The Trap of Empty Data: The False Certainty of Numbers in Football Analysis

May 16, 2026. The Bundesliga returned, but the stands were empty. On television, that unnatural silence — no one in the ground even after a goal, just the players' shouts and the sound of the ball. I opened my notebook at home in London. Of the nine matches in that first full round, the home side won only two. Within seventy hours I had written it down: “Home advantage was never about the crowd.” The claim spread and was shared thousands of times. The problem was that I was wrong. The following year, home win rates reverted almost exactly to their previous level, and I had to issue a public correction. The strange thing is that the correction video was watched more than the original claim.

This piece is about that mistake, but it is not self-flagellation. It is about the biggest trap in football analysis — the one I fell into and the one the whole industry falls into every day. The trap, in plain terms: we draw confident conclusions from empty data, because the shape of a conclusion looks credible. A number from nine matches, a clean table, a clear verdict — and the reader thinks analysis has happened. In reality there was no analysis, only noise.

“Home advantage” is not a new idea in football. Historically, home teams win more; behind it lie the unconscious pressure of the crowd on refereeing decisions, the away side's travel fatigue, a familiar pitch and environment, even the dimensions of the ground. In top leagues, the home win rate usually sits between 45 and 48 per cent, with the rest split between draws and away wins. In other words, four or five of every ten matches go to the home side — that is the long-run picture.

In May 2026, because of the coronavirus, the league returned to empty stadiums. And that is when one claim spread fast: no crowd, so no home advantage. Broadcasters, podcasters, club officials — everyone seemed to speak in the same key. Some proposed cutting season-ticket prices, some said football was now a neutral-venue game, some said the crowd played no measurable role at all. The numbers seemed to testify: in the first round, home sides were barely winning.

But looking at those numbers, one question should have been asked, and nobody asked it — how many matches underpinned all this confidence? The answer: nine. And nine matches can never justify a conclusion that rewrites the character of an entire season.

To understand what a nine-match conclusion means, basic statistics is enough. In top leagues, home teams historically win about 46 per cent of matches. So the expected number of home wins in nine games is roughly four. Winning two means two fewer than expected. In statistical terms that is nothing rare — deviations of that size occur again and again within normal variance. What we thought was a “signal” was pure “noise.”

The problem is not in the numbers. It is in our trust in them.

This is the real lesson. An empty or incomplete dataset, if presented cleanly, still looks credible. And this is not only football's problem. Anyone who builds a report — a sports analyst, a business analyst, a journalist — knows that a table, a few categories and a few figures make a piece look substantial. But if the cells of the table are empty, nothing remains except the beauty of the format. I have seen it many times: an analysis report stacks one confidence level after another — “high,” “medium,” “certain” — while not a single information point sits inside. That is the most dangerous tendency in modern analysis: a complete format with empty content.

I recognise this problem because I am part of the machine. Writing within hours of full-time is my trade. The habit of speed has brought me readers, and the same habit has made me wrong. In May 2026 I believed an empty dataset because it looked dramatic.

The past decade has seen an explosion of data in football. Now every shot has an xG, every pressing sequence has a PPDA, every pass on the pitch is captured by tracking cameras. The numbers have grown, but so has the temptation to misread them. Because the more metrics there are, the more opportunity there is — with one metric or another you can prove whatever you like. This is what statistics calls “p-hacking”; in football's language, it is gathering numbers to fit a claim.

This flood of numbers has a side effect. Readers now believe anything that comes with a number, because numbers feel neutral. But numbers never speak on their own — an analyst makes them speak. With the same statistic, two analysts can reach two opposite conclusions, and both can argue with numbers in hand. That is where the analyst's responsibility lies — in deciding which question the number is an answer to.

So I follow a rule: before publishing any claim, it must pass two tests. First, the economics test — does the argument hold up on the balance sheet? Second, the sample test — are there enough matches behind it? In August 2026, when Neymar joined PSG for €222 million, everyone was saying it was a sign of market instability. I wrote that it was not abnormal — it was the rational price of the last unclaimed global football brand, priced like a broadcast-rights deal. I used to think €222 million was an outlier. Then I watched the whole market copy it. The lesson is clear: the heavier the claim, the heavier the accounting behind it must be.

In the same way, after spending twenty-six days in Russia for the 2026 World Cup and attending eleven matches in person, I understood that the set piece is not a passing phase. England beat Panama 6-1, with five of those goals coming from dead balls; across the tournament, nine of England's twelve goals came from set pieces. After watching that match, my notebook read: the tournament is now a set-piece sport. Notice that the conclusion came from eleven matches watched in person and from my own notebook — not from the mood of a television studio.

That does not mean small samples are always false. Sometimes a shock really is a signal. On 22 November 2026, in Lusail, Qatar, Argentina lost 2-1 to Saudi Arabia — their first World Cup defeat to an Asian nation. Within six hours I wrote that Argentina would still win this World Cup. Before the tournament