Ten Empty Fields, Nine Filled Dimensions: The Costliest Misreading of N/A in Football Analytics

An 11:40 pm deadline mail. Twelve hours before a knockout tie, a scouting file opens — nine sections, a table inside each one, a conclusion column inside each table, a confidence tag under every column. The document looks finished. The format holds, the architecture is intact, the vocabulary sits where it should.

I turn to page one, then two, then three — and that is when the thing appears that formatting never shows.

Nine of the ten analytical fields are empty. Title N/A, source N/A, article type “unclassified”, the summary box blank. The field named “Information Points” holds not a single item. The field named “Entities Involved” holds no team, no player, no competition. And after all that — nine dimensions, a six-row risk matrix, a five-term glossary, four key-risk warnings, four signals requiring ongoing tracking. Everything present. Except the one thing a report exists to carry: facts.

This was not a match report. It was the post-mortem of a pipeline, sent onward in the shape of analysis.

A football analytics report travels through three stages. First, ingestion and deconstruction: the source text is read and its information points, named entities, time sensitivity and source quality are pulled out. Second, the nine-dimension analysis built on those points. Third, editorial review. The whole chain rests on one condition: that the first stage works.

In the document that reached me, the first stage had failed — and the second stage ran anyway. Tables were built, headings were set, confidence tags were applied, even the “Recommended next step” was written. This is less a technical fault than a design fault: a template that renders in every circumstance will render on empty input too.

The fingerprint of the failure was clear. Nine of ten fields empty, and the tenth — the domain label — correctly populated: football. That pattern tells you the article body never reached extraction. The HTML or PDF parse came back blank, the URL sat behind a paywall, or the feed itself was broken. The classifier did its job, which was only to say the material was football — and beyond that there was nothing.

In a tournament cycle this fault costs more than on an ordinary week. The night before a knockout tie, report volume multiplies, the deadline halves, and the reviewer who sees page one signs on page two. Nobody treats a document with nine filled sections as empty, even when it contains not one fact.

Read the ten-field audit line by line, because the damage starts here. The title is blank — the subject was never fixed. The source is blank — source tier was never graded, and without source tier there is no way to price a transfer rumour. Author stance and article purpose are both blank — the direction of the piece’s bias is unknown. Time sensitivity is unassessed — whether the information holds until tomorrow or until the next hour is undetermined. Entities are missing — which shuts club finance, league positioning, compliance and dressing-room, four dimensions at once.

There is one professional rule here that every football desk should pin to the wall: “N/A – insufficient information” does not mean “no risk identified”. It means: not examined.

The difference is not semantic, it is arithmetical. The risk matrix carried six categories — sporting, financial, personnel, rules, public opinion, systemic. All six came back “Undetermined”. A reviewer skimming six rows draws the conclusion: no red flags. All six rows are actually saying: nothing was looked at. Reading an empty cell as a green cell is the most expensive error in analytics.

The risk runs one way. A false positive enters circulation, takes fire, and gets corrected. A false negative never enters circulation, because it is invisible. A report that says nothing about FFP/PSR exposure reads exactly like a report that says “no breach” — same typeface, opposite meaning.

Put it in numbers and the whole condition of the document opens up. Ten confidence tags were applied — one High, two Medium, seven Low. The single tag carrying the most trust was not about a player, a club or a system; it was about the pipeline having failed. The most solid conclusion in the entire document was that the document had nothing to say. All four information-value dimensions — sporting, industry, timeliness, reference — stalled at one star. The glossary defined xG, PPDA, FFP/PSR and contract year; the document contained not one value for any of them. A full dictionary, an empty dataset.

I have an old benchmark for this. In 2026–21, my final year of university, campus shut, I coded the entire ISL season remotely as a video analyst — 110 league matches, three Goa venues, behind closed doors. Every goal kick, every second-ball recovery, every restart delay went into a spreadsheet. One hundred and ten closed-door matches turned silence into a dataset. A row in that sheet never went blank, and when it did it meant my coding rule was wrong — not that nothing had happened on the pitch. Behind closed doors, the crowd disappeared and the tactical truth got louder. The N/A cell in a report deserves the same treatment: count it, do not forgive it.

This costs money fastest in the transfer market. Where a fee is recorded, there is a ledger — instalments, amortisation, add-ons, sell-ons. Where the player is a free agent, the fee line is empty by design, and signing-on fee, agent commission and wage structure take its place. The transfer ledger confesses more than the unveiling press conference ever will. The difficulty is that in the free-agent case the ledger’s most important lines are missing by construction, and it is precisely where the ledger is blank that the most confident prose gets written. Large signing-on payments for free agents escape the scrutiny that a transfer fee has to absorb — because the scrutiny machinery stands on the same pipeline whose failure is recorded in this document.

The document did suggest a way out, and it was its most useful part: narrow the coverage to four dimensions — tactical, results and opinion cycle, risk profile, media narrative — and state plainly what was not examined. A coverage caveat is not an apology, it is a deliverable. A report that will not name its own limits is asking the reader to keep the accounts — and the reader never does.

The first instinct will be: fix the crawler, re-supply the URL, re-run stage one. Necessary, but that is a bandage, not treatment. The real defect sits in the template. A nine-dimension skeleton was built that renders on any input — so it renders on empty input, with equal confidence. A system that is not permitted to say “nothing” will lie, because lying is its default. The fix is procedural, not technical: if the information-point count is zero, the report must not render. Hard stop. An error message, not nine sections.

The second fault sits in incentives. A filled report gets approved. An empty one gets returned with “please complete”. The pressure therefore runs in one direction only: toward filling. Any analyst who has worked a deadline night knows where filling starts — with an estimate.

And a third point, less comfortable. The ISTJ in me audits the outlier before crowning it. Here the outlier is not a spectacular number, it is an empty cell. The industry runs the other way: the loudest figure gets audited first, the quietest cell never. Spending on models keeps climbing — xG, PPDA, passing networks — while spending on ingestion stands still. Nothing on the foundation, the whole budget on the roof.

The next report that lands on my desk gets three checks, and the checklist travels to any football desk. One: does “Information Points” contain at least one named entity? No name, no report. Two: does every N/A read “not examined” or “no risk”? Three: does the document carry a publication date and a time sensitivity? Football information without a date is a rumour in a better font.

The question is not how a document survived with no facts inside it. The question is how many of the other reports on your desk are only “finished-looking”. Next time a file opens with nine sections, check the field called Information Points before you turn page one.