A Nine-Dimension Analysis With Zero Data Points: The Failure Sits in Extraction, Not in the Team

A July night in Chengdu, and Xuan Hy Lo street is still lit. I am sitting in a small content room on the fourth floor, watching a colleague drop a six-page analysis file into the group chat. The structure is beautiful: nine dimensions, each with a table and a conclusion section. Everyone nods. Then I scroll down and read cell by cell: "insufficient information to assess," "insufficient information to assess," "insufficient information to assess" — repeating all the way to page six. No team is named. No tournament is identified. No number appears. What landed in the chat was an empty mold wrapped in a nice typeface. What chilled me was not the empty mold, but the room's reaction. Someone suggested publishing it immediately, with the headline: analysts have reached no conclusion. Vietnamese and Chinese sports content has moved past the phase of competing over who says the most shocking thing. Now it competes over who has the prettier skeleton. An article is judged by its number of structural layers, tables and subsections, rather than by the number of verifiable facts it contains. The cost does not show up immediately, but it has a history. In January 2026, when I was a first-year sports management student in Chengdu, I wrote a two-thousand-word piece on the AFC U23 Championship final between Vietnam and Uzbekistan. My argument: pushing the captain and centre-back high up the pitch in the 88th minute was a suicidal decision, and the 119th-minute concession followed from it. That piece drew 47,000 reads in three days. What I remember is not the number, but how it spread: the largest football fan page at the time shared it with the caption a different perspective, and nobody went back to check the replay. I was right about the conclusion, but if I had been wrong, nobody would have caught it. Four years later, in November 2026, I was making content for a new sports platform and wrote a seven-part series on Japan at the World Cup. Part three — Japan's 2-1 win over Germany — hit 1.2 million views, and it became my first paid sponsorship, worth 18 million dong. That series survived because every instalment was anchored to something checkable: the high-pressing window in the final fifteen minutes of each half. Thirty days inside the World Cup: where tactics are not drawn on the whiteboard. In July 2026, I published information that a Chengdu Rongcheng striker was about to move to the Middle East for 8.5 million euros. The deal collapsed at the last minute when the buying club found a hamstring issue in the medical. I absorbed 1,200 comments calling it fake news, until the player himself confirmed it on a livestream. Those three stories say the same thing: an analyst's value lies in whether he holds facts. Back to that six-page file. Its failure sits in three layers, and all three are process failures, not professional ones. The first layer is labelling. The topic classifier tagged the source document esports, but it tagged it from metadata: the URL, the tags, the channel name. Not a single line of the body text was read. A topic label drawn from metadata does not confirm that an article has content; it only confirms where the article was published. The second layer is extraction. The list of facts returned empty: no information, no entities, no headline, no source. Technically this usually signals one of four situations: the source is video, the source is an image, the source sits behind a paywall, or the page is JavaScript-rendered and the crawler cannot execute it. The striking part is that the empty state was never blocked. The third layer is consumption. The empty file was passed down to the deep-analysis layer, which behaved exactly as designed: it returned nine dimensions, each marked insufficient information, each accompanied by a principle that is entirely correct — cannot be assessed never means no risk exists. Operationally, that is a disaster, because the final output is a long document with tables and sections that looks identical to a real assessment. In football, when a linesman raises a wrong flag, we have VAR to overturn it. In analytical content, we have no VAR; we only have the reader's reflexes. When I dissected Japan against Germany, I had no nine dimensions. I had one question: why did Japan accept ceding the game for seventy-five minutes in exchange for the last fifteen? To answer it I needed three things: a heat map of duels, the count of defensive-to-attacking transitions, and the 75th-minute substitution. With data, the conclusion stands on its own. Without data, the conclusion is just an empty mold with my name attached. I have to interrogate myself here, because the flaw I just described is my own occupational disease. For years I have sold skeletons. I taught colleagues that an analysis needs a hook, a context, a core, a counter-argument, a close. I told them structure makes readers believe, and I was right — to the point where structure began generating belief without any substance holding it up. When a document carries nine subheadings, the reader's eye fills the gaps between them with the assumption that the author has data. Formal completeness is the cheapest and most effective camouflage in this industry. If someone tells me that insufficient information is an honest conclusion and enough to publish, I half agree. It is honest at the level of wording and irresponsible at the level of execution. A piece saying we know nothing still occupies the slot of a piece saying here is what we know, and readers cannot tell the two apart when both are long and both have a table of contents. There is another possibility I must leave open: perhaps the source article genuinely had nothing worth analysing. Even then, the correct conclusion is the source contains no analysable content — not a ten-page document describing that emptiness as though it were a finding. My verifiable prediction: within twelve months, at least one major sports outlet in Vietnam or China will publish a multi-dimension analysis in which more than half the cells read insufficient information, and it will be quoted as expert opinion. I would very much like to be wrong. The gate I propose is cheap: before calling a document an analysis, it must contain at least one named game title, one named entity, and three sourced facts. If not, the correct status is invalid input, not no findings. The mistake is not in the final shot; it is in the second I saw the system break before anyone else.