The Empty Report: When Sports Analysis Is Generated From Data That Does Not Exist

The Empty Report: When Sports Analysis Is Generated From Data That Does Not Exist

2:47 a.m., Seoul, quiet enough that I could hear the fridge click off between two breaths. On the screen was a forty-one-page PDF sent by an analytics group. Bold title. Table of contents. Neatly ruled tables. A whole chapter titled "Risk Assessment." A small line in the footer stating that this document did not constitute betting advice.

I read from page one. By page forty-one, I realised I had just read a completely empty document.

No tournament was named. No team. No player. No patch, no date, no source. Every cell in every table carried the same sentence: "Insufficient information to assess." But the column headers were still there. The risk hierarchy was still there. The closing summary still rated information quality on a five-star scale, and every category received one star, with a note explaining that the star reflected "the diagnostic value of confirming failure."

Forty-one pages. Three thousand nine hundred words. Actual information: none.

I sat in the empty stadium of my own profession and heard myself more clearly than ever. In the empty stadium, I heard myself more clearly than ever.

What kept me awake was not that the document was wrong. It was not wrong. It was not right either. It simply sat there, formatted so beautifully that if you skimmed it on your phone while waiting for the subway, you would nod and think you had just read professional analysis.

That was the moment I understood: the biggest problem in sports content this year is not a lack of data. The problem is that we have built a machine that manufactures the appearance of professionalism even when there is nothing inside.

The report factory

To understand why a file like that exists, you have to look at the economics of sports content in East Asia, where I have worked for close to a decade.

A sports writer in Seoul in 2026 is not competing with other writers. She is competing with an algorithm. Search platforms this year reward what they call "information gain" — the amount of new information an article adds relative to everything already online about the same subject. This is a good standard. It punishes recycled work. It forces writers to bring a perspective nobody has had.

But a good standard, when it meets a high-volume production system, produces a bad behaviour: manufacturing the appearance of new information. I call that the report factory.

The report factory runs on four stages. Collect. Summarise. Format. Publish. None of those four stages requires a single fact. They only need to produce a product that looks like it contains facts.

In football, this pressure has been building for fifteen years, ever since optical and event data became cheap. A single match now generates hundreds of thousands of data points. Nobody reads them all. So the industry created a middle layer: people who translate data into narrative. That middle layer is real, valuable, and contains people who do the job very well.

Esports took that road three times faster.

I cover esports for a Korean audience. The rhythm of this industry is far crueller than football. Some titles patch every two weeks. Some competitions run almost year-round, season after season, region after region. A team can play three matches in seven days, each on a different patch, in three different formats. There is no summer break. There is no transfer window long enough for the news to settle.

That rhythm creates a thirst for content that can never be quenched. And that thirst feeds the report factory.

I have seen that factory from the inside. In 2026 I started out as an esports player and then a tournament organiser, before moving into media. As an organiser, I was the one commissioning content. I once paid for "deep analysis" pieces that I knew full well had been written in thirty minutes using a public spreadsheet.

The frightening part is that those pieces performed well. They were shared. They were quoted. They drew viewers to my tournament.

I am not telling this story to accuse anyone. I am telling it because I was a link in that chain, and because I believe an industry only fixes its flaws when someone stands up and says the flaw starts with them.

Three layers of failure

After reading that forty-one-page file, I spent two weeks reconstructing its path. What I found was that the problem was not in one place. It was in three layers, and all three are fixable.

Layer one: a data pipeline that does not know it is empty

This is the technical layer, and the easiest to fix.

A modern sports analytics pipeline usually has two stages. Stage one extracts: it reads the source article, pulls out information points, identifies entities, assesses time sensitivity, scores the source. Stage two analyses: it takes stage one's output and expands it across multiple analytical dimensions.

The problem is this: when stage one fails, it does not raise an error. It returns a valid but empty structure. An empty list of information points. Entities unidentified. A blank one-sentence summary. Article type recorded as "unclassified."

And stage two, instead of halting, keeps going. It still outputs all nine analytical dimensions, all the tables, all the rating scales. Because the input structure is valid. Because no validation gate stops it.

I once asked an engineer about this. He said something I have never forgotten: "Systems are not afraid of bad data. Systems are afraid of empty data, because empty data triggers no alarm at all."

In esports, this failure is far more lethal than in football. Football has a fixed object: a ball, two goals, twenty-two players. You can infer a great deal from the structure of the match. Esports cannot. An esports analysis that does not identify which game it is about cannot infer anything. You do not know which patch is live. You do not know the tournament format. You do not know which region is strong. You do not know whose publisher sets the transfer rules.

Without the game title, every conclusion is fabrication.

And that report did exactly half the job: it did not fabricate. It simply left blanks. But it left those blanks inside a shell pretty enough that readers could not tell they were blanks.

Layer two: readers cannot tell the difference

This is the hardest layer, and the one I belong to.

If you hand an ordinary reader two texts — one genuine analysis, one empty analysis formatted identically — the rate at which they misjudge will be very high. Not because they are stupid. Because the signals of quality have been systematically counterfeited.

A good analysis and an empty analysis share almost the entire vocabulary. Both use the words "metric," "trend," "pattern," "risk," "probability." Both break paragraphs short. Both bold a few phrases. Both end on an open question.

The only distinguishing signal is the density of verifiable specifics: proper nouns, numbers with units, absolute dates, citations. And that signal only works if the reader is willing to spend thirty seconds checking.

Thirty seconds. On a two-thousand-word article. In a daily stream of information.

I used to believe sports readers did not need protecting. That they were smart enough to filter for themselves. I was wrong in a subtle way: they are smart, but they are tired. And a tired person does not check sources.

One evening I sat next to a friend who works in finance, and we read the same analysis of a football team. He finished and said, "This is well written." I pointed to three places in the piece, each one a judgment with no basis. He read it again, was quiet for a while, and said, "And I believed it."

That sentence is why I am writing this. Not because there are villains. Because there is a system that makes decent people believe things that do not deserve belief.

Layer three: the incentive to publish

This is the layer nobody wants to discuss.

If you run a sports content channel, you have two options on a day when no match is worth talking about.

Option one: publish nothing. You lose today's traffic. You lose part of your ad revenue. The algorithm registers that you have slowed down and reduces your distribution for several days.

Option two: publish an analysis. It may be empty, but it holds the rhythm. It holds the slot on the homepage. It holds your standing with the algorithm.

It is no surprise that most choose option two.

In football, the solution to this problem has existed for a long time: dig into human stories. On a day with no match, you can write about a seventy-eight-year-old fan who has missed two home games in forty years. About someone who walked two hundred kilometres to see a cup final. About an injured player recovering in his hometown.

Esports is harder because its community is young and most of its stories have only one layer. But that is a difficulty to be overcome, not a difficulty that grants an exemption.

The real problem is this: telling human stories takes time. And time is the one thing the report factory does not have.

What I learned at Jeonbuk

In 2026 I was nineteen, a first-year university student. I was allowed into the press area for the first time as a student reporter, for a match between FC Seoul and Jeonbuk Hyundai on matchday four of the K League.

That day, while everyone filmed the goals, I watched the technical area. The Jeonbuk coach kept giving strange signals. I did not understand all of them, but I recorded every movement: the timing, the position, the direction of his pointing.

After the match I wrote a piece predicting Jeonbuk would shift to a left-leaning defensive block in their next game. That idea ran completely against the prevailing view. Jeonbuk won 2-1, exactly as I had analysed.

But the story is not the correct prediction. One person getting a call right does not make a profession. The story is that I wrote down the evidence: which movement, which minute, the direction the full-back moved. Readers could argue with me. They could say, "That movement was coincidental." But they could not say I had made it up. Because every detail was checkable.

That is the entire difference between an analysis and an empty report. Not quality. Not length. Not the professionalism of the format. It is whether the reader can check you.

At first, my male colleagues sneered: what does a girl know about tactics. But when they rewatched the tape against my notes, they had to concede. The place that once doubted me became the place where I found my answer.

From then on I applied one rule: the headline may shock, but the body must be full of data and specifics. If the headline promises more than the body delivers, I have deceived the reader.

And that is exactly what the forty-one-page report did. It promised deep analysis. It delivered a skeleton. It delivered not one single fact.

Thirty days after a mispronunciation

In 2026 I was twenty, chosen as a field commentator for a university radio station during the World Cup in Russia.

In the semi-final between France and Belgium in Saint Petersburg, I mispronounced the name of N'Golo Kanté three times in a row in the first half. Three times. Listeners called in to complain. I nearly quit the profession.

What saved me was not confidence. It was a very concrete decision: over the following thirty days I rewatched every France match from the group stage to the final, recording the pronunciation of player names as I went, and learning how to place emphasis so the commentary had rhythm. By the final between France and Croatia I pronounced everything correctly, and the very listeners who had complained sent compliments.

A mispronunciation, but the right voice — one I did not know I had.

I tell this story because there is a technical lesson inside it, and that lesson applies directly to the problem of the empty report.

When I mispronounced Kanté's name, listeners knew immediately that I was wrong. My information had broken at one specific point, visible and audible, and fixable. That error had a sound. It incriminated itself.

An empty report is different. It makes no sound. No name is mispronounced. No minute is misremembered. It passes through the reader's ears without leaving a single anchor to hold on to.

And precisely because of that, it is more dangerous.

Since then I have formed a habit: I read my drafts aloud before publishing. If a passage, read aloud, contains nothing specific — no name, no number, no date — I know that passage is empty. Reading aloud is the fastest way to detect emptiness, because emptiness can only be heard, not seen.

A summer with no audience and forty-seven people

In 2026 I was twenty-two, doing my master's. Every sports competition in the world was suspended. There were no matches to commentate. The report factory lost its raw material, and many places began to fabricate.

I chose the opposite path. I made a short podcast series called "A View From the Empty Seat," in which I invited fans to recount their most memorable stadium memory.

Over three months I interviewed forty-seven people. A seventy-eight-year-old woman in Busan who had not missed a home game in forty years. A young man who walked two hundred kilometres to see a national cup final. A father who told me he only ever talked properly with his son through football.

The episode with the widest reach drew more than fifty thousand listens in its first week.

A summer with no audience, but we still rehearsed for the audience's imagination.

What I learned from those forty-seven people is very simple: there is never a shortage of material. There is only a shortage of effort in finding it. When there were no matches, I did not run out of stories. I only ran out of easy stories.

The report factory does exactly one thing: it always takes the easy road. And when the easy road is blocked, instead of taking the hard one, it manufactures the appearance of the easy road.

A million reads and what it cost

In 2026, in Asian qualifying for the World Cup, South Korea were held 1-1 by the UAE in the third minute of stoppage time. The chance of progressing all but collapsed.

Public opinion blamed the coach. I wrote a piece with the opposite argument: do not blame the coach, look at five mistakes by the players themselves.

I cited specific numbers: the team misplaced twenty-three passes in the final fifteen minutes, and the main striker touched the ball only eight times across ninety minutes. Those two numbers, placed side by side, told a clear story about a team losing control in midfield and running out of routes to goal.

The article drew more than a million reads on a major Korean content platform within twenty-four hours. Several players later said publicly that they had read it and reconsidered their play.

But I want you to notice a detail that is rarely mentioned: both of those numbers are verifiable. Anyone can pull up the match tape and count. They are not speculation. They are facts placed next to each other to produce new meaning.

That is the whole formula. There is no mystery. Two verifiable facts, placed in a new structure.

And that is also why I write sharply about player performance but never attack players as people. I am harsh about performance. I do not judge the human being. The difference is this: performance has data to argue with; dignity does not.

Gegenpressing, athletics and the honesty of numbers

There is one tactical angle I want to bring in here, because it is the perfect example of how data can be used to lie while remaining technically valid.

Gegenpressing has been decoded. This has become almost a consensus among serious analysts. But the public has not caught up. For most viewers, a team that runs a lot still looks like a modern team.

That is the gap.

In many mid-tier leagues, teams no longer press to win the ball high and create chances. They use sheer volume of running to turn the match into an athletics contest. They run so the opponent has no time to think. They run so the game ends with few goals and many collisions. They run to secure a point.

And in the stats table, those teams lead the league in distance covered.

An empty report will read the distance number and write: "This team has a superior physical foundation, indicating a modern pressing philosophy." The reader nods. The number is real.

But the conclusion is false. High distance covered says nothing about the quality of pressing. It only says how many metres were run. To know whether the pressing is real, you have to look at pressure intensity metrics: pressures per opponent pass, time to recover the ball after losing it, average recovery position.

That is the difference between data and meaning. Data is everywhere. Meaning has to be built.

Based on my experience watching matches, I have noticed mid-tier teams across several Asian and European leagues sliding steadily into this style. It is safe. It is cheap. It does not demand high individual technique. It turns a football match into a fitness test, and turns the stats table into a medal for teams that did not win.

The empty report is that style's best friend. It takes a real number, attaches an unreal meaning to it, and publishes.

That is why I never write analysis based only on aggregate statistics. If I have not watched the match, I do not write. A number I have not seen with my own eyes is a number I have no right to explain.

Why format is the most effective tool of deception

There is one question I asked myself throughout the two weeks after reading that PDF: if the content is empty, why did nobody notice?

The answer lies in cognitive psychology.

The human brain uses surface cues to assess the quality of information, because that is the energy-efficient approach. When a document has a table of contents, numbered chapters, tables, a risk scale, and a disclaimer in the footer — the brain registers: this is the product of a serious process. And a serious process, by experience, usually produces trustworthy results.

That is a reasonable inference in most cases. And it is exploited in this one.

The empty report does not deceive you with content. It deceives you with form. It dresses itself in the full uniform of professionalism while wearing nothing underneath.

I ran a small experiment. I took two passages — one excerpted from the empty report, one from a genuine analysis — and stripped out every proper noun, every number, every formatting element. I gave them to ten acquaintances. None could tell them apart. When I restored the formatting to the empty passage, seven out of ten said that passage was "more professional."

Seven out of ten. Merely because it had tables.

This is why I believe the fight against empty reports cannot happen only on the producer side. It has to happen on the reader side, by teaching people to see through format. And it has to happen on the platform side, by rewarding the density of verifiable detail rather than rewarding pretty structure.

Where I might be wrong

I will argue against myself, because a piece that does not argue against itself is just another empty report with an extra layer of confident tone.

Possibility one: perhaps I am undervaluing structure. A report with a clear skeleton, even with thin content, can still help a newcomer to a field by teaching them which questions to ask. Seen that way, the empty report is not fraud but an apprentice's teaching aid. I may be demanding a level of expertise from the public that they neither need nor want.

Possibility two: perhaps I am imposing football's standards on esports. Football has a mature, public, standardised data ecosystem. Esports does not. Publishers own the data and decide who gets access. Under those conditions, some of the blanks in esports analysis are not the writer's fault but a consequence of data ownership. I may be blaming the wrong people.

Possibility three: perhaps I have myself published near-empty reports, differing only in that I had enough detail that readers could not tell. If so, this article is a confession disguised as criticism. I do not rule that out.

Possibility four, and this is the one that troubles me most: perhaps the public genuinely wants reassurance, not truth. An empty report tells the reader that everything is fine, that a system is watching, that professionals are handling things. In a year when everything is uncertain, that reassurance has value. And if the public pays for reassurance, then the report factory is merely meeting demand rather than deceiving anyone.

I leave those four possibilities here without resolving them. Because I am not certain. And a person who is not certain is better off saying so plainly.

The sixty-second check

If you read a piece of sports analysis and want to know whether it is real or empty, do one thing.

Count the verifiable specifics in the first three paragraphs.

Proper nouns with dates attached. Numbers with units attached. Events with sources attached. If the first three paragraphs contain fewer than three such specifics, there is a very high chance you are reading a product of the factory.

This is not a perfect test. A good analysis can open with a story containing no figures at all. But in most cases it is a good enough filter, and a fast one.

The widest stadium is not the one with the largest crowd, but the one where people are willing to listen. An empty report listens to no one. It only speaks.

A verifiable prediction

I will leave two predictions, and I accept being held to them.

First: within three years, at least one major sports content platform in South Korea will introduce a ranking metric that measures the density of verifiable facts, in order to separate genuine analysis from reformatted content. Platforms that do not will gradually lose reader trust to smaller channels, where writers are forced to be accountable under their real names.

Second: during the current regular season, as competitions enter their decisive phase and relegation pressure rises, the gap between aggregate physical metrics and actual tactical quality will become increasingly obvious. The teams that run the most will not be the teams that win the most. Anyone using distance covered to predict results will be systematically wrong.

As for that forty-one-page report, I have kept it. Not on my computer. In a drawer, printed, next to a pencil.

Every time I am about to finish a piece too quickly, I open it, reread the first page, and ask myself: in tonight's article, is there a single detail the reader can check?

If the answer is no, I have not finished writing. I have only just started publishing.