The Empty Report: When Football's Data Machines Sound Wise Without Saying Anything

Monday, nine in the morning, third-floor meeting room of a magazine office in Marseille. A young colleague places in front of me a forty-page document about Sunday's match. The automated analysis system has just produced it. There are tables. There is a risk matrix. There are nine sections stretching from tactics, club finances, governance and dressing-room health all the way to media narrative and the transfer market. Every field is filled in.

I read it all, then read it again. After forty pages, I know nothing more about the match I watched for two hours.

Those forty pages are valid in every formal respect. All the keys are there. All the fields are there. All the hierarchies are there. The machine reported a successful run, because every field existed. No field contained a fact. Each one was a variation on the same sentence: insufficient information to assess.

I wrote a line in my notebook that I later realised applied to more than software: a system can confirm itself without ever confirming reality.

The overhaul has already happened, and what did it leave behind

Over roughly the past fifteen years, European football has undergone a quiet but thorough transfusion. In 2026, when I began writing for local radio stations, the analysis department of a top-flight French club usually consisted of one man, one notebook and one television. By 2026, when I was working as an editor in Marseille, every training session had three cameras recording the positions of twenty-two players to the hundredth of a second, and software producing numbers that no coach of the nineties could have dreamed of.

What did that change bring? A great deal. It helped small clubs find players before the market saw them. Brentford bought Ivan Toney from Peterborough for a modest fee and carried him into the Premier League. Brighton signed Moisés Caicedo from Independiente del Valle for around four and a half million pounds in 2026, then sold him to Chelsea for a fee above one hundred million pounds in August 2026. No traditional scout could have combed through thousands of hours of footage in Ecuador just to find a nineteen-year-old midfielder.

But at the same time, a new class formed: a class of reports nobody reads to the end, of models nobody validates, of numbers generated to reassure a board rather than to describe a match. And inside that class, emptiness wears its neatest clothes.

This story does not begin in the 2010s. It begins in 2026, when Charles Reep, a former Royal Air Force officer, sat down to hand-record thousands of matches and concluded that most goals came from moves of three passes or fewer. That conclusion was circulated as scientific truth and helped shape the long-ball game in England for decades. Only much later did people point out that his sample was far too small and his notation method omitted almost every move that did not end in a goal.

This is the first lesson my trade taught me. A conclusion can be numerically correct, neatly presented, widely cited, and still wrong about the match.

When every field is filled in

The striking thing is that an empty report does not look like a mistake. It is the output of a properly designed chain. Someone builds a schema, a framework dictating which fields a report must contain. Someone builds a validator, a program that only proceeds when every field exists. When real data fails to arrive, the program still does not stop, because the framework is valid, because every field has been declared, because every cell has been assigned a default value.

In software, this is called a green build with empty data. The dashboard lights up. No red warning appears. And the whole system is confident it has just done something useful.

Football has green builds of its own. A team holds sixty-eight percent of possession, fires twenty shots, posts an expected-goals figure of 2.3, and leaves the pitch beaten without scoring. The statistics say they deserved better. The scoreboard says they got nothing. And in the post-match meeting, people will argue about whether the team played well, when the better question is why the model failed to see what twelve thousand people in the stands saw from the twentieth minute onward.

On 18 May 2026, at the Olympic Stadium in Athens, Johan Cruyff's Barcelona had more of the ball than Milan, combined more, and lost four-nil. Thirty years later we have hundreds more metrics to measure the distance between control and decision, yet that distance has not narrowed. It has only become easier to hide.

On 22 November 2026, at Lusail, Argentina held nearly seventy percent of possession and lost two-one to Saudi Arabia. Three days later, in Doha, I listened to a supporter say his team won on belief. I did not write that line down. I wrote that the Saudi defence pushed high like an unlocked door nobody dared walk through. A young editor called the piece too poetic. I accepted it, because there is something the statistics cannot measure: the moment of hesitation.

A player receives the ball at the edge of the box. He has four hundredths of a second to decide. Inside those four hundredths, the model is of no help. It can only say that in four hundred similar situations, most players shot. It cannot say that the night before, this player did not sleep, and that the image of his three-year-old son with a fever was still in his head.

Data does not score goals, but it knows where the ball will go. The problem is that it knows this after the ball has already gone.

A gap is not a clean bill of health

There is a logical error I have seen repeated in almost every meeting room I have sat in: people read the absence of a bad signal as the presence of a good one. No red warning, so everything is fine. No bad debt recorded, so the accounts are healthy. No complaint filed, so no problem exists.

In November 2026, Everton were docked ten points for breaching the Premier League's profit and sustainability rules. In February 2026, the sanction was reduced to six points on appeal. In March 2026, Nottingham Forest received a four-point deduction. Earlier, in February 2026, the Premier League announced more than one hundred charges against Manchester City, a file that has run for years.

In none of those cases did anything collapse overnight. The financial reports were still published on time. The auditors still signed. The cells in the spreadsheets were still full. The trouble is that a system can only speak about what it has been programmed to see. A framework with no field for doubt will always return the finding that there is no doubt.

The same happens on the pitch. A team keeps three consecutive clean sheets and looks as though its defence has settled. But if across those three matches the opposition managed two shots on target in total, then the clean sheet measures luck rather than structure. In the fourth match, against a real attack, that defence falls apart, and nobody understands why. Nobody understands because nobody reread the data that never made it into the report: entries into the box, defenders beaten, times the goalkeeper had to leave his line.

A football team is not merely eleven people; it is a system of equations that knows how to run. And a system of equations is only trustworthy when we know which variables were left out.

The cup and the illusion of systems

I have a long-standing suspicion about the way people tell the story of small clubs making deep cup runs. On 11 May 2026, Wigan Athletic beat Manchester City one-nil in the FA Cup final, and in the same month they were relegated from the Premier League. In the 2026-13 season, Bradford City, then in the fourth tier of English football, eliminated Arsenal on penalties and Aston Villa over two legs to reach the League Cup final, where they lost five-nil to Swansea on 24 February 2026.

Those stories are beautiful, and they are usually told as proof of a sound system. Look closer and most of them are the product of a favourable draw plus one evening on which every shot found the right corner. A single explosive match does not prove the model right. It proves only that football, on one particular night, allowed something unrepeatable to happen.

When a small club goes deep, people immediately go looking for data to explain it. But data cannot explain a ball striking the post and bouncing in during the ninetieth minute. It can only record that the event occurred and assign it a small probability so everyone can nod along.

The human eye and the machine eye

I have often been challenged for writing too slowly. In 2026, in Kazan, during the match France won four-three against Argentina on 30 June, I did not write the scoreline in my notebook. I noted the moment a body changed direction. In the sixty-fourth minute, Kylian Mbappé received the ball near the halfway line, and over the following twelve seconds or so, the Argentine defence dissolved in a way no statistics table can describe. My editor complained the piece had no numbers. A young coach in Ligue 2 called to ask permission to use it as teaching material.

Mbappé's speed is not for running; it is for cutting a blade across time. When he accelerates, he does not go past one particular defender. He goes past an entire way of playing, an entire generation of defenders trained to read the game by retreating. That is what I wanted to write, and it is exactly what a data model would never suggest I write, because it has no field called time.

The pandemic did not destroy football; it left the body behind and let the soul find its own way home. In 2026, when the stadiums closed, my freelance contract was cut and I withdrew into a study of forty-seven classic matches from 2026 to 2026. I made a note about something I had never noticed before: the crowd is an instrument. When the noise is gone, players communicate with their eyes far more, and passes become slower by roughly a tenth of a second. No automated report told me that. I had to sit and listen.

Based on my experience watching matches over thirty years, I can say one thing almost with certainty: the most important signals usually come from details that have never been named. A defender turns his head to check on a teammate twice in ten seconds. A midfielder takes the ball and hesitates half a beat before passing sideways. A coach rises from the bench and sits straight back down. No system records any of it, because nobody ever thought to give those things a name.

The transfer market, where there is no referee

The transfer market is the domain where the empty-report problem becomes most dangerous. The transfer market is a match with no referee, in which every number is a free kick. There is no whistle. There is no offside. There are only numbers placed next to each other and a signature at the foot of the page.

A club receives a thirty-page scouting report on a player. It lists height, weight, minutes played, key passes, expected goals per ninety, market value according to a valuation website. Every field is filled in. And not one field answers the only question that matters: how will this player behave when his team loses four straight matches in December, when the temperature drops to three degrees, when the stands jeer, when his wife cannot speak the local language.

That gap is not a failure of data. It is a failure of believing data covers everything simply because it covers every field. Expected goals measures the quality of a shot under average conditions. It does not measure a referee's standards in a tense derby, it does not measure how high a team must push when chasing a game, and it certainly does not measure a twenty-three-year-old playing his fiftieth match of the season.

The same logic extends further out. In esports, people calculate the average career length of a competitor and know it is far shorter than that of a footballer. Yet the youth system and post-retirement support in that field are close to non-existent. The spreadsheet is full; the safety net is empty.

What remains

That night, I reopened an old piece of mine from 2026 about André-Frank Zambo Anguissa. I was thirty-seven then. The desk had assigned me a short piece on the home team's pressing under coach Rudi Garcia. I did not write the short piece. I sat down and logged one hundred and twenty-seven ball recoveries by Anguissa in the opponent's third of the pitch, then wrote a three-thousand-word piece on my personal blog calling pressing a net of rhythm. It was shared one thousand four hundred times, more than any magazine article that year.

I tell that story to make a point: a number only has value when it is placed inside a specific story about a specific person in a specific moment. Remove the story and the number becomes noise. Remove the person and the number becomes a polite lie.

The dream of football never lives in the result; it lives in the instant before the ball touches the ground. That instant appears in no report, because it exists only in the eyes of someone sitting close enough to see a player's knee drop before he shoots.

The forty pages on my desk this morning will not disappear. They will grow longer, thicker, more detailed, and there will be more risk matrices that nobody dares skip. In a few years, a generation of young editors will read them before watching the match, instead of watching the match first and reading afterwards. I am not sure whether that is good or bad. I am sure of one thing: a complete report will never replace sitting long enough to see the thing nobody thought to count.