When the Data Has Nothing Left to Excavate: The Boundary Between Analysis and Hallucination in Table Tennis Journalism

There is a moment in sports journalism when I learned how to stay silent. It was when I sat in front of a screen with an empty set of notes, a nine-dimension analytical framework already built, and twelve data fields all reading N/A. No article title. No source name. No player name. No date. Not a single piece of information to begin with.

I have been writing about table tennis for nine years. I started in 2026 as a fact-checker at a sports magazine, which means my first job in this profession was not to write well but to find where the writing was lying. Since then, I have grown used to the fact that every analytical piece must begin with a single question: do I actually have data to excavate, or do I only have a hole and a shovel?

This time, the answer is that the shovel has no soil.

Every defeat is a layer of sediment; others see dirt, I see strata. But strata only exist when material settles. A nine-dimensional analytical framework with full tables, risk matrices, and industry transmission diagrams is beautiful, structurally correct, and entirely empty. This is the most dangerous trap of modern data analysis: a system can complete flawlessly in form while the content inside is zero. The technical world calls it silent degradation — the pipeline runs to completion, returns a valid schema, raises no error, and the reader downstream believes they are holding a real analysis.

What is frightening is not that a system fails, but that it fails politely.

If a data extraction tool throws a red error on the screen, I know exactly what to do: run it again, change the source, open OCR, look for an archived copy. But when it returns a full table with N/A in every cell, it is speaking to me in the language of fabricated completeness. And if I were a writer without discipline, I would fill those empty cells with my professional memory. I would recall a match. I would recall a player. I would reconstruct a story that sounds entirely plausible, and the article would be published, and no one would discover that the whole building was constructed on air.

I have watched this happen in the industry. In 2026, when I wrote my first piece for a paid football outlet during the Euros, a veteran reporter remarked that my writing was too emotional and lacked tactical expertise. I did not argue. Three days later, I sent him an analysis of how Italy shifted from defence to a high press, with data on distance covered and ball recoveries. He apologised in front of colleagues. But the lesson I drew was not that I was better than him. The lesson was that data only has power when it genuinely exists. If I had sent him an analysis full of tables but without a single verifiable number, he would not have apologised — he would have buried me.

In table tennis, this boundary is far thinner than in football. Football has ninety minutes and dozens of cameras. Table tennis has seven-second rallies and one camera at the corner of the table. A player can lose an entire set because a sidespin serve drifts two centimetres long. Without point-win data per rally, without rally-win rates, without recorded spin speeds, every technical judgement is a guess dressed in terminology. And a dressed-up guess, in this profession, is the most polite form of lying.

There is a paradox that I think everyone working in sports data must face. The world table tennis ranking system operates on a 52-week rolling mechanism: points expire on an annual cycle. That means a ranking figure without a snapshot date has no analytical value. Not low value — zero, or even negative, because it creates the illusion of understanding. An article stating that player A is ranked third in the world without saying when will leave readers with a completely distorted picture of that player's points-defence pressure. In table tennis, points-defence pressure is an enormous psychological variable, and it changes month by month.

I dig for treasure amid the ashes of a match. But the ashes must be the ashes of a match that actually happened.

This is why I hold one non-negotiable principle: when the input data is zero, the only correct answer is to say that the input data is zero. Not to write a speculative piece with a risk label attached at the end. Not to construct a story and then add a disclaimer. Because readers do not read to the last line. They read the headline, the opening paragraph, one table of numbers, and they carry that belief home with them.

I once sat in the stands of the Chengdu stadium at sixteen, recording every misplaced pass by the home side in a 0-4 defeat, staying until the ninetieth minute when the local reporters had all left, and discovering a young player who had run two kilometres more than the team average. That two-thousand-word piece was read by no one. But it taught me the most important thing of my entire career: the value of an observation lies in whether another person standing in the same place can verify it. If I had not had that notebook that day, I would have been just a girl saying this team does not run much. No one believes a girl without a notebook.

Failure is not a stopping point; it is the deepest geological layer of truth. But that truth can only be excavated when there is soil to dig. And when there is no soil, the most professional thing an analyst can do is put down the shovel, record the date and time, and report that the survey area contains no specimens.

There was one detail in that empty set of notes that made me pause longer than anything else. The entity description field instructed that entities be identified from the information above. But there was no information above. It was a self-referential instruction — a loop with no exit, generated by a template design that no one had reviewed. And as I read it, I realised this was no longer a table tennis story. This is the story of the entire data journalism industry: we build ever more sophisticated analytical frameworks, and we increasingly forget the simplest prerequisite — there must be content before there can be analysis.

In the third tier, there are no stands, no cameras, only people playing for their own stories. I learned from those matches that the crudest source is always more honest than the processed one. A scribbled notebook is worth more than a beautiful spreadsheet with no source data. A ten-second recording in a press room is worth more than ten pages of auto-generated analysis.

Do not rush to bury a season because of the scoreline; dig one more metre, where the foundation may lie. But I want to add a second clause after nine years: if after one metre there is still nothing, be honest and say you are standing on rock. Not every season has a foundation. Not every match has treasure. And not every article should be written.

In professional table tennis, people distinguish two kinds of error: technical error and perceptual error. A technical error is when you mis-hit a ball. A perceptual error is when you believe you have won while the ball left the table long ago. Data journalism is the same. A failed extraction is a technical error, fixable by running it again. But writing an analysis from empty data without admitting it — that is a perceptual error, and it cannot be fixed with tools, only with discipline.

An empty summer teaches us to hear the echo of old matches. In 2026, when the pandemic halted every tournament, I stayed in my room for three months and rewatched classic matches, diagramming every minute, noting every movement, and writing an eight-thousand-word piece on the collapse of a defensive system. I could write that piece because I had old matches to watch. If I had had no recordings at all, I would have written nothing. That emptiness was a provable emptiness, and precisely because it was provable, it was honest.

The emptiness of a nine-dimension framework full of N/A is an emptiness disguised as abundance. It is far more dangerous.

I think this is the moment table tennis needs a new standard, simple and non-negotiable: any analytical report must declare the number of real information points it relies upon, with the date and provenance of each point. If that number is zero, the report may not exist in the form of an analysis. It may only exist in the form of an error log. In table tennis, an umpire is never allowed to award a point for a rally that did not occur. In journalism, it should be the same.

Returning to the opening question. What do I have when I sit in front of an empty set of notes? I have nine years of experience, hundreds of matches watched, thousands of pages of notes, and one rule written in thick ink at the front of my notebook: never let memory replace data. Memory is the raw material for asking questions. Data is what answers them. If I invert those two, I will write very compelling articles about matches that never took place, about players who never existed, about turning points that never happened.

The treasure in the ashes is only valuable when the ashes are real. Fake ashes produce fake treasure, and the reader pays the price.

Old matches are fragments of broken bone; I reassemble them to see the shape of who I was that year. But if my hands hold no fragments, I am not permitted to draw a skeleton and call it archaeology. That is not archaeology. That is systematic fabrication.

And in a profession where every point is recorded in ink, systematic fabrication is the gravest crime an analyst can commit — graver than missing a match, graver than getting a number wrong, because it destroys trust in the entire system of record.

I still believe in the power of table tennis data analysis. I believe a rally-win rate can explain a career. I believe a movement-distance chart can predict an injury. I believe a small number in the corner of a scoreboard can retell an entire decade. That belief is lost at exactly one moment: the moment I forget to check whether the number is real.

So this time, I am not writing an analysis. I am writing a record. The record states that there is no article to analyse, no player to assess, no tournament to position, no system to dissect. There was a system that ran to completion and returned nothing, and the first thing to do is not to keep writing but to run again from the beginning with a real source.

One question to close, for those who have been in this profession longer than I have: if you receive an analytical report that is perfect in form, full of tables, full of risk matrices, full of confident conclusions, and at the top there is a small line stating that all data could not be determined — will you keep reading, or will you stop? Your answer to that question is the measure of your professional discipline, and it is also what determines whether your next article deserves to be believed.