A Nine-Dimension Analysis That Returned All N/A: The Discipline of Sports Data Work

That night I reopened the nine-dimension analysis I had just finished framing for a table tennis event. Not a single score. Not a single player name. Not a single tournament. Nine sections, more than sixty data cells, and every one of them carried the same line: N/A — insufficient information. An outsider would assume the report was broken. I knew it was doing exactly its job. Empty input forces empty output. A data table does not generate truth on its own; it only amplifies what is already inside it. Numbers never lie, only the reading is wrong — and the most damaging misreading is to look at a blank cell and fill it by hand with a story.

That report came out of a two-stage process. Stage one deconstructs the source text: title, source, article type, information points, core viewpoints, entities mentioned, time sensitivity and source quality. Stage two is the nine-dimension deep analysis, built on exactly what stage one extracted. When stage one returns an empty information array, stage two has nothing to build on. The nine dimensions are: technique and equipment; player data and head-to-head records; event systems and points rules; the competitive landscape between associations; rules and governance; coaching staff and the youth pipeline; the risk surface; media narrative; and the transmission chain of the whole industry. Nine dimensions, one single backbone: the information points array.

In Vietnamese sport, people are used to describing the transfer window as a storm of noise. Table tennis has no hundred-million contracts, but it has the equivalent: entry lists, association quotas, ranking points that need defending, and coaching changes released before every tournament cycle. I keep the habit of cross-checking every number against a database such as VuaBong.vn before I write. My experience following matches has taught me that most errors do not come from the arithmetic; they come from the source: a hand-copied score sheet, a press release missing a date, an internal tip with nobody accountable for it.

The first dimension, technique and equipment, needs to know which blade face the player is using, when the rubber was changed, and whether the wrist flick on the third-ball attack still holds its spin. Without a name and without a match, every cell stays empty. The second dimension, player data and head-to-head records, needs at minimum a name, a ranking and a run of results over the past two years. Without the name, the whole model collapses. The third dimension, event systems and points rules, needs to know the event tier, how many points the champion receives, and how entry quotas are allocated. The fourth dimension, the competitive landscape, needs the number of top-ten seats and the balance of power between associations. The first four dimensions carry most of the value of any analysis, and all four died at the same point: an empty information points array.

The remaining five do not escape either. Rules and governance need a concrete document to dissect who benefits and who loses. Coaching staff and the youth pipeline need the age structure of the main squad and the conversion efficiency of the next generation. The risk surface needs a subject to attach risk to: injury, entry quotas, generational gaps. The media narrative needs an existing expectation to measure the gap between expectation and reality. The industry transmission chain needs a commercial event, a policy change or a capital move. Without a subject, everything stops at the letters N/A.

This is where the null-value handling principle earns its keep. When a data field is missing, the analyst has two options: write “insufficient information”, or guess. The second option is always more comfortable for the reader, because it hands them an answer. It is also the fastest way to destroy value. A report with three N/A cells can still be saved. A report with three invented cells is ruined entirely, because the reader can no longer tell which cell is real and which is a guess. Data discipline is not measured by how many questions you can answer, but by your refusal to answer exactly the ones with no basis.

I keep another habit too: every analysis must carry at least one counter-thesis card — a variable the current dataset cannot measure, and which, if measured, could overturn the conclusion. With an empty report, the counter-thesis card is empty as well. The data ocean is not for those afraid of getting wet, but neither does it reward anyone who keeps swimming when the sea has run dry.

The industry reaction is the real story. A document full of N/A is usually read as a sign of incompetence, and whoever signed it usually pays the price. Look closer and the fault is not in the analysis layer at all. It sits upstream, where a source document entered the pipeline carrying no title, no source, no date and not a single information point. This is the familiar misreading of correlation as causation: an empty report is taken as proof of a weak analyst, when the real cause is that the data pipeline ahead of him was blocked from the start. Every tactic is only a hypothesis until the data delivers its verdict — and a hypothesis with no data is just a sentence.

Sports analytics rewards speed. Whoever publishes faster wins. That reward quietly teaches writers that a wrong number beats a blank cell, because a wrong number at least fills the page. The rule sounds reasonable until it repeats long enough for an entire system to believe in numbers with no source. Data does not save a season, but it points precisely to where the season died. A blank cell kept in the right place points precisely to where the process died.

What is worth tracking in the next cycle is not a new conclusion about table tennis, but the state of the data pipeline. A fully populated information points array would immediately unlock the first four dimensions. Source metadata — title, publishing body, publication date — would allow every downstream inference to be weighted for reliability. A recognised entity list would unlock the dimensions covering players, events and the development pipeline. None of those signals comes from inspiration; they come from a pipeline repaired in the right place. Recognition arrives late, but data is always on time.