The Empty Data File and Football's Habit of Filling the Gap

Late one November night in Shanghai, I opened a file and found it empty.

No competition name. No club name. Not a single number. Only nine section headings queued up to be filled: tactics and technique; club finance and the transfer market; results and the cycle of public opinion; league landscape; rules and governance; the coaching staff and the dressing room; risk profile; media narrative and expectations; the transmission chain of the whole industry.

The system had run. The input was empty. The output still came out, and in all nine sections one line repeated: "insufficient information to assess".

I sat looking at that file longer than necessary. It was worth more than every complete analysis I have written in thirty-three years, because it dared to say the thing football commentary almost never dares to say: I do not know.

My profession lives by filling the gap. What that gap is, who pays for it to be filled, and what the price is — that is the story of this piece.

Two streams of money and one silence

Modern football runs on three streams of money: media rights, sponsorship, and the transfer market. All three need the same fuel: story. A match cannot sell itself. Neither can a contract. Someone has to tell it, and the teller has to finish before the match does.

In June 2026 I was sent to Moscow as an on-site commentator. The 2026 World Cup did not begin with the ball. It began with the fear of being forgotten. Hundreds of broadcasters, thousands of reporters, all of them needing an angle nobody had yet. The pressure was not in understanding the match. The pressure was in having to say something, fast.

In the France–Croatia final, in the eighteenth minute, I noticed Griezmann standing right beside the free kick on the left angle, exactly the spot from which he had curled the ball into the box seven times before. I said on air that the ball would go between the penalty spot and the post, that Mandžukić would clear it and put it into his own net. That is exactly what happened. My colleagues were stunned; I was not, because I was not looking at the screen. I was looking at the data table I had built in my head.

But the story I want to retell is not the story of being right. It is the story of being wrong. The first time I failed on the big screen, the audience forgot. I did not.

In July 2026, in the derby between Shanghai SIPG and Guangzhou Evergrande at Hongkou Stadium, I called Hulk's name wrong three times in the first half. The forums mocked me. That night I did not explain myself. I reopened the tape, counted every touch, every pass, every shot by the Brazilian forward, then built an Excel sheet cross-checking them against the movement of the opposing defence. Since then I have written by a single rule: data first, emotion second.

That rule sounds simple. It is not, because most of the time the data is not enough.

xG and the limits of frequency

Expected goals, xG, does not predict who wins. It measures the quality of chances. A team can generate 2.4 xG and still lose 0–1. A team can generate 0.6 xG and still win 2–0. In a single match, the result is close to random. Over thirty-eight matches, it starts telling the truth.

This is where commentary slips. People take one match to prove a model, then take a model to explain one match. Both directions are equally wrong.

In the summer of 2026, global football stopped because of the pandemic. Broadcasters cut forty per cent of staff. I lost my live commentary contract. Instead of waiting, I downloaded the full event dataset from StatsBomb and wrote Python code myself to rebuild Liverpool's pressing model from the 2026–20 season. When the Bundesliga returned in June, I tried predicting results using xG and sprint counts. I got eleven of fourteen right.

Eleven out of fourteen sounds good. But I learned most from the three I got wrong. In all three, I had imposed a rule of frequency on an event that happens only once: an injury in the eighth minute, a red card nobody saw coming, a referee's decision. Those three matches taught me something I still use every day: when a result falls outside the data series, do not delete it. Write it down.

The transfer market: the buyer picks the wrong reason to be right

Transfers, in the end, are the story of a buyer choosing the wrong reason to be right.

Every summer, some club pays above a player's market value in the final twelve hours of the window. I call it the panic premium. It does not come from tactical need. It comes from the need to reassure the audience.

How clubs handle that number is worth watching too. Long contracts let a club spread the transfer fee across the years. A deal worth one hundred million dollars signed over eight years sits on the books at roughly twelve and a half million a year. In the summer of 2026, UEFA closed that route with a five-year cap. An accounting rule changed how an entire league buys players.

This is where data and emotion touch. Nobody buys a player because of a balance sheet. But the balance sheet decides who is still allowed to buy next season.

Financial rules are a data system

The Premier League operates its Profitability and Sustainability Rules, known as PSR. Everton were docked ten points on 17 November 2026, reduced to six on appeal in February 2026, then given two more points in a second case that April. Nottingham Forest were docked four points on 18 March 2026. Manchester City face one hundred and fifteen charges dating from February 2026.

Fans look at that and see punishment. I look at it and see a data pipeline. The rulebook is only correct when the input is correct. When a club misstates one item, the whole chain of conclusions downstream bends with it. That is why I read financial rulings the way I read audit reports, not the way I read league tables.

Media rights: the marriage nobody likes

Media rights are the marriage nobody likes, but everybody waits to see the paperwork.

In June 2026, Apple signed a ten-year deal with MLS worth two and a half billion dollars, starting from the 2026 season. It was the first time a technology platform bought an entire league as a package rather than buying local rights deal by deal. On the other side of the world, the Saudi Pro League spent close to a billion dollars in a single season to pull stars in. Two moves, two logics: one buys long-term stability, the other buys short-term attention.

The real question is not the value of the contract. It is this: who owns the audience data once the contract is signed. That is the only asset in this industry that does not lose value with the season.

Vietnam: a data gap at home

I was born in Vietnam and work for the Chinese market. That distance gives me a view few people have: looking at Vietnamese football through the eyes of a market that is a few steps ahead on data infrastructure.

On 5 January 2026, in Bangkok, Vietnam beat Thailand 3–2 and won the Southeast Asian championship. The naturalised striker Nguyễn Xuân Son was the face that emerged from that campaign. A month earlier, the road to the 2026 World Cup had ended differently: Indonesia advanced, Vietnam stopped.

Those two results are usually told as two separate stories. To me they are one. Both are about the same thing: the ability to measure. A regional tournament can be won with nerve and a few moments. A qualifying campaign lasting eighteen months needs data, squad depth, and a stable player-tracking system.

I stand between revenue and emotion, and I have learned that whoever holds both is the one who wins.

The V-League has crowds, players, stories. What it lacks is the infrastructure to turn stories into evidence. At many clubs, match data is still recorded by hand on paper. Nobody builds a scouting model on paper.

On the commercial side, a league only attracts rights money when it can prove the size of its audience with numbers rather than with feeling. Broadcasters pay for measurable ratings, for behavioural data, for the ability to resell advertising. A league that cannot measure itself cannot price itself.

The counterintuitive angle: emptiness is data

This is the part where I want the reader to stop.

The empty file I opened that night was the most correct result of my entire working week.

Football analysis in 2026 suffers from a disease: it produces the shape of rigour without the substance. Language models can write a nine-section analysis in thirty seconds. It has the right headings, the right tables, the right terminology. It is missing exactly one thing: the underlying data. And because the shape is right, the reader cannot tell it apart from a real analysis.

This is why I treat the ability to say "insufficient information" as the most expensive skill in the trade. A system that stops when the input is empty is worth more than a system that always returns a conclusion.

For the same reason, I do not believe romantic stories about a small town beating a giant. They do happen. But they hide the financial gap behind them. A small club beating a big club once is the business of a single evening. To do it twenty times in a season, the small club needs a better data system, not a bigger heart. Emotion wins a match. Data wins a season.

Ending

At 49, I am still rewriting the script of my own career. Not to be different, but to survive.

Thirty-three years ago I started in the sports department of a television station, with a notebook and a pen. Today I work with event data files, xG models, and balance sheets. The tools change. The principle does not: say only what you can prove.

If you read a football analysis where every section has a conclusion, ask one question: where is the underlying data. If there is no answer, that analysis is a product, not knowledge.

The fans, the people who pay for all of it, deserve to know what they are buying.