Empty Data and the Football Matches That Never Happened

In the autumn of 2026 I sat in front of a screen and rewound a single passage of play in El Clásico exactly twelve times. Not to find emotion. To measure. I plotted the ball's coordinates frame by frame, compared the position of the touchline in camera A against camera B, and placed both on the same ruler. A 1.7-metre offset. One moment, one match, two cameras in two different positions, and VAR confirming a Real Madrid goal from the wrong angle. My piece travelled more than 200,000 shares within 24 hours.

People called it a technology scandal. I called it a routine morning at work. VAR was born to correct human error, and it ended up manufacturing machine error. By the time the competition learned to read frame timecodes, everyone had forgotten the larger lesson: a wrong frame is still a frame. There is an image, an object, a person, a timestamp to confront. The more dangerous thing sits on the other side of the mirror — the empty frame.

The television screen does not lie. Only the person sitting behind it lies to himself.

I met that empty frame again recently, in a different form. A football analysis file entered a processing system, fully labelled: domain football, article type, a nine-dimension analytical framework. The article title was blank. The publishing source was blank. The publication date was blank. The list of information points was blank. No club was named. No competition was named. No player appeared. No scoreline, no fee, no league position existed. The only populated substantive field was a single line: domain, football.

And the analytical engine did exactly what it was built to do: it refused to conclude. Every cell across nine dimensions was filled with the phrase insufficient information. No tactical judgement. No financial assessment. No risk model. Just one large warning at the top of the document, stating that the most dangerous outcome of this run was the possibility that the document would be read as real analysis.

I tell that story because it is the story of the entire football industry right now, only at a different scale. A headline with no source. A rumour with no date. A transfer figure with no contractual terms. A tactical breakdown with no minute of play. Stitch those four together, republish it three times, and you have a complete truth in the reader's eyes with not one fragment of evidence behind it.

The current transfer window is in its loudest phase. That is when this genre breeds fastest, because nothing can be verified until the window shuts. A player is said to have agreed personal terms with a new club while his existing contract still has two years to run and no release clause has been triggered. A club is said to have lodged a forty-five million bid while its wage bill is already at the ceiling and it must sell before it can buy. Not one person republishing that claim bothers to check how the fee would be amortised across the contract, who holds the sell-on percentage, and where the player would sit in the dressing room's wage hierarchy.

That is the entire problem. Football is operating on empty files wrapped in full language.

I began this trade in 2026, aged twenty, reading sports bulletins on radio. There was no live data then, no probability models, no spatial heat maps. There was a notebook, a pen, and one rule I set for myself: if I cannot record a source, I cannot record a sentence. That rule has followed me for more than half a century, through newspapers being founded, through books on football, basketball, and the sport people still insist has no sweat in it.

When football entered the digital era I noticed a paradox. The volume of data multiplied a hundredfold, while the volume of evidence inside each sentence fell. There was more to measure and less to believe.

The phrase "a source close to the player" is a linguistic structure, not a source.

Try grading it the way I still grade transfer news in my notebook. A first-tier source is an official statement with a timestamp. A second-tier source is a direct quote from a sporting director or an agent, with audio that can be checked. A third-tier source is a named journalist with a correct record, at a named outlet, traceable twenty years back. A fourth tier is the outlets that live entirely on rumour, where the hit rate sits below random chance. And a fifth tier is "a source close to" — a phrase with no person, no job title, no accountability, and no consequence when it is wrong.

When you see four different outlets citing the same "source close to" in the same afternoon, you are looking at four copies of a blank sheet. I once reconstructed one of those chains with a search tool, traced it back four layers of republishing, and the origin was a thirty-character status update on an unmoderated forum. Thirty characters became national breaking news inside eighteen hours. That is the speed of reaction, and it is also the speed of self-deception.

Now to the part I consider more serious than rumour: live data sold into the betting market.

Many people assume that digitising football means fans get to watch more. True, but that is the visible half. The submerged half is that every pass, every touch, every running stride, every gap between two centre-backs is captured and sold within seconds to a single class of customer paying far more than the price of a stadium seat. Broadcasters pay for the pictures. Bookmakers pay for time — being a few hundred milliseconds ahead of you is already an advantage.

The television screen does not lie. Only the person sitting behind it lies to himself. You watch the same match, the same passage, the same frame. But you watch it after everything is finished. Someone watched it before you, and placed money before you.

That is the darkest side effect of sport's digitisation, and it does not sit with referees, nor with players. It sits in the fact that the data serving viewers and the data serving bookmakers come from the same source, differing only in delivery time.

I have said this plainly for years: do not imagine that looking up metrics makes you understand football better. It makes you more confident. Those are different things.

Back to the defence, where everything actually begins.

The France back line of that year needed no prediction — you only had to watch how they stood.

At the 2026 World Cup, before a ball was kicked, the world talked about Pogba and Griezmann. I spent fourteen minutes of footage on something else: how France's back four concealed weaknesses on both flanks during transition. Varane and Umtiti held a high line, the gap between the two centre-backs widened as the team built out, and the full-backs were pushed ahead of the ball. That system worked perfectly against opponents with nobody quick enough to exploit the space behind. Against Argentina, that space had a name.

The conclusion in my footage was specific: Pavard needed to drop three metres deeper than his group-stage average, to deny Messi the half-second he needs to receive, turn, and face goal. Three metres. Not a strategy — a measurement. The Argentina coaching staff later photocopied my piece as a meeting document. I do not mention that to boast. I mention it to mark the difference between an opinion and a piece of evidence.

I saw it in advance because I looked. Not because I guessed.

What I looked at, concretely, was static structure. The way a back four stands before the ball moves indicts the entire intention of a collective. If two centre-backs stand level and both full-backs sit deep, that team will defend as a block. If one centre-back steps and the other covers, that team is preparing an offside trap. If the midfield forms a narrow diamond, that team will press by forcing the ball wide and locking the return pass. None of that requires the ball to read. It requires eyes, time, and a notebook recording patterns across tournaments.

Modern football loves numbers, but numbers do not know fear. A defender does.

That is why I persist with the old method: watch, record, cross-check, then conclude. You can hand me expected goals, passes allowed per defensive action, possession broken down by zone. I read them all. But I read them like a map, not like a verdict. A map tells you the terrain. The verdict must come from the person walking it.

I see a dangerous habit among young analysts: treating metrics as the answer rather than the question. When a team loses with a high expected-goals figure, people say it deserved to win. Football does not run on deserving. It runs on goals, and a goal is the output of a specific sequence of decisions in a specific period of time, with specific players in a specific physical state. No model contains the fear of a defender in the eighty-eighth minute when he knows he no longer has the legs for one more sprint.

One thing I must state clearly, because it is where I differ from the crowd: the problem is not data. The problem is people who use data without declaring the source, without stating the sample, without naming the limits. A metric published without a sample size is a polite lie. A percentage without a variable definition is bait. A prediction model without a confidence interval is propaganda.

And here is where I join the two stories. The analytical engine I mentioned at the start, handed an empty file, did not invent players, did not invent scorelines, did not invent competitions. It stopped and said: insufficient information. If a machine programmed to conclude only on evidence behaves more honestly than most humans in this trade, then our problem is not technology. Our problem is habit.

That habit has a name: preferring a plausible-sounding answer to no answer at all.

I have watched enough World Cups to know: the champion is the team that corrects itself the fewest times. Not the most beautiful side, not the side with the most stars, not the side with the best group-stage metrics. The side forced into the fewest corrections — in its system, in how it stands, in how it transitions, in how it spends in the transfer market. A club that corrects itself repeatedly inside one window is a club that does not yet know what it is.

That brings me to the hardest part of this piece: where I might be wrong.

There is one possibility I must concede. I may be underrating the power of modern metrics, because I was born before them and my eye was trained a different way. It may be that the models I distrust see things the naked eye cannot: invisible pressure, space generated before the ball arrives, the cumulative effect of hundreds of small actions no single frame can narrate. I have set myself a bias-offset rule for age: every season, I spend twenty minutes reading an analysis written by someone young, trying to understand what they measure and why. Not to agree. To know what I am arguing against.

There is a second possibility, and it stings more. It may be that the very noise I condemn is what feeds this sport. Rumour generates clicks, clicks generate advertising revenue, advertising revenue pays for broadcast rights, and broadcast rights pay player wages. If I demand that everyone speak only when they have evidence, I may be demanding something that shrinks the industry by half.

But I have read enough empty reports to know my limits. An industry built on files without sources collapses when the people paying start demanding verification. And the people paying have started demanding.

Look again at how I caught the camera error in 2026. No system flagged it. No expert called me. There was a man in front of a screen rewinding twelve times, refusing to believe what he had just seen until he had measured it himself.

Seeing comes before believing. Not the other way round.

That is my whole method, packed into four words.

If you ask me for a prediction over the next few years, I will not give you a club name or a player name. I will give you a structural one. Whichever club builds a verifiable record system — where every transfer decision has a source, every analysis has a sample, every claim has a timestamp — will hold a competitive advantage larger than any financial edge money can buy. Because when the transfer market is pure rumour, the only party that knows the truth buys at a better price than everyone else.

And that is the closing note to the story of the empty file I encountered not long ago. That engine did not produce a wrong analysis. It produced an honest emptiness. In an industry busy inventing its own matches, honest emptiness is the most valuable thing you can own.

People will tell me that means a blank newspaper, a blank programme, a blank broadcast. Yes, it does. And I will answer with the single question I still use to close every conversation with a young editor.

If you have nothing to say, why must you say it?