Table Tennis: The Data Architecture Behind an 11-Point Match

On a weekend night, I sat alone in the editing room, rewinding footage of a knockout-round match. The electronic scoreboard read 11-9. But when I opened the detailed data layer — every service, every rally, every backhand flick — a completely different picture emerged. The player who won that match took 9 of 11 points on his own serve, yet lost 6 of 8 rallies that stretched beyond seven strokes. The 11-9 on the board was real. It simply did not tell anyone the true story that unfolded on the table. To the casual viewer, table tennis is a sport of rallies too fast for the naked eye. A top-level exchange lasts less than three seconds. The ball leaves the racket at speeds that can exceed 100 km/h, carrying more than 100 rotations per second. Within that brief window, two players must make a cascade of decisions: where to place it, how much spin, power or placement, attack or fall back into defence. Every decision is a bet, and every bet leaves a trace in the data. I came to table tennis through the path of a data professional. For years, I have spent most of my time building athlete databases, comparing development curves, and trying to answer a deceptively simple question: what truly separates winners from losers in a sport where a point can be decided in a split second? The answer lies not in flashy rallies, but in the architecture behind them — the rule system, the tournament structure, and the flow of data running through an entire career. What makes table tennis a particularly fascinating subject is the speed of its own change. No other popular sport has restructured itself so rapidly and so decisively. In less than two decades, table tennis changed the size of the ball, switched its scoring system, banned hidden serves, killed speed glue, and finally changed the material of the ball itself. Each time, the entire data ecosystem of the sport had to be rewritten from scratch. Start with the ball. In 2026, the official diameter increased from 38 mm to 40 mm. It sounds like two millimetres, but physically it reduced speed and spin exponentially. Players who built their careers on high-speed looping suddenly lost part of their edge. Analysts had to recalibrate every model, because the same motion with the same arm strength now produced a different ball trajectory. In 2026, the scoring system changed from 21 points to 11 points, cutting each game roughly in half. This was the greatest shock to tactical thinking. When a game is only 11 points, every point becomes heavier, the margin for error narrows, and the so-called slow start becomes an unforgivable sin. In a 21-point game, a player could lose the first four or five points probing, then mount a comeback. In an 11-point game, doing that means handing the opponent the game. In 2026, the ban on hidden serves took effect. Before that, the server could hide the ball's path with his body, forcing the opponent to guess the spin. After the rule, a large part of the serve advantage vanished, and the value of reading spin from trajectory soared. In 2026, the ITTF banned volatile organic compound speed glue, which had been used to increase bounce and speed. Many players had to relearn how to play with their familiar blade, because the feel of the ball had changed. In 2026, celluloid balls were replaced by plastic balls. The new ball had a slightly larger nominal diameter, spun more slowly, and offered a distinctly different feel. This was the last change and also the most controversial, because it was rolled out at a moment when the sport was trying to expand into international markets. Players who had spent entire careers perfecting their feel for celluloid suddenly saw the gap between them and the younger generation shrink. Looking at that whole sequence, I recognised a pattern: every time table tennis changes its rules, it changes not only the match — it changes how we define a good player. Which metrics matter, which qualities are celebrated, are decided by the rules before anyone gets to judge. Data is the match's love letter — learn to listen, and you will see everything. From that rule foundation, I can go deeper into the core: technique and tactics. In modern table tennis, playing styles are classified fairly clearly. The two-winged fast attack relies on speed and continuity. The loop drive relies on heavy spin to create an arc and overwhelm the opponent through placement adjustment. The chop, a away-from-table defensive style, relies on breaking rhythm and forcing the opponent to err. Alongside these are variations such as pips and anti-spin rubbers. Interestingly, style labels often do not match what players actually do at the table. A player labelled a fast attacker may spend most of the time controlling and blocking, waiting for the opponent to err. A defensive player may turn the match around with lightning counterattacks. The gap between label and reality is always where data opens opportunities. Based on my experience following many major tournaments, the metrics that make up a table tennis match fall into a few main groups. The first is the point-win rate on serve and on receive. These are the foundational metrics, because the serve is the only phase a player fully controls. The second is performance by rally length: short, medium, and long rallies tell different stories. The third is scoring at decisive moments — when the game score sits at 9-9 or 10-10. The decisive-moment metric deserves the deepest dissection. The same player might win 65% of points across a match, yet take only 40% at the end of games. This is the number that separates champions from talents. A champion does not necessarily hit prettier points; a champion is the one who maintains technical structure when pressure peaks. That is why I always separate ordinary points from decisive points in every dataset. Beyond technique, the structure of the competition system also shapes the data. The rise of the WTT tournament system completely changed how table tennis operates at the professional level. Instead of a sparse calendar, players must now race through a tiered series from Grand Smash, Champions, and Star Contender down to Contender. Each event carries a different points gradient, and players must calculate which events to prioritise. The world ranking operates on a rolling 52-week mechanism. Points earned are valid for one year, then deducted gradually. That means a player may hold a high position while in reality facing enormous points-defence pressure, and an early loss can make their position collapse. This is a concept fans rarely notice, yet it determines how national teams plan their schedules. I once spent months tracking how associations manage the points of top players. The most notable conclusion: the world ranking does not measure current form, but past stability. A player in superb form over the last three months can still sit behind a player who has oscillated all year. Only by understanding this can one read the true meaning of the numbers. At the macro level, world table tennis still revolves around the axis of China and the rest. This is a major difference from football or tennis, where power is distributed relatively evenly. In men's singles, the gap between China's leading group and European and other Asian challengers has narrowed in recent years but remains significant. In women's singles, the gap is even clearer. This creates a very distinctive data structure: people compare not only players with each other, but also one system with another. European rivals, especially Sweden and Germany, have produced generations of players capable of causing upsets. Asia outside China, with Japan and South Korea, is also increasingly competitive. But what I look at in the data is not how many matches they win, but how they improve. An opponent strong only at a few events is usually the result of one outstanding individual. An opponent consistently strong across many events signals a maturing development system — and that is the real long-term threat. Looking at recent history of major events, the gap between the outstanding individual and the system becomes very clear. A country may successively produce medallists, but if the number of players reaching deep rounds does not rise accordingly, that signals an excellent individual rather than a sustainable foundation. Knockout-round data, the number of players reaching semi-finals year by year, is the more reliable indicator. On the Chinese national team side, generational restructuring is always a closely watched topic. Any team that maintains dominance for years will eventually face the question of a generational gap. As the older core ages out, is the next cohort truly ready? Data on roster age structure, the number of players in the 23-to-26 bracket, and the frequency of young faces in decisive matches can reveal early gaps before they surface on the scoreboard. A frequently overlooked variable is equipment. Blades and rubbers are not just tools; they are part of the feel system. When a player changes rubbers, the learning curve often stretches from weeks to months, and during that period performance data becomes noisy. For an analyst, the key is distinguishing a change made to amplify a strength from one made to patch a weakness. These two goals produce completely different data patterns once the adjustment period ends. This is where data teaches me a lesson about its own limits. Data does not answer your question. It teaches you to ask the right one. Years ago, I believed that with enough data, one could predict a player's future down to the decimal. I was wrong. Data tells you what happened and what usually happens, but it cannot replace understanding people. The paradox of sports data is this: the more numbers you have, the easier it is to confuse correlation with causation. A player who serves often to the left corner may win more matches. But that does not mean serving left creates victory. Perhaps he only serves left when leading, and those matches are easier to win for many other reasons. Naive models jump to wrong conclusions, and clumsy analysts spread that error as truth. Throughout my career, I have kept one principle: before publishing any conclusion, write out by hand the strongest counter-evidence against myself. If the conclusion still stands after considering the counter-evidence, only then do I dare to speak. This is the price of honesty with data — you must accept that most of the time, your intuition is wrong. There is one factor that table tennis data, however advanced, struggles to capture: tension. In a sport where a game lasts only to 11 points, psychological pressure is not a secondary variable — it is the main axis. A player may have an excellent point-win rate in practice and small events, yet collapse at the decisive moment of a major final. No metric measures the moment a hand trembles at 10-10. Major knockout matches are often decided by what I call invisible points. These are points that never appear on the stats sheet: a drop shot that breaks rhythm at the right time, a serve that changes spin mid-flow, a glance toward the coach for confirmation. These are things high-speed cameras can record, but no data model can fully explain. Another dimension data must confront is the media narrative. Every player carries a story — the young hero, the avenger, the fading champion. These stories form far faster than evidence accumulates. The analyst's job is to check whether the story matches the statistical foundation. When a narrative explodes after two or three matches, the sample size is still too small to assert anything. I always ask: if you remove the crowd's emotion, does the number still hold? So I built my working philosophy around a clear boundary: data guides, people decide. The vast database my team and I painstakingly built, with tens of thousands of athlete profiles and many standardised metrics, exists not to replace expert judgement, but to ensure that judgement is made on solid ground. A bad decision based on good data is still a bad decision. For Vietnamese table tennis, the lessons from international data become ever more urgent. This is a sport in which Vietnam has tradition and potential, but still lacks a systematic data infrastructure. Building a domestic athlete database, tracking the development curves of young cohorts, and comparing with regional table tennis nations will be a necessary step if the country wants to advance to the continental stage sustainably. At the lower level of the ecosystem, the commercial flow also operates by its own data logic. The star effect drives rubber and blade sales after each major event. Brands calculate carefully which players to sign, and at which stage of their careers. But commercial value and competitive value do not always coincide. A player beloved by the media may sell more blades than a quiet champion, and vice versa. Looking at that mismatch, one understands how the market misprices belief. Back to the story in the editing room. After finishing my analysis of the match on tape, I realised the most important thing was not who won. The important thing was understanding why they won, and whether that result would repeat. A single match is one data sample. Only when placed beside hundreds of others do we begin to see the pattern. And that pattern is what deserves trust, not a lone number. This is why I abandoned vague, feel-based commentary years ago. When I saw a player with a standout expected metric but a low actual output, I learned that such a gap is usually unsustainable. Statistics tend to regress to the mean. Those who mock data models often miss this: they judge a model by the result of one match, while the model is designed to be right over the long run. I once believed in a number the whole world laughed at. Time answered for me. Of course, believing in data does not mean being enslaved by it. One of the biggest mistakes in sports analysis is turning statistics into a religion, where numbers are worshipped and people are forgotten. Table tennis is a sport of humans before it is a sport of numbers. The ball is struck by a hand, by feet, and by a mind under tension in a split second. If your analysis has no room for luck, for emotion, for a flash of brilliance that cannot be calculated, then it has lost half its value. This is a perspective I believe will matter more and more as table tennis enters the digital era. When every match is filmed from multiple angles, when sensors can measure the spin and force of every shot, when machine-learning models can predict placement, the greatest temptation will be to believe we understand everything. But the history of this sport shows the opposite: every time technology advances a step, a new facet of uncertainty is exposed. Looking ahead, I believe the coming decade will see a shift in how table tennis is analysed. Traditional metrics such as point-win rate will remain important, but they will be supplemented by positional data, spin data, and adversarial simulation models. The question will no longer be who is stronger, but what happens if plan A collapses. Teams prepared for the second question will hold a long-term advantage. And in that picture, I still hold to one foundational principle: correct architecture matters far more than data. A vast database built carelessly will only create the illusion of understanding. A smaller system designed carefully, with clear metric definitions and transparent data provenance, can produce trustworthy conclusions. Table tennis deserves such a system. I once told my team that a data monastery needs no walls — it is built by the discipline of never-ending games. That discipline lies in recording every point, checking every source, and daring to say I don't know when the data is not yet sufficient. In a world where everyone wants to speak like a prophet, admitting one's own limits is the mark of the highest level. After all, this sport taught me that a table tennis match, despite its fast and simple appearance, is in fact a complex system of rules, technique, psychology, and competitive structure woven together. To read a match is to read that whole system. The number on the scoreboard is only the entry point. The real story lies behind it. And the question I want to leave for those who follow this sport is not who will win the next match. The right question is: what foundation are we building to understand that match before it happens? Because in a sport where a point can come and go in a split second, the only thing that can be prepared in advance is how we ask our questions. A point is a moment. A point-win rate is evidence. We live on the boundary between them.