Football Analysis in the Algorithm Era: Perfect Frameworks and Empty Evidence

On 2 July 2026, in Rostov-on-Don, I sat in row eleven of the mixed zone, my notebook open at a page reading 2-0. Genki Haraguchi and Takashi Inui had just put Japan two goals ahead of Belgium within seven minutes of the second half. Around me, the Belgian supporters had gone quiet. Roberto Martinez left the technical bench, spoke quickly to his assistant, and Marouane Fellaini began warming up on the touchline.

Forty-two minutes later, that page was full of lines that had nothing to do with tactics. I recorded the sound of the breathing beside me, the way a Japanese defender bent down to pick a tactics slip off the grass, and the fact that I forgot to open the statistics app because my hands were shaking.

The final score was 3-2 to Belgium. The decisive goal came in the 90th minute plus four, from a counter-attack that organisers measured at roughly 24 seconds: Thibaut Courtois gathered Japan's corner, threw the ball upfield to Kevin De Bruyne, the ball travelled through Thomas Meunier, past a dummy from Romelu Lukaku, and finished at the far post with Nacer Chadli's tap-in.

In those 24 seconds, the entire analytical framework I had brought with me became meaningless.

Every major tournament produces a particular kind of document: the analysis table. The columns list categories. The rows list conclusions. A reader skimming it sees completeness, professionalism, the impression that everything has been verified. The deeper a national-team tournament compresses emotion, the more this document type multiplies, because everyone wants an opinion immediately after the final whistle.

I have a habit of reading such tables backwards. I start with the data-source line at the bottom, and only then read the conclusions. Six years of working in Barcelona taught me something simple: most analysis tables floating online contain not a single original data point. They have a skeleton, section headings, technical vocabulary, and inside each cell is a sentence restating what everyone already saw on television.

In June 2026, I spent nine months following a seventeen-year-old midfielder at the La Masia academy, a boy with twelve appearances for the B team that season. Outlets competed to write sensational pieces about him. I cross-checked his match data against the precedents of five young talents in the same position across the previous ten years, verified across three sources, and kept daily notes colour-coded. When the long-form series was published, a young coach at the club wrote to confirm that every single data point was accurate.

Trust works that way. It does not come from speed.

To understand why these hollow documents are dangerous, you have to look at the modern game's data supply chain. Upstream are the companies that collect match events, logging every pass, every duel, every metre run. In the middle are the distribution platforms, where data is packaged into indices. At the end sit the two groups of customers who pay the most: clubs and betting companies.

The second revenue stream is larger than the first. Live data sold to bookmakers is the least-discussed dark side of football's digitalisation, because it turns every passage of play into a variable that can be wagered on within seconds. When a writer states that team X "is controlling the game" without a single metric attached, the reader has no way to distinguish an observation from the stands from a blurred copy of an odds table.

An analytical framework only has value when every cell in it traces back to a specific data point, with a source, a date, and someone accountable for it. Without that, the skeleton is just a handsome form of silence.

Over the past few years, the cost of producing a professional-looking analysis table has fallen to near zero. A language model can assemble nine sections, full tables, and correct terminology in seconds, even when the input is a single headline. This is the technical reason why skeletons multiply while evidence thins: producing form is cheaper than producing content, and the reader pays in time.

I have held exactly that skeleton in my hands. A nine-section document, each section with tables, terminology, risk classification by level. The headings were all present: tactical and technical analysis, club finance and transfer market, results and public-opinion cycles, league context and team positioning, rules compliance and governance, management and dressing room, risk profile, media narrative and expectations, industry transmission. Read closely, every cell said the same thing: insufficient information to assess.

What stood out lay elsewhere. That document openly acknowledged it had no content, and devoted a whole section to warning about what it called "false authority" — the risk that a complete-looking skeleton gets mistaken for a real piece of analysis. I read that line and thought of the hundreds of tables I have seen online, the kind that never admit to being empty.

European football already has precedents showing that the gap between skeleton and evidence causes real damage. In February 2026, the Premier League charged Manchester City with 115 breaches of financial rules, spanning seasons 2026-10 to 2026-18. In November 2026, Everton were deducted 10 points for breaching loss thresholds; the deduction was later reduced to 6 points on appeal. In March 2026, Nottingham Forest were deducted 4 points. Each case is a file hundreds of pages long, with dates, contracts, balance sheets, hearing transcripts.

Yet most of what fans receive about these cases is a few unsourced summary lines. The framework is there. The evidence gets stripped out somewhere between the courtroom and the timeline.

Telling them apart is fairly concrete. A trustworthy analysis can always answer four questions: what event occurred, on what date, which specific metric, and who confirmed it. Not all four are always required, but at least one must be answered with a clear verification address.

Back to Rostov-on-Don. Reconstruct the match through a skeleton and you get a tidy table: Japan lead 2-0 at minute 52, Belgium level at 2-2 at minute 74, Belgium win 3-2 at minute 90 plus four. The trend column will read: Belgium came back through height and aerial balls. The conclusion will be neat: fitness and aerial duels decided it.

That reading skips precisely what decided the match. Belgium's first goal came from a set piece, a Jan Vertonghen header on 69 minutes. The second also came through the air, Fellaini heading in Eden Hazard's cross on 74 minutes. But the third goal, the killer, involved no aerial ball at all. It was a 24-second counter starting from the very situation Japan had prepared for most carefully: a corner kick.

Akira Nishino's side had spent a great many hours training set-piece defending. In the 90th minute plus four, what beat them was not the opponent's height, but the decision-making speed of Courtois and De Bruyne. The decisive gap in that match lay in time, not in space. No column in a standard statistics table measures it.

I was in the mixed zone that night, one of three female reporters present. I watched the Japanese players collapse, and I watched their coach pick a tactics slip off the grass. I once believed data was everything. That night I understood that tempo and mental state can shatter any statistical forecast, and that most models have no variable for the moment a team loses its capacity to adjust.

In Moscow, I learned that a match can end, but its echo does not. Three years after Rostov-on-Don, when the pandemic halted the leagues, I stayed with a second-division club I had been following since before. Across one hundred days of empty stadiums, I phoned twenty-seven players and recorded their diaries: sessions in living rooms, matches on rooftops, frozen contracts, depressions nobody logged, men forced into early retirement at twenty-nine.

No data table on the market measures those things. The metrics kept running: passes, distance covered, ball-recovery rates. But with the stands empty, I could hear the coach shouting more clearly than the ball rolling, and I understood that what was being measured was no longer what was happening.

Not long ago, I reread my notes from those nine months at La Masia. There, every session was identical in structure, in drills, in duration. But that boy changed every day, in a way no table could capture. That seventeen-year-old did not need me to believe in him. He needed me to stand still and see.

The counterintuitive part sits here: a piece of writing full of emotion and error is easier to detect than an empty table presented beautifully. Readers have reflexes for tone. They have no reflexes for format.

When a document wraps itself in the form of a professional report — clear sections, standard terminology, risk warnings graded by level — that form itself operates as substitute evidence. The feeling of "this appears to have been verified" is far stronger than the reality of "not a single line was verified". That is why the things that look most like analysis are often the things that most need to be questioned.

The football industry has advanced enormously in explaining the past and far more slowly in predicting the future. We can reconstruct a finished match in astonishing detail, and still be routinely wrong about the next one. That gap is not a failure of data. It is the consequence of choosing to measure what is easy rather than what matters.

I am not against data. I am against using data as a ritual. Every club has someone singing; only a few clubs have someone listening. A table that records its source, its date, its author, and contains at least one cell admitting "not enough basis" is more useful than a table stuffed with conclusions that trace back nowhere.

The signal I am watching in the coming period sits at the source end. Watch how many analyses of this major tournament state clearly where the data came from, on what date, confirmed by whom. Every season is a cycle of rhythm, and I have learned to count the rests as well — including the rests named "not enough information". Whoever dares to write those three words more often is the one worth trusting.