When the Locker Room Source Cannot Verify: Lessons from an Empty Analysis

There was a moment in my career — June 2026, in a hotel in Kazan — when I sat in front of a screen with two blank pages and could not write a single word. It was the evening before South Korea faced Germany. I knew one thing: Lee Seung-woo had pulled a thigh muscle in a closed training session. But I had no training-load data, no MRI result, no confirmation from the team's medical department. I had one sentence. And I did not write. Three days later, Lee came on in the 65th minute and completed 12 accurate passes. He called me afterwards and said: "Thank you, hyung, for not selling me out."

I tell this story not to boast. I tell it because this week I was handed a document labelled "Stage-2 Deep Professional Analysis — Esports Domain". It was thick, with tables, a nine-dimension framework, a risk assessment section, a recommendations section. But when I turned to the very first part — the input verification — the line that appeared was: "input is effectively empty." No tournament name. No team name. No player name. No patch version. No article source. No publication date.

My notebook — the one with the worn spine that I have carried for nineteen years — has one page with exactly four words on it: "No facts, no story." I wrote that line in March 2026, after the K League season opener when I mispronounced Lee Seung-woo's name three times in a row on live broadcast. Getting a name wrong is getting a person wrong. Getting a number wrong is getting the truth wrong.

But what made me sit down and write this article was not the document's failure. It was how the document handled its own failure.

The first thing worth noting: it did not fabricate. In nineteen years following football teams and nearly seven years following professional esports teams, I have seen far too many analyses born from nothing. A team loses three in a row, and immediately there is an analysis of "internal crisis", "conflict between coach and captain", "player X losing form". Those pieces read smoothly. They have structure. They have voice. They have no facts whatsoever.

This document did the opposite. Instead of filling the gaps with speculation, it marked every cell explicitly: "N/A — insufficient information". Nine analytical dimensions, from patch analysis to industry transmission analysis, all left with their substantive conclusions empty. And at the end, it included a sentence I think every sports journalist should tape to their monitor: "The absence of risk flags in this output must not be read as evidence of absence of risk. It is strictly an absence of evidence."

This is the difference between a disciplined writer and a text-generating machine. A disciplined writer understands that silence is a valid answer. A machine does not.

But the second part of the story is the part I want to talk about.

In the "hidden information" section — what can be inferred though it was never stated — the document made an observation more valuable than all nine analytical dimensions combined. It pointed out that the failure signature here is very specific: the "domain label" field was correctly populated (esports), but the "information points" field was entirely empty, the "entities involved" field contained an instruction rather than data ("identify from the information points above"), and the title, source, and type were all N/A.

That is not the signature of an empty article. It is the signature of a process that broke at one very specific point: the domain-classification step ran, then the process exited before the information-extraction step could begin. The "entities involved" field containing an instruction rather than data is the most direct evidence — it shows the entity-extraction step never executed, rather than executing and failing.

For a reporter, this is a survival skill. When a source tells me "I don't know", I have to distinguish three possibilities: the person genuinely does not know, the person does not want to say, or the person was never asked the right question. These three situations require three completely different approaches. Confusing them is the fastest way to lose a story — or worse, to publish a wrong one.

In 2026, when I was one of the few reporters allowed into the stadium for the FC Seoul versus Gwangju FC match, I faced exactly this situation. FC Seoul's captain — a man I had interviewed many times — was being fiercely criticised on social media for inappropriate conduct toward the media. I knew the real cause: family pressure was breaking him down. But I had no evidence I could publish. No medical documents, no quotable testimony, nothing beyond what I had observed with my own eyes.

If I had written the piece with a full structure and concluded that "family pressure is the cause", I would have had a piece that read beautifully. And it would have been a lie dressed up in formatting.

Instead, I wrote about the mental pressure on players during the pandemic. I used the numbers I actually had: roughly 70% of K League players experienced stress during the lockdown period, roughly 45% of matches were played in empty stadiums. Those numbers I could verify. What I could not verify — one person's private story — I left outside the article. Not because I did not know. Because knowing is not the same as being permitted to write.

This is what I want to say to the people building automated sports-news processing systems. The document in my hands has a recommendation at the end, and I will translate it into the language of the trade: if the input is empty, stop and raise an alarm, do not generate text. Because text generated from nothing is not neutral text. It is harmful text. It takes the place of truth, and it carries the credibility of a trusted format while saying things nobody can verify.

I have seen this in esports more than in football. Perhaps because esports moves faster — one season can contain dozens of patches, hundreds of matches, thousands of hours of live broadcast. Perhaps because the esports community is hungrier for information, and speed is valued above accuracy. Perhaps because younger fans have not yet developed the habit of asking "what's the source?"

But there is one thing I learned from nineteen years of carrying a notebook: what cannot be verified cannot be called truth. It may be a rumour, it may be a hypothesis, it may be professional instinct — but it is not truth. And an industry built on things that are almost truth will collapse at exactly the moment it most needs to stand firm.

I went back to my desk. Outside the window, Seoul was entering autumn. In my notebook, the June 2026 page is still there, with the two blank pages I did not write.

There is a question I think anyone in this profession must answer for themselves, whether they write about football, about esports, or about anything else with a human being behind it: when you have nothing in your hands, do you have the courage to say "I don't know"? Or will you fill the gap with a beautiful structure, a smooth voice, and a promise you cannot keep?

I chose silence once, in 2026. Three days later, a player called me and said thank you. That is the only evidence I needed.