When the Report Comes Back Empty: The Discipline of Silence in Sports Analysis

At 7:40 p.m., a nine-section report landed in my inbox. I opened it and found exactly one thing repeated: the string "N/A" filling every table. No title. No source. Content type listed as "unclassified." The information points were an empty array, and everything downstream was empty with it — no tournament, no team, no player, no timestamp, no reliability grade for the source.

Someone in the room that evening said the familiar sentence: "Just write something, we have two hours." I have sat in that room many times, and every time somebody turns the void into a commentary that sounds very sure of itself. The report allowed for only two possibilities: the source article was never fetched, or it was fetched and the extraction layer returned an empty object. Telling those apart mattered more than any sentence written that night, because the remedies diverge completely — one requires re-fetching the source, the other requires fixing the extraction schema and running it again.

In 2026 I was thirty, working for a young sports channel. Korea were playing Iran in World Cup qualifying and I was assigned the pre-match analysis. I built the argument on xG and progressive passes and concluded the national team should control possession rather than sit and counter. The coach kept a 5-4-1. The match ended 0-0, and Korea needed the final matchday to secure their ticket. The next day a male colleague said, in front of the whole desk, that women do not understand football and can only cling to numbers.

I did not argue. I downloaded all 38 qualifying matches from all five confederations and analysed them again from scratch. "That mistake taught me that data never lies, only the reading of it is wrong." From then on every judgment needed at least two independent sources, every figure needed a provenance note and an error margin, and every piece carried a methodology section at the end. Longer, slower, and with no room left for "in my opinion."

Our trade has a worrying reflex: after every big match, dozens of player ratings appear within twenty minutes. I once checked a few of those against full match footage and found most of the scores had been written from three minutes of highlights. Nobody lied; the gap was simply filled with memories of the prettiest moments. A sports data system runs on two distinct layers. The first pulls facts out of a source: a tournament name, a team, a person, a number, a date. The second interprets what the first has pulled. The second cannot invent what the first missed, in the same way a video analyst cannot describe a passage of play he never watched. When the first layer returns empty, the second has exactly two honest options: say there is not enough information, or say nothing. Every third option is decoration applied to fiction.

Leicester City in 2026-23 was the opposite case, and it shows what decent data looks like. After 14 rounds they sat second from bottom. My model flagged an anomaly: Leicester’s expected goals were running above their actual output, while actual goals conceded exceeded expected goals conceded by 7.8. That gap was not evenly distributed — it clustered in a narrow band of situations. Wout Faes made a direct error leading to a goal in three consecutive matches. I wrote that Brendan Rodgers needed to move to a back three to cover pace in the inside channels. Three weeks later Rodgers was sacked; Dean Smith came in and did use a back three, and Leicester were still relegated.

That outcome splits into two separate lessons. The 7.8 pointed at a mechanism, and the mechanism was verifiable by eye — I rewatched every goal conceded and counted the occasions Faes was dragged out of position before the ball went in. In the other direction, a correct diagnosis does not mean the team is saved. A good analyst identifies the mechanism; he does not promise the result.

"The cancelled Seoul derby of 2026 was a stress test for every prediction algorithm." The K League was suspended indefinitely, the Seoul World Cup Stadium stood empty, and I worked remotely on FC Seoul’s first ten matches. Average distance covered was 98.7 km per match, third lowest in the league; the rate of tactical fouls in their own half was climbing — a signature of systemic loss of concentration rather than individual decline. I wrote a critique of the tactics. The desk refused to publish it on the grounds that "the timing is sensitive." The piece stayed in my archive, and I added five seasons of the club’s fitness data while I waited. An unpublished piece is still a record; a published piece built on nothing is a permanent error.

Then Isak Hien. In 2026 I scanned 49 European domestic leagues looking for centre-backs for Korean clubs. Hien was 24, Ethiopian-born, playing for Hellas Verona: 2.9 successful tackles per match, and progressive passes in more than two-thirds of his matches — the signature of a defender who can start attacks. I set his profile against Virgil van Dijk at the same age. The piece got attention, but my recommendation to the national team’s scouts was rejected on the grounds that there was "no direct source." Four months later Atalanta signed him and he became a pillar of their 2026 Europa League title.

I tell that story because it exposes an operating rule of the trade: however strong the data, it is discounted without someone who has watched with their own eyes. So I added a third verification layer — contacts among video analysts in Europe — and split the writing into two parts: a data section for newcomers, a deep section for scouts. The approach has precedent in my own career. At the 2026 World Cup, after Korea lost 0-1 to Sweden, I struck up a conversation in the mixed zone with a Belgian agent who had spent two years watching a Senegalese player in the Belgian second division with the naked eye. I pulled the player’s public data: top speed 34.2 km/h, 61 percent successful dribbles, but only 18 touches in the final third per match and a very poor counter-pressing index. He was surprised I had never watched the player. That conversation opened two further contacts in the VIP area, and since then I have always built interview questions on numbers rather than on sentiment.

Back to the empty report. The most telling thing about it was the consistency of the emptiness: even the mechanical fields, title and source, were gone. When a pipeline fails at ingestion, the mechanical fields die first. When it fails at interpretation, the title and source survive and only the conclusions are blank. The string "N/A" looks identical wherever it appears, but its position tells a very different story.

Here is where it is easiest to go wrong: an empty dataset and a dataset of zeroes are two different objects. Empty means nobody has measured. All zeroes means somebody measured and found nothing. Blending the two is the fastest way to turn ignorance into a report with a table of contents.

By the same logic, a "low risk" label is the most dangerous output in any sports report. No team named, no tournament identified, no date — a "low" here would be read as a clean bill of health for a subject nobody has examined. In my risk register, the correct status of all six categories — competitive, financial, personnel, rules, public opinion, systemic — is "not screened," not "no problem."

If I had to set a minimum threshold for a piece of sports analysis to be allowed into the world, I would pick five inputs: the tournament or match name, the named entities, at least one quantitative fact, the publication date, and the outlet. Below that threshold every conclusion belongs in the lowest confidence band, and I write that into the body of the piece rather than hiding it in a footnote.

There is a counter-movement I have to state plainly, because it is the blind spot of the data people themselves. Correlation is not causation, and a model flagging an anomaly does not mean the anomaly caused the outcome. Leicester’s 7.8-goal gap could come from individual errors, or it could come from a midfield leaving so much space that the centre-backs were forced to step up. I chose the first reading because I had rewatched the footage, not because the number said so on its own.

The reverse also holds: emptiness is not evidence of harmlessness. A report that mentions no unpaid wages, no match-fixing suspicion, no injury to a key player means only that nobody has looked. In this industry the most dangerous document is a clean report produced by someone who read nothing.

And there is a professional pressure no model captures: the person who says "I don’t know yet" gets replaced on air by the person who says "I know." "I don’t trust intuition, I trust numbers that speak after being asked the right question." But to ask the right question, you first have to admit you have nothing to ask it of. "I once bet on a bad dataset and got a correct lesson in return." That lesson was not about reading more numbers; it was about recognising when there is no number to read.

The signal I will be tracking in the next cycle does not belong to any club. It belongs to whether data providers dare publish their error margins and their empty-field rates inside the product itself. "Every season is a ritual, and the analyst is only the one who records the omens." A decent record-keeper has to record the blank pages too — and state clearly that he could not read anything on them.