When Table Tennis Data Returns to Zero: The Discipline of the Analyst

2:47 a.m. in Nha Trang. I sat in front of two screens — one showing a spreadsheet with fourteen pre-formatted metric columns, the other playing back a WTT Contender semifinal I had rewatched four times. I was waiting for a consolidated ball-tracking data file to cross-check against what my eyes had seen. At 3:12 a.m., the file came back empty. Not a single row. Not a single number. Just column headers and white space running to the bottom of the page. A layperson might think an empty file is trivial. But for anyone working in sports data, the most dangerous moment is not when a number is wrong; it is when the number does not exist at all. When a number is wrong, you have an object to inspect, to exclude, to correct. When the number is absent, all you have left is temptation — the temptation to fill the gap with intuition, with memory, with something that sounds perfectly plausible at two in the morning when no one is checking. I once fell into that trap. Not tonight, but years ago, when I still believed a good analyst is someone who always has an answer. I was wrong. And the file returning to zero tonight, in a strangely repetitive way, is reminding me of that mistake. Data does not forgive emotion. And that is why I converted. Context: why an empty file is big news To understand why an empty dataset could make me write a whole article, a word is needed on how the table-tennis analysis industry operates. Over the past fifteen years, table tennis has shifted from a sport analysed by feel to a sport analysed by metrics. People talk a lot about xG in football — expected goals, computed from shot position and quality. Table tennis has equivalent metrics that get far less attention: the point-win rate when serving, the point-win rate in the last two points of a game, the index of transitioning from defence to attack within a rally, and the pressure index at mid-distance. I came out of football. In 2026, at age 55, I calculated xG myself in a spreadsheet for a V-League match in which a team held 71 percent possession and took 22 shots but still lost. The piece "Possession is not victory" was shared more than three thousand times and turned me into a data writer in the Vietnamese market. Since then, I have required every article to carry at least two advanced metrics as its main arguments, not as decoration. When I moved part of my work into table tennis, I assumed everything would be easier because the court is small, the point count is low, and every point is recorded explicitly. I was wrong. Table tennis has a data paradox: every point is recorded transparently, yet the context needed to understand that point is extraordinarily hard to reconstruct. A serve at 10-all is not the same as a serve at 3-all, even though the scoreboard shows only the number. A player changes the racket face, changes rhythm, changes placement within the same rally — and without detailed ball-tracking data, all you have left is the final result. And the final result, as I keep telling my regular readers: the score is only the verdict. The rally is the testimony. A player who wins 3-0 may have won thanks to three lucky rallies at decisive moments. A player who loses 1-3 may have controlled most of the match. If you read only the result column, you are reading the verdict while skipping the entire testimony. That is why an empty data file is not a small matter. It means you have lost the ability to reconstruct the testimony, leaving only the verdict. And when only the verdict remains, the analyst begins to play the most dangerous game in the profession: telling stories out of nothing. The nine-dimension framework and the death of each dimension For years I have built a nine-dimension analytical framework for table tennis. It is not a product to show off; it is a discipline for self-checking. Whenever I analyse a match, a player, or a tournament, I walk through nine layers: technique and tactics, player data and head-to-head records, event system and points, the competitive landscape between associations, rules and governance, coaching staff and talent pipeline, risk surface, public narrative, and finally industrial transmission. The purpose of these nine layers is not to guarantee an answer, but to pinpoint exactly where I currently have nothing. Tonight, all nine layers return zero. The first layer is technique and tactics. To say anything about technique, I need at least one style label: loop-drive, fast attack, chopping, pips, penhold reverse-backhand. To say anything about tactics, I need the scoring structure of a match: serve-and-attack, backhand flick, short-push control, mid-to-far-table counter-looping. When the data file is empty, no style label exists. There is nothing to analyse. I remember, some years ago, confidently labelling a player's style as "European far-table play" based on two matches I had watched. By the third match, he was playing close to the table, and I realised I had built a story from too small a sample. I fear a wrong model more than a wrong judgement, because it is wrong systematically — and that time it was a wrong model. The second layer is player data and head-to-head records. To say anything about rankings, I need the total points, the points structure, the expiry calendar. To say anything about match-ups, I need a head-to-head table, and more importantly the ability to distinguish a true nemesis from a loss caused by form. Modern table tennis has the WTT system's rolling 52-week deduction mechanism. This is something worth saying a great deal about if you have a points ledger. Without a points ledger, I cannot know how many points a player is defending, how many are about to expire, and where the ranking-drop risk lies. An analyst without ranking data is like a captain without a compass who is still forced to shout a heading. The third layer is the event system and points. Table tennis has a clear hierarchy: the Olympic Games, the World Championships, the World Cup, WTT Grand Smash, WTT Champions, WTT Star Contender, WTT Contender, continental events, domestic events. Each tier has different points weightings and participation obligations. But to position an event within the Olympic cycle, I need its name and its date. Without a date, without an event, I cannot know where it sits on the four-year axis — the points-accumulation phase or the selection-lock phase. I remember, when building my home-field model during the football shutdown of 2026, that I collected 3,100 matches from the 2026-2026 season in Europe's top five leagues to calculate an average home advantage of 0.42 xG, then predicted the home-win rate would drop from 43 percent to 27 percent when the Bundesliga resumed without fans. It happened exactly as predicted. But the important thing was not that the prediction was right; it was that I needed an event name, a date, and a concrete context before daring to publish a number. When football died, I realised my home-field model had rooted itself in a false context — a context that always had fans. Table tennis is the same: an event without a name, without a date, is a false context. The fourth layer is the competitive landscape between associations. This is where modern table tennis differs sharply from football. In men's singles, the game is far more open than in women's singles, where one association has dominated for years. But to say anything, I need at least two entities at association or player level that can be set against each other. Without entities, there is no landscape. I can recite the list of the world's top players by heart, but reciting a list is not analysis — it is lookup. Analysis requires a subject, and a subject must be named. The fifth layer is rules and governance. This is a layer I especially enjoy because the history of table-tennis rule reform holds great lessons. In 2026, the ball went from 38mm to 40mm, reducing speed and spin. In 2026, the 21-point system became 11 points, increasing luck and per-point pressure. In 2026, the hidden-serve ban took effect. In 2026, the ban on speed glue containing volatile organic compounds affected a whole generation. In 2026, celluloid shifted to plastic, changing ball feel worldwide. Each reform created clear winners and losers. But to discuss a specific reform, I need to know which reform the article is about. With no rule mentioned, this layer cannot activate. I cannot, and must not, write a piece on "the impact of rule reform" when no specific reform is in hand. The sixth layer is coaching staff and talent pipeline. To assess a head coach, I need a name, a coaching philosophy, and the binding relationship between a personal coach and a player. To assess the pipeline, I need the age structure of the main squad and the depth of the U18 to U23 reserve cohort. The generational-transition question in table tennis — is the main tier ageing or young, is there a vacuum in the 23-to-26 band — is a very real question, but it needs at least a roster or a named cohort. Without a roster, I can only speak in abstract possibilities, and abstract possibilities have no value for readers. The seventh layer is the risk surface. This is the layer I regret most when it is empty. Injury risk from schedule density is a theme I have pursued for years: no medical team can save a player who plays two matches a week across a whole season. The risk of a slump after a technique overhaul, the risk of fluctuation after an equipment change, the risk of having a style decoded, the risk of dispersing energy by entering too many events — all are items that need to be screened. But with no player, no schedule, and no equipment change, I cannot screen any of them. And importantly: when risks cannot be screened, the biggest risk becomes decision risk. That is the risk of acting on a document that looks complete but is in fact empty. The eighth layer is public narrative. A player at the peak of public opinion or under doubt, a national team over-hyped or under-rated, a selection decision stirring controversy — all require a concrete claim or framing to be placed on the heat cycle of public opinion: budding, accelerating, climaxing, or receding. Without a source, I cannot even distinguish mainstream media from a fan community or self-media. This is foundational input data, and it is empty. The ninth layer is industrial transmission. From equipment and youth development, through events and associations and clubs, down to broadcasting and commerce and derivative markets. Star effect, equipment-brand influence, the coaching market, WTT commercialisation, capital flows — all need at least one named commercial actor or policy signal. Without a brand name, a broadcaster name, a host-city name, I cannot draw a single transmission arrow. A transmission map with three empty boxes for upstream, midstream, and downstream is not a map. It is three empty boxes. The contrarian angle: the gap is not for filling By now, my regular readers may have spotted the common thread across the nine layers: none returned information, yet each offered a way to fill the gap that sounds perfectly reasonable. I could assign a style to a player whose full matches I have never watched. I could comment on a ranking whose points I never checked. I could tell a generational-transition story without a roster. And the article would read very smoothly, very professionally, very credibly — until someone checks. This is the contrarian angle I want readers to carry away: in sports analysis, fluency is not evidence of accuracy. On the contrary, the more fluent an article is without an evidentiary anchor, the more likely it is to be far from the truth. The reason is simple: fluent prose is designed to persuade, while numbers are designed to verify. When you have numbers, you do not need fluent prose to cover anything. When you have no numbers, fluent prose is the only thing left — and it becomes a tool of camouflage. I call this the "gap-filling syndrome". It is so common it has become an industry habit. Analysts feel pressure to have a conclusion, editors feel pressure to have a piece, readers feel pressure to have an answer. No one wants to read a conclusion that says "insufficient information". So the gap is filled with plausible-sounding propositions: a statistic slightly exaggerated, a correlation upgraded to causation, an anecdote generalised into a rule. Each step of filling the gap is small, but added together they build a model rooted in a false context. Correlation is not causation. This is the most basic principle of anyone working with data, and also the most violated. A player changes his rubber and wins three straight matches — we attribute causation to the rubber change. But those three opponents may simply be weaker, the player may have just recovered from injury, or a three-match sample may be too small to say anything at all. Likewise, a national team changes coach and results improve — we attribute causation to the new coach, when the cause may be an easier schedule. Without a control group, without a test, you are only telling stories. With table tennis, the problem is subtler still. Points are recorded one by one, so we easily assume everything is quantifiable. But being quantifiable does not mean being quantified correctly. A player who wins a third game 11-9 may have led 10-4 and let the opponent claw back four points before closing it out. The same 11-9 result, two entirely different stories. Without point-by-point data, you are comparing two things that cannot be compared. The death of the player-data layer is not just a lack of numbers — it is a lack of context to make numbers meaningful. I know the feeling of being rebutted. When a metric I published is shown to be wrong, the first reflex of a long-time practitioner is to defend that metric. I have learned that reflex is a bad one. A statistic being refuted is not my failure — it is evidence that the system is self-correcting. Treating updated data as evidence that the model is maturing, rather than that I was wrong, is the difference between an analyst and an intellectual conservative. And in the case of tonight's empty file, admitting I have nothing is the only way not to become a fabricator. There is one more temptation I must name: the temptation to apply stereotypes. I was born in Japan, where organisation and hierarchy are revered. I live in Vietnam, where table tennis thrives at the grassroots but lacks professional data infrastructure. I read a lot about European and Chinese table tennis. Precisely for that reason, I am very prone to applying Japanese or European table-tennis stereotypes to a context with nothing to compare. Before comparing, I must set aside a dedicated, original data analysis from that very context. Without an original analysis, any comparison is just projecting a stereotype onto a gap. One memory stays with me. Years ago, working on a piece about Croatian football, I counted pressured passes myself by advancing video frame by frame because no ready data existed. I found that the trio Modric, Rakitic, and Brozovic made 4,321 passes, with Modric alone reaching an 87 percent pass-completion rate under pressure. I published a prediction that Croatia would reach the final. They did, and lost 2-4 to France. The piece hit 120,000 reads, the highest on the site. But what I learned was not that the prediction was right — it was how to weave numbers into emotional storytelling so that even non-expert readers understand. And the bigger lesson: if I had not counted myself, I would have had nothing to write. Croatia 2026 taught me that a pass under pressure is not just technique, but a manifesto. And a manifesto must have evidence. Tonight, there is no pass to count. No manifesto to issue. And that is precisely the lesson. Signals for the next round Back to the empty file at 3:12 a.m. I closed the spreadsheet. I did not write a single line of analysis from it. I marked the file with an internal note: return to sender, re-run the extraction step, check whether the fault lies on the retrieval side or on the source-data side not existing. That is the correct action, but it says nothing to readers. To readers, I want to leave a different signal. Over the coming weeks, as WTT events continue, watch for three things. First, analyses of table tennis that name no specific player, no specific date, no specific scoreline — read them as you would read a speech, not an analysis. Second, pieces attributing causation to an equipment change or a coaching change based on only two or three matches — ask where the control group is. Third, pieces publishing an impressive number without saying how many matches it was computed from, over what period, under what definition — that number is decoration, not evidence. I still hold one simple belief after nearly fifty years observing the sports industry: the most valuable thing about an analyst is not the ability to always have an answer, but the ability to say "I do not yet have enough data" and then get back to work. A profession matures only when it dares to leave the blanks blank. And a reader is protected only when the writer dares to let them see that blank, instead of filling it with something that reads very smoothly at three in the morning. If this piece makes readers uncomfortable because it offers no conclusion about any player, then it has done its job. Because in table tennis, as in any sport, the biggest gap always lies where we most crave an answer — and there, discipline is all we have left.