A Nine-Section Table Tennis Report With Not a Single Data Point

On Friday evening I opened a table tennis report file on my screen in Guangzhou. Nine sections, fully numbered. Tables aligned. A colour-coded risk matrix. Every conclusion carried a confidence label. At a glance, it was the kind of document any sports newsroom would want before a WTT swing.

Then I read it properly.

The assessment column said: insufficient information. The opponent column: insufficient information. The metrics column: insufficient information. The information-point list was empty. Entities involved: unresolved. The original article's title: none. Source: none. Article type: unclassified. Time sensitivity: not assessed. Source quality: not assessed.

Nine sections of professional reporting, and exactly one cell carried a real score: the analytical-process risk cell, rated High, with likelihood recorded as certain, already occurred.

The tactics board has no room for noise. But a board with no pieces on it says nothing either.

The architecture behind that file has two stages. Stage one reads the source article and extracts title, outlet, article type, one-sentence summary, author stance, purpose, the list of information points, entities involved, time sensitivity and source quality. Stage two takes that output and runs nine dimensions: technique, tactics and equipment; player data and head-to-head records; event system and points rules; the landscape between the leading group and the rest; rules and governance; coaching staff and the talent pipeline; the risk surface; public narrative and expectation; and transmission into the wider industry.

Stage two's principle is simple: every conclusion must be anchored to a Stage-one information point. No anchor, no conclusion.

This time Stage one returned exactly one usable field: the domain label, marked table tennis. The other ten fields were blank or unclassified. In other words, the system correctly identified this as a table tennis document, then failed at reading its contents.

In table tennis that gap cannot be filled with instinct. The WTT ranking system runs on a rolling 52-week window: old points expire, new results must replace them, and a player can lose a seeding position simply by having no event to play for two months. The three majors — the Olympic Games, the World Championships and the World Cup — measure something quite different from the annual tour. Without an event name, without a date, without a player, any read on the landscape is fabrication with formatting.

Based on my experience following matches across WTT events, I know one thing: most of what circulates online as table tennis analysis lacks anchors, not words. A player gets described with adjectives, then handed a competitive read by intuition.

A valid shell, an empty core

The notable part is that the report was still generated in full. Operating rules require the nine-section frame to be emitted in every case, including when the input is empty. The result is a product that is structurally correct down to the last colon and contains not one verifiable judgment.

I have met this kind of document many times in my career. It usually comes out of beautiful processes: a template, mandatory fields, confidence labels. The only thing missing is data. And because the template is beautiful, readers believe it.

Picture a tactics board with magnets placed at the right positions, the right distances, the right tactical shape — but not one magnet carrying a name. To an outsider it still looks like a diagram. To a practitioner it looks like a jigsaw exercise.

Why complete formatting is itself a risk

In sports data reporting, form is not neutral. A table with ruled lines, a matrix with colour, a confidence label reading medium — all of them transmit authority. A reader skimming sees structure, and structure is automatically read as competence.

In table tennis the trap runs deeper than in other sports. The discipline comes with a reassuring set of metrics: point-win rate, rally-win rate, serve efficiency, receive efficiency, points won inside the first three shots, win rate against players from other associations. Fill numbers into those cells and the report looks like a piece of research. But without a player, without a match, without an opponent, those cells are empty labels waiting for numbers.

An empty report is dangerous because it carries enough form for the reader to assume the information has already been checked.

Data does not lie; only the reader misinterprets. This time the data had nothing to say, and the misinterpretation was filling in the blanks.

Three honest inferences

The classifier worked while the extractor failed. The fault sits in ingestion and text parsing, not in topic recognition. That is valuable diagnostic information: it narrows the search for the fault to one link in the chain.

Alongside that, the entities-involved field is defined as dependent on the information-point list, yet the template carries no cross-field validation gate. An empty list means an empty field, and no warning is raised. A template without a validation gate will reproduce this failure indefinitely.

The remaining trace sits in the time-sensitivity field, which carries the value not assessed — a value quite different from a blank cell. That wording shows the deconstruction step was cut off mid-run rather than completed and returning empty. Every tactical system is a confession. This system confessed that it never read the article.

None of those three inferences is about table tennis. All three are about the analytical machinery itself. That is the entire residual value of the report, and also its limit.

The minimum viable input

For the nine dimensions to function, Stage one must return a minimum set. First, title, outlet, publication date and article type — news, analysis, interview or self-media commentary. Alongside that, at least three information points, each with a citable anchor: a result, a quote, a ranking change, an equipment detail, a selection decision. Without a concrete entity — a player, a coach, an association or an event — plus a clearly specified discipline, men's or women's singles, doubles or team, any read on the competitive landscape is meaningless. The final condition: time sensitivity and source quality must actually be graded, not left pending.

With those four groups, all nine dimensions become executable. Without them, the null result is the only answer that holds.

The standard I hold myself to

In 2026 I wrote that Belgium's three-man defence would break because the midfield was covering too much ground. That team reached the semi-finals and my inbox filled with angry replies. Since then I have kept a private rule: before praising or criticising anyone, the criteria must be written down first, scored numerically, and the things the numbers cannot answer must be stated explicitly.

For table tennis those criteria run to four lines: point-win rate in long rallies, serve efficiency at decisive points, win rate against players from other associations, and points-defence position inside the 52-week window. Without those four lines, praise is just emotion presented neatly.

The shelf life of an analysis

Table tennis information has a short shelf life. Rankings update weekly. Entry lists close and reopen. An analysis of form that arrives two weeks late must be reframed from current form into historical record, and that reframing must be stated openly. That is why a null result needs re-running within days, not weeks.

The counterintuitive part

The only risk figure in the file is rated High. It says nothing about table tennis. It speaks about the analytical product itself: a complete nine-section report can leave a reader believing nine sections of conclusions have been delivered.

There is a reverse reading worth considering. Perhaps the process did not break; perhaps the source article was nothing more than a metadata stub — a link, a description tag, no body text. In that case the null result is the correct answer, and the error lies elsewhere. I do not have enough evidence to separate the two possibilities, and I am not going to pretend that I do.

What is certain: under the same conditions, another process would have filled the blanks. It would pick a player, assign a form curve, construct a landscape, and publish a fluent read. That product would attract more views than this empty file. It would also be wrong in every cell, and nobody could check it, because there would be no anchor to check against.

An empty arena says more than thirty thousand spectators. An empty arena with no line-up says only one thing: nobody has walked in yet.

The verification test next time

Next time, before reading a piece of table tennis analysis, I will count the empty cells. A report willing to stay empty where it should be empty is a report that knows what it is doing. A report that fills every cell is a report selling the reader reassurance.

With this file, the verification lies in the re-run. If Stage one returns a title, an outlet, a publication date and at least three information points, the nine dimensions will mean something. If not, the null result remains the most honest answer I can give a reader.