The Art of Saying 'Not Enough Data': A Lesson from Kazan 2026

Kazan, June 27, 2026. Minute 93. Kim Young-gwon puts the ball past Manuel Neuer. In the press tribune, more than two hundred writers reopen their drafts at once. I had filed fourteen minutes earlier.

What I remember most about that night is not the roar in the stadium but the frantic clatter of keyboards from colleagues rewriting from scratch a story they thought they understood. Everyone already had a conclusion saved on their laptop. The problem was that the conclusion had never had a single piece of data holding it up.

My piece opened with a dry metric: Germany's PPDA at the 2026 World Cup group stage stood at 15.2. Put plainly, every time a German player committed to pressing, his side had already allowed fifteen passes. That number does not say Germany were weak. It says Germany were slow. And on the other side of the pitch, Son Heung-min was being placed into exactly that gap.

Nobody wants to read an article that opens with PPDA. But by the time the match ended 2-0 and Germany left the tournament bottom of Group F, that piece had drawn 120,000 reads, the highest in the newsroom that week.

I tell this story not to praise myself. I tell it because of one detail in it that most of football media walked past: the most useful analysis I have ever written was not the one with the boldest conclusion. It was the one where I dared to state clearly what I did not know.

Professional football runs on belief in stories. A team that wins three in a row is called mentally strong. A team that loses three in a row is called a dressing-room crisis. Both labels are stuck onto the same dataset, and that dataset usually contains nothing but scorelines.

In 2026, in my first month as an intern at a sports outlet in Seoul, I filed an analysis of the K-League. I counted every set piece, logged the timing, reconstructed the ball's path, and concluded that the champions' share of goals from dead-ball situations sat well above the league average. An editor returned the draft with a line I still remember verbatim, to the effect that women knew nothing about tactics. I did not argue. I rewatched every minute of footage, annotated each dead-ball phase, and attached an appendix stating the sources and the method. The piece ran.

Since then, every article I write carries a small section at the end: sources and method. Not as a shield. So readers can check the work themselves.

In 2026, when the pandemic emptied stadiums and the outlet's revenue fell seventy percent, the desk met to discuss speculation pieces, the kind that ask what would have happened if the season had not been suspended. I turned the assignment down. Instead I sat alone and rebuilt a match database covering more than six hundred games already played, typing in each figure by hand, cross-checking each source.

The whole world stopped turning, but my ghost football database kept breathing.

Back to Kazan. What made that analysis hold up was not a scoreline prediction, because I made none. It was that I connected three separate layers of data that nobody had placed side by side.

The first layer was pressing intensity. A PPDA of 15.2 put Germany among the slowest teams in the group stage at winning the ball back. Joachim Löw's side controlled possession well, but controlling the ball and controlling space are two different jobs. Losing the ball in the middle third left the German back line retreating a step too late.

The second layer was defensive-line height. I measured the average position of Germany's four defenders on every ball loss and found enormous variance: at times the line pushed up to the halfway stripe, at others it dropped almost to the edge of the box. That inconsistency made the space behind the full-backs appear on a rhythm, as regular as a trap setting itself.

The third layer was the opponent file. Son Heung-min had just come through a Premier League season with a top speed among the fastest in the division. A fast player who likes receiving the ball in the space behind a defence, meeting a high line that cannot hold its shape. The match-up was almost embarrassingly simple.

Germany did not collapse for lack of talent. They collapsed because nobody read the whisper of the numbers.

The bridge between the two football cultures I follow daily has a feature few people notice. Metrics in Europe are recorded densely; data in Asia is often sparse and inconsistent in how it defines things. When I compare the PPDA of a K-League side with that of a Bundesliga side, I have to rewrite the definition of a defensive action so the two align before the numbers can sit next to each other. Skip that step and I produce a conclusion that sounds persuasive and means nothing.

Data does not stop at tactics. It speaks about money too.

In recent transfer windows I have spent most of my time on a category of deal the media barely notices: the loan with an obligation to buy. On paper it is a loan. Financially it is a completed transfer, recognised a year later.

The structure has spread because it solves a problem for both sides. The big club defers the outlay, keeps cash inside the current accounting period, and still knows the player is coming. The small club receives money in the future but pays for it upfront in risk. For twelve months the player still wears their shirt, still takes a starting place, still trains as a long-term asset. By the time the payment lands, it is too late: the squad was built around someone who is no longer there.

In the Premier League, Chelsea paid 106.8 million pounds for Enzo Fernández in January 2026, then went to 115 million pounds for Moisés Caicedo in August of the same year, a British record at the time. Numbers like that set the reference frame for the whole market, and every smaller club gets dragged along. Brighton sold Caicedo and reinvested in players they bought for a fifth of the price. Most other clubs do not have that scouting system. They have an obligation to buy.

Their breaking point is not in the dressing room. It sits in the third column of the spreadsheet I filter.

This is where I want to talk about what technical documentation calls null handling, and what football calls admitting you do not know.

When I built the match database, I followed one iron rule: if a cell has no trustworthy source, it stays empty. No estimated values. No league averages used as filler. No plausible-sounding invented figure.

That rule makes the table far uglier. White gaps appear on screen, and each gap is a question without an answer. But those gaps are exactly what saved me from drawing a wrong conclusion during a transfer window.

What the ghost database taught me was not the volume of matches. It was that I know precisely which matches I have no data for. The list of matches with missing information is nearly as long as the list of complete ones, and I keep both in the same file.

Applied to football, the principle is simpler than most people assume. A scouting report that states the author lacks sufficient data to judge a player is worth more than a three-page report concluding the player has potential. The second creates comfort. The first creates work to do.

At match level it is clearer still. If a team concedes twice from long-range shots while its opponent generated just 0.4 xG across the match, the probabilities say that result does not reflect the quality of the performance. But if I have no shot-location data, I have to say I have no shot-location data. I am not allowed to infer it from the score.

People watch goals and cheer. I watch a seventeen-minute probability chain to understand why it happened.

In football analysis, the thing more dangerous than a wrong number is a number with no source.

A wrong number can be caught. It has units, a sample, a method, and sooner or later someone will check it. A number without a source cannot be caught, because there is nothing to check it against. It exists as a belief, and belief travels faster than data.

I once spent a week tracing a distance-covered figure for a midfielder that had been quoted widely on social media. More than twenty articles repeated it, and not one pointed back to original data. They all pointed at each other. The number was not wrong. It simply did not exist in a checkable form.

Football analytics is committing a systemic error, distinct from a technical one. We calculate far better than we did a decade ago. We build models ordinary fans cannot read. But we have not built the habit of stopping when the data is empty. With nothing to analyse, the industry's default reaction is to analyse by feeling, then call that feeling expert intuition.

I keep two categories of numbers strictly apart: numbers that prove, and numbers that have not yet answered. The second kind is not worth less. It points to exactly where my understanding is blank, and that is always the most productive place to dig.

Data practice is not about prophecy. It is about never being fooled twice by the same lie.

At thirty-three, I believe every number is a witness that never perjures itself. But a witness only speaks when someone asks the right question.

The season is moving through its rounds, and the signal I am tracking is not in the table. It is in the scouting files left empty over the past month, in the clubs that declined to sign a player for lack of data instead of buying on a hunch. If the number of empty files rises in the coming weeks, that will be the healthiest sign professional football has produced in years.

And if every file is full, check what it was filled with.