Zero Data, Nine Dimensions: When the Spreadsheet Returns an Empty Truth

A 4,000-word report was on my screen at 2 a.m. Nine sections. Each had its own table, an assessment column, a risk column, a recommendation column. In every data cell, the same line repeated: insufficient information. No athlete named. No tournament. No date. Only the skeleton of a table-tennis analysis system, correctly assembled, then left completely empty.

What stopped me was not the emptiness. It was the form. The tables still lined up. The terminology was still exact: points-defence pressure under the rolling 52-week mechanism, technical-overhaul window, foreign-match win rate, the generational gap in the 23-to-26 age band. A hurried reader could skim for ten seconds and believe this was a professional document. You had to reach the third line to see the N/A.

That document is the stage-two output of a two-stage analysis pipeline. Stage one extracts from the source article: title, source, article type, list of information points, named entities, time sensitivity, source quality. Stage two takes that result and runs nine dimensions: technique and equipment, player data and head-to-head records, event system and points rules, the landscape between a dominant table-tennis nation and the rest of the world, governance and regulations, coaching staff and talent pipeline, risk surface, media narrative, and the industry's transmission chain.

In this run, stage one returned an empty list. No title. No source. Article type read "unclassified." The entities field held an instruction rather than a list of names. Stage two, instead of inventing, printed the entire framework with insufficient-information labels, then assigned itself the highest risk rating available: analytical-integrity risk.

In fifteen years watching this industry, I have rarely seen a document indict itself so thoroughly. And rarely seen one so widely misread.

A system with a skeleton will always produce output, even when there is no subject to analyse. That belongs to structure, to design, not to operational failure. Give a nine-dimension machine a blank page and it will return nine dimensions of a blank page. Tables still get ruled. Column headers still get set. The assessment column still has a cell to fill, and that cell gets filled with the only honest thing available: a dash.

The sports analytics industry has built a great many machines of this kind over the past five years. Every major tournament, every transfer window produces hundreds of identically structured reports: twelve metrics, seven comparison dimensions, a three-scenario forecast table. Most get filled. And precisely because most get filled, readers can no longer tell a report built on real data from a report whose cells were filled to complete the form.

A second point deserves credit: this system blocked itself. It set a clear condition — no conclusion may appear unless it traces back to an information point from stage one. With stage one empty, all nine dimensions locked. In the seven-row risk table, six rows belong to the sport and all read insufficient information. The seventh belongs to the analytical process itself, rated high, with a remedy attached: block publication, re-run stage one.

One detail kept me from full reassurance. In every dimension, after the conclusions, the document still carried a section called hidden information — things not stated outright in the source but inferable. And there, with an empty input, the system still inferred. It guessed the extractor most likely returned a null or truncated payload, medium confidence. It guessed the classifier had not run against real text either, low confidence. Those guesses are reasonable. But they show that narrative pressure is strong enough to operate on a blank page.

I once fell into the opposite trap. In 2026, while interning in Chengdu, I had a fourteen-round dataset from a third-tier league. A twenty-year-old striker had scored seven goals with an expected-goals figure of 12.4. I wrote two thousand words packed with tables and got back exactly one sentence from my editor: this is a financial report, not a football piece. I spent another month rewatching every touch to understand that a number only means something when it sits inside a specific moment on the pitch.

That lesson still holds. It has a reverse side I needed several more years to see: once everyone learns to place numbers inside situations, people start placing numbers inside situations even when the number does not exist.

In 2026, when global football stopped, I was handed a task few wanted: measure the effect of missing crowds on results. I built a dataset of 2,471 matches across five major European leagues from 2026 to 2026. Average home points: 1.54. Against 494 matches played behind closed doors from May to August 2026, the figure fell to 1.21. A gap of 0.33 points per match. I wrote a 4,000-word report: hypothesis first, data second, method stated, and the places where the data could not support a conclusion published alongside the rest.

When the stadium empties, data is the only spectator that does not leave its seat.

That report was shared, but it did not travel. What travelled, in June 2026, was a different piece I wrote about Croatia. I analysed three qualifiers and two friendlies, calculated an average PPDA of 9.2 — meaning opponents were allowed only 9.2 passes before being challenged, among the lowest in the tournament. I argued Croatia would strangle Argentina's midfield. The match ended 3-0. I called the pattern correctly. The article drew 1,200 reads. The same day, a colleague's piece mocking Messi drew 50,000.

The number spoke first, but people only listen once the truth has become legend.

Which is why I do not treat that empty report as a scandal. It is the surviving evidence that the pipeline still has a gate. The real scandal sits in the other ninety per cent: complete, tidy, numeric, and traceable to nothing.

Market mechanics explain why. A platform paying rights fees to stream a table-tennis event or a World Cup qualifier faces pressure to fill the hours around the match. Producing a match is expensive. Producing an analysis table is cheap. When the cost of content falls far below the cost of rights, volume beats quality, and a report with twelve metrics always sells more easily than a report with one conclusion.

For readers, the damage is not a single wrong article. It is the loss of the ability to tell them apart. When twenty pieces open with a shock metric, carry a comparison table, and close with three scenarios, readers pick the one with the strongest headline. That selection mechanism does not reward accuracy. It rewards confidence.

In the market I cover — Chinese table tennis — the pressure runs heavier. A dense selection system, a rolling competition calendar, the largest fan base in the world and a news stream that never stops. Each round produces hundreds of articles, most written within two hours of the final point, before technical data is released, filled with emotional description. That empty report, useless as analysis, was the only document that day willing to state outright that it did not yet know anything.

I have also seen the consequence elsewhere: inside the dressing room. Data analysts now sit in national-team meetings, presenting models coaches have no time to verify. Their conclusions are often arithmetically correct and rhythmically wrong, because a player's rhythm does not live inside a model. When a report is presented with enough confidence, few ask how many real data points it was built from.

Three questions belong in front of any sports report, mine included. Is the named subject actually present in the source. Does each conclusion trace to an event, a match, a dated figure. And does the document leave room for the possibility that it is wrong.

The empty report answers all three. It names nobody. It concludes nothing. It spends its final section stating that it cannot yet state anything.

More data is not what I want to see next season. This industry already has too much data and too few places for it to speak. What I want is a newsroom willing to publish a blank page rather than fill it with a three-scenario forecast nobody can verify.

Emotion writes the script, data draws the map. I only draw the map.

And the most honest map is sometimes one with a clearly marked blank region, rather than one filled edge to edge with imagined contours. Three signals will show whether the industry moves that way: the share of reports that state their sample size, the number of newsrooms publishing a minimum threshold before an analysis clears, and whether platforms still pay for volume rather than for verifiability.

Trusting data is like a cold early morning: few people wake up in time to see it.