Nine Sections of Analysis, Zero Facts: How Esports Analysis Is Fooling Itself With Templates

2:14 AM in Seoul

I opened the document at 2:14 in the morning, right after killing the mic on a podcast episode about the transfer window. The filename was a meaningless string generated by a system, the kind of name nobody gives their own child. Inside: nine sections. Section one covered patch and meta. Section two covered tournament format. Section three covered teams and players. And so on down to section nine, which covered the transmission effects across an entire industry.

Every section had a table. Every table had columns. I counted the cells. Two hundred and four.

I counted the cells containing an actual fact. None.

Not a tournament name. Not a team name. Not a player. Not a single number on win rate, transfer fee, payroll, or patch release date. Not one line stating where the original article was published, on what date, by whom.

At the end, under "Comprehensive Assessment," the conclusion read: no substantive assessment is possible.

And the document was formatted flawlessly. It had a table of contents. It had a glossary. It had a four-category information value rating. It had a five-step remediation path. It had a disclaimer. I read it a second time looking for errors. There were none. That is exactly what woke me up.

One hundred and seventy-three cells of N/A

Two hundred and four cells, one hundred and seventy-three reading exactly N/A. The remaining thirty-one were longer, but said the same thing: insufficient information.

The first table was patch impact, four rows: meta direction, beneficiaries, losers, key data. All four empty. The second was roster assessment, four columns: paper strength, role fit, chemistry, bench depth. All four empty. The third was a risk matrix, six categories: competitive, financial, personnel, rules, public opinion, systemic. All six empty.

One table I read more carefully than the rest. It rated the document's own information value on a five-star scale. Competitive value: 0 out of 5. Industry value: 0 out of 5. Timeliness value: 0 out of 5. Reference value: 1 out of 5, with a note that it retained value only as a negative control — a worked example of correct null-value handling and a failure signature for future runs.

Those were the only numbers in the document that described something real. And the thing they described was the document itself.

Where templates come from

I entered this trade in 2026 as an esports player, then a tournament organiser, then a broadcaster. Back then an on-screen analysis panel had three lines: gold, minions, kills. Nobody called it an analysis system.

Then Korean broadcast needed frames. Each frame lasted ninety seconds, needed a table, needed a numeric column, needed a closing line. No analyst can go on air and say "I haven't seen enough data." So the frame was designed first and the data was pushed in afterwards. When there was no data, assumptions went in, and assumptions were read aloud in the same confident tone.

The frame migrated. From broadcast to print. From print to transfer coverage. From transfer coverage to automated article generators. With each migration it grew another dimension, because each dimension is another subheading, and subheadings are what make a document look like work was done.

On 18 March 2026, the FC Seoul – Suwon Bluewings derby, I proposed dropping Park Chu-young into a false nine instead of starting Dejan Damjanović at centre-forward, while colleagues in the studio laughed. FC Seoul lost 1-2. But the team produced 17 shots against their own previous-season average of 9.5. I carried that number into the argument and the K League community erupted.

Seoul that year did not rebel; it simply showed that tactics are written after the match is over.

I used precisely the mechanism I am attacking here: pick one true number, then build a story that stands on top of it. The 17 shots were real. They did not prove what I wanted them to prove. They proved the team shot a lot. The rest was me.

Three mechanisms that make an empty document look valuable

First: the template is a product, not a tool. What gets sold is coverage. Nine dimensions times six rows is fifty-four rows, and fifty-four rows look exactly like a twelve-hour workday. Nobody audits the inside of each cell. People audit whether all the sections exist. That is why this document has nine sections instead of one: nine sections are safer than one, even when all nine are empty.

Second: constraint collision. The system is bound by two rules at once. Rule one: never fabricate when data is missing. Rule two: fill the template. Obey both perfectly and you get exactly this document — one hundred percent compliant, zero percent informative. And because it followed the rules, nobody is entitled to scold it. That is the hardest failure in this trade to catch: a legitimate failure.

Third, and the most dangerous: confidence laundering. It runs in four steps. Step one, the document says "insufficient information." Step two, an editor reads "no risk identified." Step three, a host reads "the situation is stable." Step four, the audience hears "this team will compete." Each step loses only a sliver, and after four steps a blank cell has become a claim. Nobody lied at any step. All four people were polite. And the result is an assertion built out of nothing.

A system designed never to say "I don't know" will always find a way to lie within the rules.

The Germans did not die from a lack of talent; they died because they trusted their diagram more than the feet on the pitch. Sports analysis has its own version of that disease: trusting the template more than the blank cell staring back at you.

In fairness, I have to put a number on the table against my own argument. Over the past twelve months I counted forty transfer analyses I read end to end. Thirty-one contained at least one specific number. Nine contained none. But of those nine, six reached conclusions far stronger than the other thirty-one. Which means numbers save nobody, and the absence of numbers kills nobody. My sample is forty articles, I know, and I will not pretend it is science. But it is enough to stop me from claiming that any table equals any sin.

Two ways a data pipeline dies

This morning's document diagnosed itself better than most sports writing I read this month. It distinguished two failure modes.

Mode A: the ingestion layer died first. The source article was never fetched — dead link, login wall, JavaScript-only page, anti-scraping block. The analysis layer above it had no text to work with.

Mode B: the extraction layer died. The article arrived, but the parser returned an empty object because of a schema mismatch, truncated output, or a field-mapping error.

These need different fixes, and this is where the empty document teaches the trade something. It identifies a clean diagnostic signal: if even the mechanical fields — title, source, type — are blank, the failure is upstream. A working parser always picks up at least the headline. No headline means it never saw the article.

I have watched this failure in esports data for five years. A tournament reschedules, the publisher API returns a blank page, and the on-air analysis panel still runs on time because it was pre-loaded the night before. Nobody presses stop. Nobody has a stop button. And that evening an analyst explains why Team A beat Team B using three numbers from a different match.

Absence of evidence is not evidence of absence.

That is the one line in the whole document I would tape to the studio wall. It sits in the risk section, and it concerns the three stories this industry buries most often.

Unpaid wages. No table says "this club owes three months of salary" until someone posts it publicly. An analysis that omits it has not cleared it.

Match-fixing. The kind of thing a six-row risk matrix never catches, because it does not live in match data. It lives in somebody's bank account.

Injury. And here I have to confess.

In 2026 I did exactly what this document refused to do

In April 2026 the world stopped. Stadiums emptied. Colleagues went quiet; I built a simulation model from FIFA 20 data and proposed a thirty-minute first half to raise tempo, with a number attached: a 23 percent reduction in muscle injuries, based on an analysis of 450 K League matches.

The Korean referees' council objected. ESPN Asia republished it. A wave of debate followed. And when football returned, I wrote the most-read piece of my career: my idea failed, but the spirit of breaking rules won.

My thirty minutes during the pandemic taught me this: football does not need more time, it needs fewer illusions.

But this morning, counting blank cells, I have to say the part I never said on air. That 23 percent was never measured. It was interpolated. I had 450 matches of running data, collision data, card data. I did not have muscle-injury data at that granularity for 450 matches, because nobody publishes it. I took a reference rate from sports medicine literature, applied it to the model, and called the output a simulation result. That is decorated fabrication. That is filling a blank cell with an assumption typeset in bold.

Had the system behaved the way it behaved this morning — write N/A and stop — I would not have had the million-listener piece that year. And Korean sport would probably have lost nothing.

Where I might be wrong

Three possibilities, before any verdict.

First: the empty document is not a failure but a product doing its job. It is a pipeline QA report, not an analysis. I measured a technical artefact with a journalist's ruler and scolded it for not being journalism. If so, the error is mine, and it is the most common error in my profession: misreading the genre, then judging it.

Second: the culprit is the deadline, not the template. Templates are simply what is available when the deadline arrives. Give people two more days and the template stays, but the cells get real data. Blaming the table is blaming the scale because the person standing on it lied.

Third, and the one I hate: the client is the patient. Someone in Seoul, at 2:14 in the morning, opens a document with zero facts and still commissions three thousand words from it. If you want an assertion, you will always find someone to write it. The market does not pay for honesty; it pays for certainty. And I am part of that market, not its victim.

All three are partly true. But a piece without a verdict is just a conversation.

My verdict: this morning's document is the most honest artefact esports analysis produced this month, and precisely because it is honest it is useless to the majority. The real offender is a system that taught a generation that a blank cell is an embarrassment to be covered up.

What I will bet on

One falsifiable prediction, and I want readers to keep it for comparison.

Within thirty days, at least one transfer analysis will be published with the same nine sections — squad context, tactical impact, regional read, financial effect, risk, forecast — containing not one transfer fee, not one contract length, not one release clause. It will draw at least five times the engagement of any piece brave enough to write "we do not have enough data to conclude."

The test is simple. Open any transfer analysis you are reading and count the verifiable numbers: fee, contract length, salary, clauses. If the count is zero and the piece still concludes, you are reading a nine-section empty document in makeup.

And the question I leave with my colleagues, in Seoul or anywhere: if your data pipeline returns a blank page tomorrow, will you go on air and say the one sentence — "I have nothing today" — or will you open the template and start filling?