The Blank File in Mid-Season: What Football's Data Tables Never Say

Wembley, the night of 11 July 2026, minute 120. In the middle of my notebook sat a blank square, the spot where I usually sketch expected goals. The Euro final between England and Italy ended 1-1 after extra time, and Italy won 3-2 on penalties. Bukayo Saka walked to the spot, and I could not write a single line. I closed my laptop and went down to the lower tier, where England supporters sat on bare concrete. A father held his small daughter. The child had no idea what had just happened; she simply reached up and stroked her father's cheek. A child wiping her father's face after three missed penalties — that is how football teaches people to live. That night I wrote three thousand words without naming a single player who missed. The blank square in my notebook stayed blank. The piece was shared more than twelve thousand times, and it taught me something no model had ever told me: what holds a reader is not in the spreadsheet.

In 2026 I graduated from the Journalism Academy, joined Bong Da newspaper, and picked up a correspondent's card for The World of Sport in Madrid. The newsroom then had a fax machine, a yearbook, and a shared telephone. Thirty-five years later, every outlet I work with runs a live dashboard: expected goals, passes per defensive action, PPDA, progressive carries. Every editor asks the same question: where are the numbers? And every time I file a long piece that opens with a breath rather than a figure, someone reminds me that readers in 2026 want data.

This season, in the third round of fixtures, I opened a research file for a deep-dive analysis. The first page read: headline — none; source — none; core arguments — not provided; entities involved — not identified; time sensitivity — not assessed; source quality — not assessed. All nine analytical dimensions carried the same verdict: insufficient information, cannot assess. A perfect blank file, as clean as a pitch before kick-off.

My first professional reflex was to fill it with guesses. That reflex was trained into me, and it is dangerous. Based on my experience watching matches across several World Cups and European Championships, I know that an analysis built on data that does not exist is simply an invented match presented in a confident voice. So today I am doing the opposite: writing about the white space itself, and about three places where modern football is still measuring on samples far too small.

In 2026 Liverpool won the Premier League for the first time in thirty years, and there was no trophy parade. On 25 June that year I drove around the city at midnight, filming supporters standing in their doorways, lighting candles and singing You Will Never Walk Alone through phone speakers. The candles that year did not light Anfield, but they lit an entire season without crowds. No dashboard measured that. The stadium became an absent character, and that absence carried more emotional data than any model records.

France against Argentina on 30 June 2026 is the mirror image. France won 4-3; a nineteen-year-old Mbappe scored twice and won a penalty. Mbappe does not run past defenders; he runs past the prejudices of an era. Yet while colleagues around me logged sprint speeds, formations and pass counts, what I recorded was the hush that fell over the stands in Russia with each of his runs. It took me years to understand why my memory of that match is thicker than any data sheet I have ever read.

Now to the hardest part.

The transfer market has a blind spot about sample size, and that blind spot costs more than any tactical error. In July 2026 Joao Felix joined Atletico Madrid for 126 million euros at nineteen, after barely more than one top-flight season at Benfica. In January 2026 Mykhailo Mudryk moved to Chelsea in a package that could reach around 89 million pounds, with fewer than fifty top-flight appearances. That same month Enzo Fernandez joined Chelsea for around 121 million euros, largely on the strength of one World Cup. Three files, three valuation models built on a sample smaller than any confidence interval I have seen in any other industry.

Statistics offers a simple rule: the smaller the sample, the wider the confidence interval and the larger the risk. A nineteen-year-old with forty top-flight matches may be a generational talent, or he may plateau at twenty-two. Data cannot distinguish between those two possibilities. The transfer report, however, always presents them as distinguishable, because the people paying do not buy probabilities — they buy certainty. The transfer market does not sell footballers; it sells dreams priced by fear.

Look at how goalkeepers are valued, where the distortion is even clearer. In August 2026 Kepa Arrizabalaga moved from Athletic Bilbao to Chelsea for 80 million euros, then a world-record fee for a goalkeeper; the same summer Alisson Becker joined Liverpool for around 66.8 million pounds.

At the goalkeeper position, distribution has been sanctified while declining reflexes still command the same price. The reason is structural. A goalkeeper faces roughly one hundred to one hundred and thirty shots on target in a season. With a sample that small, save percentage swings more from noise than from ability, and advanced metrics blur the problem rather than solving it. When two goalkeepers post similar numbers across a season, the better distributor is handed a fee differential that can reach tens of millions of pounds — evidence drawn not from defensive data, but from an attacking skill that is hard to separate from the system around him.

The same holds for failures. A goalkeeper who peaked and then declined retains market value far longer than a striker declining at the same rate. Scouts talk about experience, composure, command of the back line. None of that appears in a spreadsheet, and none of it can be refuted with a spreadsheet. Belief is defended by the very thing it cannot measure.

Esports repeats the same error, but on a more human scale.

Esports is where young men without divine feet still touch glory with their fingertips. It is also where a career is compressed with cruelty. A professional footballer can play at the top for fifteen years; an esports professional usually finishes before or around twenty-five, many at twenty-three. Meanwhile, the youth system in esports is largely built ad hoc by team organisations, and post-retirement support is close to non-existent: no mandatory transition scholarships, no industry pension fund, no standardised pathway into a second career.

In-game data is suffocatingly dense: kills, damage per minute, teamfight participation, retreat indices. Data on what happens after the hands leave the keyboard is almost blank. An industry measures every movement across eighteen minutes of play and measures almost nothing across the forty years of life that follow.

Back to the blank file on my desk. After reading it carefully, I realised it was not empty at all. Nine dimensions marked insufficient information are themselves data: they show that the input source does not exist, that any tactical claim made now would be fabrication, and that the only correct action is to declare the situation unassessable. In my trade, that is a valid and rare conclusion, because it forces me to say what editors do not want to hear.

The 2026 World Cup in Qatar taught me this lesson in its purest form. Before the tournament I did not fly straight to Lusail. I spent three days walking the migrant labour districts of Doha, talking to Nepali workers who had built the stadiums. One afternoon I found a discarded old boot by a fence. On 13 December 2026 Argentina beat Croatia 3-0 in the semi-final; Messi opened the scoring from the penalty spot in the 34th minute. I placed the old boot beside Messi's boots in a piece called Shoes of Two Worlds. Across the whole match I recorded no goals. I recorded my own silence.

The popular story in football today is that data killed the romance of the game. I think the opposite is true, and this is the part I will argue to the end. Data did not kill romance; it created a new kind of romance, in which supporters place their faith in models instead of in curses. The structure of that faith is identical: pick a memorable symbol, ignore the confidence interval, and retell the story as though it had been proven.

The greatest error in modern football analysis is not using the wrong data; it is turning silence into an assumption. When a file contains no information, the newsroom reflex is to fill it with inference and then present that inference in the tone of a conclusion. When a goalkeeper has one good season, we write about a career breakthrough. When a nineteen-year-old scores seven goals in ten games, we write about a generational talent. In both cases the sample is too small to support the claim, but nobody wants to be the first to say we do not yet know anything.

There is an economic reason for that reluctance. An analysis concluding that there is not enough data will not be shared. An analysis concluding that this is the heir will be shared. The system rewards certainty, and the penalty for caution is silence. Thirty-five years in this trade have taught me that most of football's biggest mistakes come not from misreading data, but from being too afraid of white space to endure it for one more week.

There are twenty-five rounds left this season, and I will still be in the stands with a notebook full of blank squares. Every season that passes is a book closing; careful readers find themselves inside it. What I want to know next round is not who leads the table, but who among us has the nerve to say the three words this industry fears most — I do not know — while the page is still white.