Nine Dimensions of Analysis, Not a Single Fact: What the Football Data Industry Is Actually Selling Us

Tuesday night, 2:47 a.m., an eleventh-floor apartment in Incheon. An acquaintance from a scouting network sends over a twelve-page PDF titled Stage-2 Deep Professional Analysis, packaged as neatly as a sponsorship contract. I make coffee, sit down, and read from section one to section nine.

Nine sections. Thirty-four data cells. Four financial comparison tables. Two industry transmission diagrams. A glossary at the end carefully explaining xG, xGA, PPDA, PSR, FFP, salary caps, transfer amortization, sell-on clauses, contract years.

And in almost every cell, a single sentence: insufficient information.

No club name. No player name. No scoreline. Not one metric. The only field filled in completely is the domain label: football.

I read it a second time, but not looking for data anymore. I read it the way you read an artifact. And I did not feel angry. I found it beautiful the way a mirror is beautiful.

My career began with being wrong. I was wrong at the Russia World Cup, and it was the best thing that ever happened to me. That shock taught me one thing of value: to tell analysis apart from the appearance of analysis. That PDF, by admitting its own emptiness, became the clearest lesson about this trade I have come across in years.

Football analysis has grown into a publishing industry. In Korea, where I work, each round of K League 1 generates dozens of tactical reports. Data companies. Broadcasters. Deep-analysis accounts posting at midnight. Every one has a frame. Every one has terminology. There is so much of it that I once asked a young colleague which report he remembered best, and he paused and said: none of them.

That does not happen because he is lazy. It happens because most of this content is produced to look like analysis rather than to contain information. The economics of the trade push in exactly that direction. A data company sells its clients formal completeness, not truth. A publisher pays for volume, not depth. An algorithm ranks clear structure, not accuracy. The result is a machine that manufactures documents that are flawless in layout and hollow in substance, and the PDF on my phone is that machine's purest product.

I have read a great many real reports. On August 13, 2026, I still keep the habit of note-taking during every match I watch: touches, pass completion, distance covered. That habit began in August 2026, after Korea drew 0-0 with Iran in a World Cup qualifier. I wrote that Korea could not beat Iran playing as they did, got heavily abused for it, then spent a week rewatching the tape and counting every misplaced midfield pass. I was right about the problem and wrong about how I framed it. Korea had 61 percent possession and only two shots on target.

A number like that has value because it is attached to something measurable. It can be wrong. It can be disputed. Someone can open the tape and say: you missed three passes. That is information.

In 2026, when the pandemic froze football and stadiums reopened in silence, I did not write a nostalgia piece like my colleagues. I sat down and logged 245 Bundesliga and K League matches after the restart. Home advantage fell from 55 percent to 42 percent with no crowd present. That number does not say everything, but it says something specific and verifiable: singing in the stands is a tactical variable, not decoration. An empty stadium is the most honest mirror football has ever had, and I will repeat that with a caveat I once forgot: it is only one of many mirrors. Noise, money, and crowd fanaticism are facts no less important.

Compare two documents: one containing 245 matches and a 13-percentage-point drop, and one containing thirty-four cells reading insufficient information. The first can be shattered. The second cannot, because it never stood up.

That is the point I want to name. An analysis is only worth something when it can lose. An empty structure is an invincible structure, and invincibility is the surest sign of zero.

Looking back at the nine sections of that PDF, I realised the machine had followed every step of the trade correctly. The tactics section asks about system, formation, build-up style. The finance section asks about broadcast revenue, wages, net debt. The transfer section asks about fair valuation, panic-premium risk, contract structure. The governance section asks about salary caps and financial fair play. The media section asks about source credibility and agent motive. This is the question set of someone who actually works in the field. The problem is that no answers were entered, and the machine still published all nine sections, in full format, with full solemnity.

Put another way, it proved something this industry does not want to hear: the modern football analysis template has been standardised to the point where it can run without football.

This is not harmless. It has a direct effect on how we talk about tactics. Gegenpressing was decoded several seasons ago, yet the reports still praise it as an invention. Meanwhile, in mid-table teams, what actually decides results is fitness and running volume, to the point where many matches become track meets with a ball. If your template has no cell for kilometres covered, you will never see it.

The transfer market behaves the same way. Valuation models rate young potential very highly and can barely assess dressing-room chemistry at all. A report full of amortization and sell-on clauses can still miss the real reason a signing failed: the man could not talk to anyone in the room. There is no cell for that in any template I have ever read, and I have read hundreds.

At the top of the pyramid, the sports rights bubble peaked long ago. Streaming platforms are losing money to buy rights, repeating the exact mistake traditional television made two decades earlier. The hollow analysis machine did not create that bubble, but it serves it very well: when you have to fill thousands of broadcast hours and tens of thousands of articles, you need structure, not truth.

So where might I be wrong?

One possibility I have to admit: that PDF might simply be the product of a pipeline failure. A scraper hits a paywall, a parser ingests a bot-block page, and the input goes empty while the output stays fully formatted. If so, I am using a technical incident to indict an entire industry, which is the cheapest hot-take reflex I know. I always have to ask myself that question before publishing: would this piece still be worth reading if nobody argued about it?

But even if it is a technical failure, the failure reveals a design. A machine with no gate that stops it when it has nothing to say is not a broken machine. It is a machine built never to stop.

The second possibility bothers me more: maybe I also sell structure. Every piece I write has a frame, a rhythm, an opening, a close. The only difference between me and that PDF is that I accept the risk of being shattered. I can be mocked live on radio, as I was after predicting Korea would beat Portugal 2-0 through possession at the 2026 Qatar World Cup. They won 2-1 through a 91st-minute Hwang Hee-chan goal. I rewatched forty minutes of tape and found Son Heung-min had run 11.2 kilometres but touched the ball only 38 times. Korea won through a high press, not the possession football I had imagined. I wrote a piece admitting the error with the data attached, and it was shared by sports journalists themselves.

My detractors read every line I write more carefully than the people who love me. They read carefully because I handed them something that can be broken. That is my entire asset, and it does not exist in any report template.

Third possibility: maybe audiences do not buy information, they buy confidence. A twelve-page document with nine sections and a glossary makes a reader feel they have just accessed something professional, even when not a single event is named. If that is true, the empty machine is not deceiving anyone. It is meeting a real demand, and the only deceiver in this story is the hot-take writer standing in the street complaining.

I also have to be careful about a more familiar trap: using Vietnam and Korea as a shortcut comparison. I am not doing that here, because this is not a story about two football cultures. It is a story about a content industry that operates identically in Seoul, Hanoi, Tokyo and Manchester. If I want to test this observation, I have to put a third football nation on the table and see whether the conclusion holds. It holds. That is why I believe it.

Data whispers while the whole stadium is screaming. I learned to listen. But I also learned that some documents are so silent that their silence is louder than any scream.

My prediction, offered so that it can be checked: within eighteen months, football analysis will split into two clear branches. The first sells structure, machine-produced, cheap, infinite, and increasingly indistinguishable from itself. The second sells judgement, expensive, rare, and obligated to come with the capacity to admit error. The first will take most of the market. The second will hold most of the memory.

And a smaller prediction, verifiable within twelve months: at least one major Asian sports outlet will publish a match preview with a footnote admitting it did not have enough data. Not because it is honest. Because honesty, like everything else in this industry, will become a market trend.

My job, and perhaps yours, is to have a question ready for that day: when every report learns to say I do not know, what standard will we use to tell the honest one from the one that is simply selling?