Nine Layers of Reading a Table Tennis Match: The Discipline of Not Fabricating Numbers

Late at night in Guangzhou, I open a nine-page analysis of a table tennis match. Every cell is empty: no player name, no match, no metric, no event. A complete analytical framework with a hollow core. A newcomer to the trade would fill it with guesses just to finish the job. I sit still. After twenty-two years standing between the court and the statistics board, I have learned that an honest empty result is worth more than ten pages of invented conclusions.

That is the most expensive lesson the craft of table tennis data analysis has taught me, and also the thing very few people in this profession are willing to admit.

Context: too much data, too few readers

Table tennis generates a huge amount of raw data but it is poorly mined. Each high-tier WTT match produces thousands of data points: spin speed, placement, win rate in each zone of the table, successful forehand loops, points lost after a short serve. Yet most online commentary circles around feeling: this player has nerve, that opponent is raw. Feeling is easy to state, hard to verify, and no one is held responsible when it is wrong.

I built a nine-layer framework to read a table tennis match, borrowing the metric mindset from football and then adjusting it to the rhythm of this sport. Those nine layers are not ceremony. Each layer is a question that forces the data to answer, and each answer must leave a trace of its source.

Layer one is technique, tactics and equipment. The blade, the rubber, the hardness of the sponge: a change shifts every ball trajectory. A player who switches rubber needs about six to eight weeks for the hand to re-learn the power, and in that window their numbers lose value compared with their own numbers from the previous season. That is the first trap: comparing a person with their own past while forgetting the equipment changed. More tellingly, not every equipment change is recorded. Some players switch rubber mid-season without announcing it, and analysts only notice when the ball trajectory shifts visibly on video. At that point, any comparison of metrics before and after becomes meaningless without an adjustment variable.

Layer two is player data and head-to-head record. The world ranking is a summary; the raw data is the testimony. A top-five player can lose repeatedly to an opponent outside the top twenty, if that opponent's style attacks their weakness directly. Head-to-head must be read within a time window, usually the last two years, not accumulated across a whole career.

Layer three is the event system and the points rule. WTT runs a rolling deduction mechanism over fifty-two weeks. A major title can be wiped out after exactly one year, and that changes how teams schedule, how they choose events to concentrate their effort. Points do not merely reflect form; they shape it.

Layer four is the competitive landscape, China against the rest of the world. This is the layer the media loves most and the one most easily inflated. Seats in the world top ten, titles at the majors, the depth of the under-twenty-one generation: those three numbers tell a different story from sensational headlines. A nation's dominance is not measured by one win but by the gap between its next generation and everyone else's next generation.

Layer five is rules and governance. A small change to the ball, to the scoring system, or to the service rule creates winners and losers. Table tennis history is full of such reforms: the forty-millimetre ball, the eleven-point rule, the ban on hidden serves, the speed-glue ban. Each time, one group of players benefits and another is pushed out, and people usually remember only the second group.

Layer six is the coaching staff and the youth pipeline. Layer seven is the risk surface: competitive risk, selection risk, generational-gap risk. Layer eight is the public narrative and expectation. Layer nine is the whole industry's transmission chain, from equipment and training to events, commerce and media.

But a bigger risk sits inside the analytical profession itself. More and more data specialists are entering the locker room, drawing conclusions about a player's psychology and tactics without ever standing in the training hall. Their conclusions often detach from the actual rhythm of the match. They read charts, but not the breath of the athlete at the deciding point.

Core analysis: when the data is empty

Those nine layers only mean something when there is real data to pour in. And this is where I want to pause longer, because it is the boundary between an analyst and a rumor seller.

When the data is empty, there are three choices. One: invent a conclusion that sounds reasonable. Two: stay silent. Three: say plainly that there is not enough information to assess. The first brings page views. The third brings honesty. My profession lives on the third, even though it never goes viral.

In a table tennis match, raw data shows us what the naked eye skips. For example, the win rate in rallies that stretch beyond five shots. A player can win short exchanges easily but collapse when the match drags on. That is a sign of physical foundation, not of nerve. But the crowd always calls it nerve, because nerve is a better story to tell.

I was once mocked for using a model to show that a reigning champion risked elimination from the group stage. I did not rely on feeling. I relied on the gap between expected attack and expected defence, computed across many matches. When the result came, thousands of apologies appeared online. I did not keep them as trophies. I kept the lesson: a conclusion must come with a probability, say a thirty-two percent chance of advancing, not an absolute declaration.

That is also why I always state the source of the data: who collected it, how, across how many matches. Without that part, the number is only decoration. A metric without a source is an ownerless metric, and an ownerless metric can be bent by anyone.

In the current phase, as the transfer market and national teams both enter their cycle, the noise grows louder. Agents create stories to raise a player's value, and numbers about transfer fees, wages and release clauses become bargaining weapons. Fans read rumors; analysts must read the contract structure behind them. The structure is the real story; the headline is only the surface.

Another mistake is absolutizing a single number. When someone says numbers do not lie, they usually forget to ask under what conditions the number was measured. A beautiful home win rate may merely reflect that a player was given an easy draw early in the event. Strip away the context and the number becomes a legally sanctioned lie.

The contrarian angle

People assume a good analyst is someone who always has a conclusion. The reality is the opposite. A good analyst is someone who knows when a conclusion is not permitted.

In table tennis, a small sample is a deadly trap. A player who wins three matches in a row against a strong opponent proves nothing except that they won three matches. But the media will build it into a nemesis. Then when the opponent wins again, the nemesis story vanishes, and a new story is built. No one checks the old one.

Numbers do not lie, but the people who read them do. A metric that is right in one context can be entirely wrong in another. A player who performs well at home with a familiar crowd can collapse when the stands are empty. I collected data from hundreds of matches without spectators and saw the home win rate fall sharply, while the average score also dropped. When the stands are empty, I see the truest team.

So when someone asks me to predict a table tennis result and I do not have enough data, the right answer is that I do not know. That is not weakness. That is discipline. And that discipline, in an industry that lives on rumor, is the hardest thing to sell.

Takeaway

If you read a table tennis analysis in which every number leads to a neat conclusion, be suspicious. Real data always leaves a few cells empty. The analyst's job is to point out those empty cells, rather than fill them with plausible guesses. Next round, try asking yourself: how many matches does this conclusion stand on, and who collected the number. If the answer is no one, then that analysis is just a good story.