conclusions: count the named clubs, players and coaches on the first page. If the count is zero, close the tab. If it is three, find the denominator under every number — how many matches, how many seasons, how many of those minutes repeat. A number without a denominator is not a number, it is a draft.

And for the next match, do not take a model into the ground. Take one question: what is this number divided by? When you know the answer, turn the sound off — I turned the sound off, and then the game said something.