The Sound of the Ball on the Table and an Empty File: The Discipline of Verification in Table Tennis Analysis

1:12 a.m. In a small newsroom in Shanghai, I opened the tracking file for a tournament on the WTT circuit and found every cell empty. No set-by-set score. No win rate on serve. No list of players. Not even the name of the event. In the middle of the spreadsheet there was one line: insufficient information for analysis.

The story still had to be filed before six in the morning.

In nine years of following table tennis, I have stood in that moment many times. It forces a choice that readers almost never see: either say plainly that you have nothing to say, or fill the gap with language that sounds entirely reasonable.

I chose the first. But to understand why the first is so hard, you have to look at the machine that produced the empty file.

Context: a process that runs on two stages

Picture the production line behind an in-depth table tennis story. It runs on two stages.

The first stage is collection. Here people record what can be counted: the score of each set, the flow of live points, which player served at which moment, the share of points a player wins inside the first three strokes of a rally, the share won in rallies lasting more than seven strokes, the share won on decisive points from 9-9 onward, the number of points lost to service faults, the number of backhand direction changes, the number of times a player stepped in to attack the third ball.

The second stage is analysis. It does exactly one thing: it connects the facts from stage one in order to find a cause.

When stage one is empty, stage two has nothing to connect. Any conclusion it produces is guesswork wearing the clothes of analysis.

This is the part readers rarely see. They see fluent prose with numbers, opinions and conclusions. They do not see that behind it there may be a blank spreadsheet, and that the writer has convinced himself that a few estimated figures are good enough.

In table tennis, the gap between a grounded judgement and an ungrounded one is far narrower than in other sports. This is a game of extremely short intervals. A top-level rally lasts under three seconds. The service decision is made in about a second and a half. A topspin loop only means something when placed next to the opponent's position on the previous stroke.

Which means that without positional data and stroke-sequence data, the writer is left with feeling. And feeling, in this sport, is frequently wrong.

I still remember why I entered the trade. In 2026, seventeen years old and still a schoolboy in Shanghai, I stayed up all night after France beat Argentina 4-3 in the World Cup round of sixteen. I did not stay up for the four goals. I stayed up for the vast space behind Argentina's back line. I sat and redrew Didier Deschamps' formation, using arrows to trace Kylian Mbappe's acceleration paths. He touched the ball thirty-nine times, scored twice, assisted once and created three chances from thirty-metre bursts. The piece ran two thousand words, its diagrams drawn by hand in PowerPoint, and reached fifty thousand views in seventy-two hours.

From that night I set a rule for myself: never write about tactics in words alone. The rule followed me into table tennis, and it is the reason I cannot rest easy with a blank spreadsheet.

The principle I carried over from football is simple: the map of a 4-3 is not in the scoreline, it is in the running lines the camera forgets. In football those lines are a midfielder's movement off the ball. In table tennis they are the small adjusting footwork before the third-ball attack, the very thing the broadcast camera almost always cuts away from.

In 2026, when the pandemic emptied the arenas, I sat down with historical data and found that home advantage in a major national league fell from 46.2 percent to 41.9 percent once the crowd noise was gone. The lesson I kept was not the number itself, but how I found it: testing a new phenomenon against a sample large enough to speak for itself.

What happens when the data does not exist

Take a concrete example of how the two stages are supposed to fit together.

A player wins a match three sets to none. The easiest reading is that he was comprehensively superior. But if stage one has recorded everything, the picture is usually quite different.

Suppose the spreadsheet shows that he won sixty-eight percent of the points on his own serve, but only forty-seven percent on his opponent's serve. His win rate in rallies lasting more than seven strokes was just forty-one percent. And in the 9-9 region of the third set he won all four points, every one of them ending inside the first three strokes.

Read those three groups of figures side by side and the story changes entirely. It becomes the story of a player with an excellent serve, who loses the long exchanges, and who survives on his ability to finish early at the most pressured moments. This is a very familiar pattern for a game built on the third ball.

I call the early points the visible part and the long rallies the submerged part. The visible part decides the score. The submerged part decides how far a player can go in later rounds, once opponents have read the serve.

Without stage one, a writer will overlook the submerged part, or worse, will invent it by intuition.

In contemporary table tennis the submerged part keeps growing heavier. Since the 40-millimetre plastic ball became standard, spin has fallen, rally speed has risen, and the gap between the leading group and the rest has narrowed in extended exchanges. Backhand-dominant play has gradually displaced the dominance of players who lived entirely on the forehand. Players in the world's leading group, such as Wang Chuqin or Sun Yingsha, no longer win by waiting for opponents to err; they win by actively manufacturing tempo.

But tempo is not visible to the naked eye in a thirty-second clip. To see it you need data on when the ball was struck and where it was struck. Without that data, every description of a tempo change is literature.

I do not believe in random goals. Every rally is a link in a chain of the twelve strokes that came before it.

I wrote that line for football, but it holds even more strongly in table tennis. Here the chains are shorter, repeated far more often, and therefore easier to misread.

In Vietnam the problem is sharper still. Based on my own experience watching matches, most domestic and regional events are not recorded with detailed data. A player like Nguyen Anh Tu or Mai Hoang My Trang can perform brilliantly at a regional tournament, yet what remains afterwards is often only the scoreline and a few reconstructed highlight rallies. Writers in Vietnam therefore work with a higher coefficient of uncertainty than colleagues in places with complete collection systems.

That does not mean in-depth writing is impossible. It means the discipline of verification must be higher, not lower.

When all that remains of a match is the score, the thing to do is say so plainly: this is the score, the rest was not recorded well enough to support a conclusion. A sentence like that takes three seconds to write and can save an entire analysis from becoming speculation.

There is another detail the data layer never touches: the human being behind the racket. At 9-9, the service decision is not merely a technical choice. It is the product of having lost the previous two points with a sidespin serve, of what the coach said during the interval, and of the pressure of an Olympic or World Championship place hanging in the balance. The spreadsheet records the serve. It does not record the reason. A good writer knows where the data stops and where the story begins.

Three metrics are enough to build a minimum analytical frame: win rate on own serve, win rate in rallies over seven strokes, and win rate in the decisive region of the score. These three answer three different questions: who controls the rhythm, who can endure a long exchange, and who holds firm when the match tightens. Adding a twelfth metric does not make a piece deeper. It only makes it longer.

The real risk is not missing data

The biggest risk in this trade is not a lack of data. It is the reflex to fill the gap.

When a spreadsheet is empty, there is a pressure that is very hard to resist: write something that flows. The writer begins to deploy terminology that sounds authoritative exactly as the voice of the game, and yet never shows where on the table it happened or in which stroke of the rally. Those words are not semantically wrong. They are technically meaningless.

In a two-stage analysis file whose collection stage is empty, the most honest handling is to mark every field as insufficient information and then stop. Stopping there is the correct output of a correct process.

What is surprising is that this risk is rarely logged as a professional risk at all. People check whether a story has enough figures, enough sections, enough length. Few check whether those figures actually exist in the source file.

That is why I rank the worst risk in analytical work as chain risk: an empty file slips through the collection stage, is then processed downstream as though it were complete, is then published, is then cited again, and by the time anyone notices it sits inside thousands of search results.

A miracle is the name we give to the part you have not yet analysed.

I do not write that to disparage beautiful moments. I write it to remind myself that every time we call a win miraculous, it is usually a time when we have not yet found the link that explains it. And the explaining link almost always lies in the data: a small change in service placement, an adjustment of stance, a coaching decision in the middle of a set.

There is one more rarely discussed risk: the illusion of the small sample. A player wins three matches in a row with a backhand-dominant game, and a conclusion about a trend appears immediately. Three matches are three data points. No conclusion stands on three points. But three matches are enough to make a headline.

Sports writing lives under publication pressure, and table tennis has a crowded calendar. That is why verification discipline has to be designed in advance rather than left to a writer's willpower while a deadline burns.

What to do before the next match

If one concrete thing is to be taken from that night of the empty file, it is this: set a clear time limit for the verification step, and when the clock runs out, publish your level of certainty instead of a conclusion pretending to be certain.

For me that limit is ninety minutes. Within those ninety minutes I re-check sources, cross-reference independent data sets, and if there is still nothing, I file a short piece containing only what has been confirmed. Readers get less, but what they get is real.

The match ends, but the record keeps talking. That is the part I keep for myself.

Tomorrow, another spreadsheet will open, and another line will sit among empty cells. The writer's job is not to fill it, but to decide whether he has the nerve to say that it is empty.