The Millimetre Offside Line and the Empty Report: When Football Measures What It Cannot Measure

The report ran to nine pages. It had a tactical table, a six-row risk matrix, and a glossary of technical terms at the end so the reader would not get lost. Every cell was filled in. And every cell said exactly one thing: insufficient information.

I read it at two in the morning in Manchester, after the desk sent it over with a one-line note: “Can't use this. Rewrite it. Need it by nine.”

Seven hours. A match with no data. A document with no content. And a broadcaster sitting in a rented flat with a cup of tea that had gone cold long ago.

The first reflex was to open a blank file and start typing. No event? Describe the emotion. No numbers? Talk about spirit. Nothing at all? Write about identity and hunger — the two words this industry uses to plug every hole.

I got four lines in and stopped. Three of those four lines contained a claim I could not verify.

That was the moment I realised I was holding the rarest thing football analysis produces in any given week: a document that refused to conclude.

What happened was a data pipeline failure. Stage one — extraction — returned an empty payload: no title, no source, not a single information point, not one identifiable entity. Stage two carried nine deep-analysis dimensions, and it did exactly one thing: it ran a gate check, found the gate shut, and stopped.

Instead of filling the gaps with plausible-sounding sentences — some club, some transfer fee, some xG figure — it wrote “insufficient information” nine times and closed.

My editor read a broken document. Professionally, he was right: nobody pays for a blank page. But that blank page was saying something very specific about how this industry operates, and that deserved to be printed.

In 2026 I worked at a sports radio station in Manchester. The France–Belgium semi-final on 10 July 2026 finished 1-0, Umtiti heading in on 51 minutes. In the first half I mispronounced Nacer Chadli's name three times and the switchboard took complaints. That night I redrew France's transition map with fourteen decisive passes from Mbappé and posted it to my personal blog. Two days later it had twenty thousand views.

A colleague's verdict: too academic, too dry.

I still remember that line better than the twenty thousand views.

Then in 2026 The Athletic put me on crisis clubs. On 26 May 2026 Sunderland lost 1-2 to Charlton in the League One play-off final at Wembley, the winner coming in the 94th minute from Patrick Bauer. Mid-match I tweeted that relegation was a chance to cut out a financial tumour. A veteran reporter pushed back hard. I went out, interviewed twelve people — creditors, supporters, club staff — wrote four thousand words, and the piece was shared twelve thousand times.

Those two stories differ in one respect, and that difference is this entire article.

In 2026 I had a conclusion before I had data: four lines of prose, then go find the numbers. In 2026 I had twelve interviews before I had a conclusion. The industry rewards the first and punishes the second with silence, but only the second is still standing a week later.

The empty report is neither. It had no conclusion and it had no twelve interviews. It had a gate that was shut.

The four risks the reader never sees

What stands out is that the report ranked four risk warnings by priority, and three of the four belonged to the reader's side rather than the data's side.

First: an empty result can be swallowed by the next processing stage and quietly treated as a completed analysis. A null result can travel an entire production line without anyone stopping to ask.

Second, and this is the one that stings: fabrication risk under pressure to deliver. The report called this the greatest integrity hazard in the situation and recommended preserving the null result — no entity may be introduced that cannot be traced to an original information point.

I have been on the other side of that recommendation. At twenty-eight I had seven hours, a blank page and an editor waiting. I know the feeling exactly. It is not the feeling of wanting to deceive anyone. It is the feeling of professionalism misread — the belief that silence is failure, that there is an obligation to say something.

Third: a template defect. The entity field still contained an unsubstituted instruction string, a signal that the entity-extraction step never ran. It is a small detail, like a scratch on glass, and it says the problem is not the source article.

Fourth: the piece may have been legitimately rejected upstream — not an article, paywalled body, not newsworthy — but the rejection reason was never propagated.

Reading those four lines, I felt I was reading a description of my own trade, written in the language of a systems engineer.

The family of false precision

The Premier League introduced VAR from 2026-20. In the first two seasons viewers were shown super-slow replays, lines drawn as thin as a hair, and goals disallowed because a shoulder or a toenail was ahead of the last defender. From 2026-22 the authorities thickened the lines to give attackers the benefit of the doubt.

That thickening was a confession. It admitted the thin line was not measuring anything real.

I am not against technology. I am against using a line to answer a question a line cannot answer. In athletics, optical timing measures to a thousandth of a second and measures correctly, because the finish is a line lying still on the track. In swimming, the touchpad records the instant a finger hits the wall and records correctly, because the touch is a discrete event — it happened or it did not.

Football measures, in millimetres, a body in motion, on a frame that might be the frame before or the frame after, against a concept of “advantage” the laws cannot define numerically.

Athletics measures in thousandths because the finish is a line. Football measures in millimetres because the finish is a human being in motion. The two measurements look identical on a screen and are entirely different in nature.

xG belongs to the same family. An xG figure is a distribution compressed into a decimal. It is useful across hundreds of shots over months. It is meaningless as proof that a side “deserved” to win a particular match. Football does not award points for deserving.

PPDA is the same story. It counts the passes an opponent is allowed per defensive action. It does not measure intent. A low-block team has a high PPDA and people read “no intensity”, when the truth may be that they are waiting for the final twenty metres — the only metres that matter.

The five-substitution rule, made permanent by IFAB from 2026-23, is another member of the family. It makes squad depth more valuable and turns the last twenty minutes into a war of attrition between two benches. Substitution counts are measurable. The tactical consequence has to be watched, and watched repeatedly.

In March 2026, with football frozen, I sat in an empty Etihad and counted Pep Guardiola shouting “play faster” twenty-seven times in one half. On an empty night at the Etihad, I suddenly heard my own applause most clearly. That number twenty-seven proved nothing about Manchester City's tactics. It proved only that when the crowd goes, you start hearing what the noise had been covering.

The file-drawer problem

In 2026 the psychologist Robert Rosenthal named a disease of the research world: the file-drawer problem, published in Psychological Bulletin. Failed trials do not get published. They sit in the drawer. Medical literature ends up containing only successes, and a generation of doctors comes to believe every intervention works.

Football has the same disease with a different drawer. Every match generates an analysis with a conclusion. Nobody publishes “I watched ninety minutes and could not work out what happened”. Over years the industry accumulates an illusion: that every match is explicable, that every result has a technical cause, that enough data will always produce enough answers.

The empty report is a page that slipped out of the drawer.

Based on my experience covering matches over nearly two decades, the hardest games to analyse are always the ones where both sides executed their plans and the result came from a moment that will not repeat. Those are the games that need the longest piece and produce the emptiest one, because there is nothing to tell except the admission.

The counterweight: false humility is as comfortable as false confidence

I have to argue against myself here, or this becomes just another performance.

There is a great temptation in worshipping emptiness. It lets an analyst never commit. Never be wrong. Never be accountable to a reader for a conclusion.

False humility is exactly as comfortable as false confidence, and slightly more dangerous because it gets praised.

The empty report is not Socrates. It does not interrogate itself. It is a blocked pipe. It does not say “I do not know”; it says “nobody gave me anything to know”. The difference between those two sentences is the whole of professional ethics, and the whole difference between an analyst and a printer.

Real humility goes and gets the data. Fake humility sits and waits for the data to come to it.

In 2026 I did not sit and wait. I called Sunderland's creditors, the people who had bought season tickets for twelve straight years, a club office worker who had just been made redundant. Sunderland went down, and I still hold the view: going down is a way of going up. But that view was only worth something because it was paid for with twelve interviews, not with a quick line on social media.

In 2026 I wrote about France's 4-2-4. 4-2-4 is not a formation, it is a test of who dares to dream. But I was only allowed to say that after redrawing fourteen decisive passes and checking each one. Without those fourteen passes, the line is just poetry.

And here is what I want to say to anyone reading the empty report as a moral mirror: it is not a mirror. It is a warning that in a nine-step process, one failure at step one renders the other eight meaningless — and step one, in football as in any trade, is always a human being.

People remember me for three mispronunciations; I choose to remember myself for three decades on the pitch. I mispronounced a name three times live on air, and it did not make me wrong about myself — because after the mic went off I sat there checking that name until I had it right.

What is worth keeping

Next time someone sends me a report that says “insufficient information”, the right answer is not “rewrite it”. The right answer is “go and get the information”.

The offside line will not disappear. xG will not disappear. Nine-page reports certainly will not. But if one day a data department in England published its drawers — the matches it could not explain, the players it judged wrongly, the models it threw away — audiences might stop trusting absolute numbers and start trusting questions.

Football has always been a play of mistakes — I have only helped make it worth watching. And one of the most watchable mistakes of this season is a nine-page document that says nothing at all, written by a machine with the nerve not to invent.

At nine in the morning I sent the desk a piece about that match. It carried no tactical conclusion, but it opened with one line: nobody has watched this game closely enough to say anything yet, including me.

Football is the only sport where a line can erase a goal, and the only one where a goal can erase a line. The writer's job is not to pick a side. The writer's job is to say clearly where he is standing when he looks at that line — and if that place is an empty room with no data in it, then to say plainly that he is standing in an empty room.