Nine Dimensions of Table Tennis Analysis: When the Data Returns Empty

On the night of August 12, I sat in front of the screen waiting for a data extraction for my table tennis column. That extraction was supposed to contain player names, tournament names, scores, service statistics, and point-win rates in long rallies. It returned exactly one line: a domain label — table tennis. Below it, a void. No names. No events. No numbers. I sat still for a while. Seven years earlier, at 29, I thought I understood data. In 2026, I published a home-made expected-goals model for a V.League match and got the prediction completely wrong. The team I gave a 65% chance of winning lost 0-3. I spent a month reviewing footage to understand what I had missed. That night, the void on the screen reminded me that there is something almost as dangerous as miscalculating: filling a void with what you want to believe. I am not writing this to complain about a broken pipeline. I am writing because the moment data fails to arrive is the most honest test of the profession. It is also the starting point for a larger question about table tennis today: what framework are we using to analyse this sport, and what is that framework hiding? The framework I use for table tennis has nine dimensions. The first is technique, tactics and equipment. The second is player data and head-to-head records. The third is the event system and points rules. The fourth is the competitive landscape, especially the China-versus-the-rest balance. The fifth is rules and governance. The sixth is coaching staff and the talent pipeline. The seventh is the risk surface. The eighth is public narrative and expectation. The ninth is the industry transmission chain, from equipment and youth development to commerce and media. When the extraction returns empty, all nine dimensions fall silent. But inside that silence there is a lesson: a complete skeleton is not an analysis. And a void is not a finding. People routinely confuse the two. An article that says "there is no data" reads as deeply uninteresting, so people tend to turn it into "there is no problem". That is the first mistake I want to describe. Imagine the opposite situation. Suppose the pipeline returned everything, and I rushed to write an analysis about a player whose match I had never watched, based on a few scattered numbers. I could tell you a very fluent story: Player A is in form, Player B has psychological issues in decisive rallies, the coaching staff is clashing with the athlete. Every one of those sentences could be written without a single piece of evidence, given enough confidence. And that is exactly the kind of article I have committed never to produce. In 2026, I wrote a prediction before a World Cup final and published it on a sports site. I relied on expected goals per match to conclude that the side I judged stronger on numbers would collapse. The piece was read more than two hundred thousand times. The result went completely the other way. I realised my error was not in the number but in the context: I did not adjust the data for opponent strength across rounds. A team facing weak opponents in the group stage will post prettier attacking figures than reality. I had taken a pretty number and used it in place of the truth. The data was not wrong; the reader was — and I was once that reader. Since then I have set myself a rule: before every analysis, I ask a control question — what is the probability this is just background noise? If the answer exceeds 30%, I stop and write honestly about the noise instead of forcing a pattern. That 30% probability is not an excuse — it is a reminder that I am only right seven times out of ten. That is why the night of August 12 became a milestone for me, in a counter-intuitive sense. The void did not confuse me. It woke me up. And once awake, I began to ask: if I had to rebuild the nine dimensions of table tennis analysis right now — with no extraction in hand, only seven years of watching and a pile of public data I must collect myself — what would I write, and what would I admit I do not know? Table tennis is a peculiar sport for data. It is fast, each point lasts a few seconds, and the measurable window is tiny. A good serve may take two-thirds of a second. In those two-thirds of a second there are dozens of decisions: topspin or backspin, long or short, to the middle or to the corner. No model captures all of it. But precisely because of that, table tennis is a sport where data can help a great deal — provided the user knows what is being measured and what is being left out. I start with dimension one: technique, tactics and equipment. This is the dimension I am most wary of, because it is where people are most confident and most prone to loose talk. Anyone can say a player "plays a modern loop" without defining a modern loop. To measure, I must return to the physical milestones of the sport. In 2026, the ball went from 38 to 40 millimetres in diameter. In 2026, scoring changed from 21 points per game to 11. In 2026, the hidden-serve rule arrived. In 2026, the ITTF banned speed glue containing organic solvents. In 2026, the plastic ball replaced celluloid. Each of those milestones is a moment when the sport changed its own rules and, with them, the way people play. When the ball got bigger, speed dropped, rallies lengthened, and the value of endurance rose. When games shrank to 11 points, a single serve became far more expensive, so the pressure on the first serve spiked. When the plastic ball replaced celluloid in 2026, spin dropped and trajectories shifted, forcing players to rebuild both service technique and spin-handling technique. These are not trivia. They are variables any technical analysis must put on the table before speaking about a specific player. If my extraction had content, dimension one would require at least one of four things: a named player with a style descriptor; a tactical review of a specific match; a description of a coach's deployment; or an explicit equipment-change statement. Without those four, any technical analysis is subjective interpretation in disguise. I learned this from my own mistake: a technical number never speaks for itself. It must be traced back to the match — conditions, opponent, point in the set, score at the time — before it is used as evidence. Dimension two is player data and head-to-head records. This is where numbers are most abundant and where people are most easily fooled. The world ranking operates on a rolling 52-week points mechanism in the WTT system. That means old points fall off after a year, and a player can drop not because they lose more, but because they fail to defend last year's points. It is an important mechanism that very few people notice when reading a ranking. A player holding a high position thanks to a single big event faces entirely different points-defence pressure from one who accumulates steadily. Analytically, those two should not be placed side by side as identical unknowns. But the ranking shows only one number — it does not show the structure inside that number. Once again, my control question helps: what is this number hiding? To analyse dimension two properly, I need a named player, a current ranking, and either a head-to-head table or a set of recent results. Without those three, I can only speak in generalities about the points mechanism. And generalities about the points mechanism, applied to no one, produce no information gain — they are just the rules read back to someone who already knows them. I will take a historical example, old enough to be uncontroversial, to show how I still work. In the period when one generation of Chinese players handed over to another, statistics on long-rally win rates and service-point win rates often moved in opposite directions. Some players imposed themselves from the third ball, winning fast but collapsing when dragged into rallies. Others played longer, starting slowly but lasting late. Looking at a single metric would have produced the wrong conclusion about who was at their peak. I do not use one metric for a human being. I always need at least three variables to begin. Dimension three is the event system and points rules. This is the dimension I believe most table tennis viewers misunderstand, because they watch each match as an isolated event. But players do not live in isolated matches. They live in a calendar. The WTT system tiers events by point value, from Grand Smash and Champions to Star Contender and Contender. Each tier has different participation requirements and point structures. At the very top, the Olympic cycle runs on its own four-year rhythm, with selection points systems built by national associations. That means a player is not simply playing to win the match in front of them. They are playing to manage their points capital across a multi-year cycle. Some events offer few points for the physical cost; others change the selection landscape just by reaching the later rounds. Analysing a participation or withdrawal decision without placing it in the points cycle is incomplete analysis. I once read a commentary describing a player skipping an event as "losing the fighting spirit" — when the calendar data showed it was a sound load-management decision. The truth is usually less dramatic than the story. Dimension four is the competitive landscape, and this is where I must be most careful when discussing the China-versus-the-rest balance. One thing that has held true for years is that in men's singles the landscape is far more open than in women's singles. In women's singles, the dominance of Chinese players often leaves the rest of the world with only a few individuals capable of an upset. In men's singles, more non-Chinese players are capable of deep runs at major events, and the depth gap is narrower. But I must state the limits of this claim clearly. It holds at the level of trend, not at the level of exclusivity. Anyone saying the men's singles landscape is now fully balanced is reading half the truth. Conversely, anyone saying China cannot be beaten at a specific event is ignoring the denominator: the number of matches. A non-Chinese player can win one big match; sustaining that across seven consecutive matches is a different problem. The difference between "can win a match" and "can win a tournament" is a difference of distribution, not of moment. To assess dimension four properly, I need at least two entities — by association or by player — that can be placed in opposition, along with data on top-10 seats, titles at recent majors, and under-21 depth. Without those numbers, any statement about the balance of power is a belief presented as an assertion. Dimension five is rules and governance. Table tennis has a dense history of rule reform that I mentioned above: the 40mm ball in 2026, the 11-point rule in 2026, the hidden-serve rule in 2026, the 2026 VOC glue ban, the plastic ball in 2026. Every reform had winners and losers. A bigger ball favours physical players and penalises pure-speed players. The 11-point rule favours fast starters and penalises grinders. The hidden-serve rule favours good serve-readers and penalises those who live by disguising spin. What is notable is that almost every rule reform was argued for on grounds of spectator appeal, and almost always a group felt short-changed. But without a specific reform on the table, talking about rules is just recounting history. Recounting history is useful, but it is not analysis. Analysis begins with the question: how does this reform change the probability distribution of winning, and who recognised the change first? Dimension six is coaching staff and the talent pipeline. This is the most underrated dimension in Vietnamese writing about table tennis, because it is hard to see. People see a champion and praise the champion. They rarely see the person who spotted that player at twelve, who chose the right coach, who decided the player would attend this event rather than that one at fifteen. In table tennis, generational transition is especially complex because developing an elite player takes a long time. The age structure of a national team, the gap in the 23-to-26 band, the handover between one generation and the next — all of these determine a federation's fate over the next decade, yet are rarely discussed before results arrive. When results arrive, people explain them through individual talent rather than structure. I always think of this when watching debates about "whether Team X will keep winning". The right question is not "will Team X win" but "how many new faces of sufficient quality is Team X's pipeline producing over the next three years". That is a measurable question — via junior-event slots, junior head-to-heads, wildcard allocations. But it is less dramatic than a question about a final. Dimension seven is the risk surface. For table tennis, the standard risks I always check include: competition load and injury, technical-transition periods, equipment-adaptation periods, style being decoded, and energy dispersion across events. A shoulder injury may not bring a player down, but it changes movement structure and therefore spin quality. A rubber change may throw off ball feel for weeks. These effects are measurable, but only with time-series data. The biggest risk I want to discuss in this piece is not a sporting risk. It is the risk of deciding based on an incomplete analysis, because at the process level, a complete skeleton looks very much like an analysis. What must be checked first is not whether we have all the dimensions, but whether we have evidence for each. That is why I often tell younger colleagues: a table full of missing data still looks very tidy. Tidiness is not evidence. Dimension eight is public narrative and expectation. This is the dimension I believe carries enormous power yet is rarely quantified. In table tennis, public narrative often runs in cycles: an individual shines, expectations rise, every match carries the expectation placed on that individual, and then a defeat flips the story. The difficulty is separating a story built on a large sample from one built on a few moments. I made this mistake once and still remember it. In 2026, when leagues paused for the pandemic, I analysed 400 matches in the Bundesliga and K.League to measure the effect of playing without crowds on home advantage. The result showed the home win rate fell from 44% to 31%. I drew a conclusion and was strongly opposed. But what I took from it was not that I was right. It was a sentence I have used for many analyses since, including table tennis: the empty stadiums of 2026 proved one thing — data without context is only half the truth. Context here, for table tennis, means: whether an event has a crowd, whether the arena is noisy, whether a scoreboard is being broadcast live. Those things are not in the scoreline, but they produce the scoreline. Dimension nine is industry transmission. Table tennis has a distinctive transmission chain: from equipment and youth development upstream, through events and associations midstream, to media, commerce and player commercial value downstream. Each change at the top flows downward, but with different lags. An equipment change may take two to three years to show clearly in elite play. A change in event structure may take one points cycle to show in participation behaviour. One point I believe matters: commercial value and competitive value do not always move together. A player with a large fan base may have higher commercial value than one with stronger competitive results. In upstream analysis, the two kinds of value must be separated. Without separation, people measure fame and call it strength, or measure results and call it commercial potential. Both are category errors. Now let me say plainly what these nine dimensions cannot do. They cannot predict who will win the next tournament. They cannot tell you which player is in the best form right now. They do not replace watching the sport. This is the point I want to stress, because it is the professional boundary of my work: table tennis does not live in a spreadsheet — but a spreadsheet helps me see table tennis more clearly. A framework has value only when it serves a better reading of the match, not when it replaces the match. And here is the counter-intuitive part I want to save for the end. In sports data analysis, people are usually praised for delivering decisive conclusions. Audiences want to know who wins. Editors want a firm headline. But it is precisely when I deliver my most decisive conclusions that I am most often wrong. Not because my data was poor, but because I conflated correlation with causation. A player with a high service-point win rate does not mean their serve caused that rate. They may have faced weaker opponents in that stretch. The venue surface may have suited them. They may simply have been at peak physical condition. Correlation is not causation. That line is repeated so often people forget what it actually means. It means: if I cannot build a causal mechanism explaining why number A leads to number B — through technique, conditions, opponent — then I may not use A to assert B. I may only use A to say: there is a relationship that needs further testing. This makes most decisive conclusions online analytically meaningless — not because they are wrong, but because they cannot be verified. I do not say this to sound superior. I say it because I once wrote a piece read by two hundred thousand people and erred at exactly that point, and had to write a longer self-critique than the original to correct it. There is another blind spot worth noting, related to how people generate conclusions in table tennis. Because the sport moves at extreme speed, a viewer's memory of a match is usually dominated by a few emotional moments — a last-second save, a missed serve in the deciding set. But those moments are rarely the root cause. They are where pressure accumulated over a whole match becomes visible. A missed serve at match point has little to do with service technique and much more to do with how much energy the player has spent by that point. Yet the viewer remembers the missed serve, not the first three sets. This is where data can supplement memory: it shows that the visible moment sits at the end of an invisible causal chain. Every model I have was built on mistakes that were once laughed at — the most genuine foundation I have. The 2026 model that mispredicted a 0-3 in the V.League is a foundation. The 2026 model criticised for predicting against a World Cup final is a foundation. The 2026 study opposed for arguing home advantage falls without crowds is a foundation. None of them was entirely right. But they taught me one thing: before believing a conclusion, ask how many mistakes it was built on, and how those mistakes changed the conclusion. That is why I want to end this piece not with a decisive conclusion but with a forward-looking question. When a data pipeline returns empty, the right response is not to fill it. The right response is to record that void precisely, pinpoint where it lies, and state clearly what is needed for it to be filled by evidence rather than imagination. For table tennis, I think this matters more than in many other sports, because the volume of usable data in Vietnam is still thin. The nine dimensions I built are not a formula for appearing rigorous. They are a checklist to ensure that, when I speak about a player, I know what I am standing on and what I am leaving out. When I do not know, I will say I do not know — and that, to me, is an honest form of conclusion. Back to the night of August 12. I closed that empty extraction window, opened a blank file, and wrote the first line: my data returned nothing. I did not delete it. I saved it as one of the most important notes of the year, because it is evidence of something I always repeat: before analysing a sport, an analyst must learn to analyse the silence of the data itself. And in table tennis, as in every sport, an honest void is always worth more than a conclusion built from nothing.