When the Data Sheet Is Empty: The Fabrication Trap in Football Analysis

On the night of 27 June 2026, in Kazan, I sat in front of a spreadsheet that had been running for three months. My xG and xA model, built on three seasons of data from Europe's five major leagues, gave Germany a 78% chance of reaching the World Cup semi-finals. By the final whistle the score was 0-2 to South Korea, and the team I trusted most left the tournament at the group stage. The model got 12 of the 16 knockout qualifiers right. It failed at precisely the point where I had placed the most faith. Since that night, before reading any metric, I always ask one thing: which cells in this sheet are empty, and why do I want to fill them?

It sounds technical, but it is a question of professional conduct. Football analysis runs in two stages. The first breaks the source into discrete information points: team names, player names, scores, timestamps, source reliability. The second builds the deep analysis, but it may only build from those points. When the first stage returns an empty list, meaning no confirmed team, no confirmed player, no confirmed match, the second stage has exactly two options: state that there is insufficient information, or invent content.

In my trade, the second option is more tempting than it looks. A thick report with every section, table and chart filled in always looks more credible than a single page admitting that no assessment is possible. That smoothness is the real danger. Data does not get emotional, but it remembers everything journalism forgets, and it also remembers everything the analyst quietly added.

I learned this a second time in 2026, when European stadiums closed because of the pandemic. I collected data from nine Bundesliga matchdays after football resumed in May. Home win rates fell from 44.2% in 2026-19 to 36.7%; average goals per match dropped from 3.1 to 2.8. Home advantage, which every older model treated as a constant, suddenly depended on whether stands were full. Home is not sacred ground, it is a frozen variable. When the stands change, the old number stops meaning anything.

In 2026 I carried that lesson into the Euro quarter-final between Italy and Belgium. Italy pressed with an average PPDA of 8.2, allowing opponents fewer than nine passes before an intervention. Belgium played on the counter and covered 17% less ground than in previous matches. PPDA is the signature, distance covered is the confession. I concluded Italy would control the game, and Italy won 2-1. For the first time, a model annotated with crowd conditions, fixture congestion and injuries predicted an important development correctly. I did not record that result as a medal; I recorded it as a process check: data, context, prediction, then verification.

By 2026, aged 23, I was tracking Enzo Fernández's move from Benfica to Chelsea for 121 million euros. My valuation report rested on World Cup data: 82% pass accuracy, 14 successful tackles. But the deal also depended on the agent's role, the structure of payment terms and the buyer's urgency. No metric measures those things. Transfers do not pick the best player, they pick the player you measured wrong the least.

The limits of football data have very specific shapes. A club can satisfy every financial fair play threshold and still collapse under a long-term wage contract signed in a panic. A team can lead the xG table all season and still finish empty-handed after dropping seven points in the last four rounds. A transfer story can come from a journalist with verifiable sources, or from a social media account shared fast enough to become news. Source reliability is an information point, not a decorative detail.

The intuitive reaction of most people in the industry runs the other way. When data is missing, we fill the gap. A striker who fails to score for six rounds is described as being in a crisis of form, when the cause may be that his team stopped creating chances for him. The same number, two stories, and only one of them has a foundation.

I trust variance more than I trust champions. A champion is the output of a sequence of events; a team holding low variance in process metrics is the output of a way of playing. The table tells the story of the past. Process metrics tell the story of the near future, and even then they only tell part of it.

In the player-data industry, fabrication rarely comes from malice. It comes from incentives. A template that is nine-tenths complete always looks more professional than one that is six-tenths complete, and nobody grades the empty part. That is why a sheet stuffed with numbers but missing a data-limitations section is usually a bad sign.

Data is a foundation, not absolute truth. In the coming annual-season cycle, the signal worth tracking is not which team leads the table, but which team starts publishing its empty cells too. Germany 2026 was a gift, because it proved that models also need to fail in order to grow. An honest analysis begins by stating clearly what it does not yet know.