The Empty Data Table: The Line Between Table Tennis Analysis and Fiction

On Tuesday morning, I opened my inbox and found an attachment from a young colleague. The file name was clear: "Match Analysis – Quarter-Final Round". I opened it. Inside was a full analysis skeleton: technique, tactics, equipment, head-to-head, physical condition, psychology, risk. Twelve section headings, each with a proper label field. But everything underneath was empty. Not a single athlete's name. Not a single score. Not one statistic. I sat still for a long while. In data consulting, a file like that usually means one of two things. Either the data-extraction pipeline upstream has broken — someone uploaded an article, but the parsing software could not read it and returned an empty shell. Or the writer deliberately left it blank, waiting for someone else to fill it in. Both meanings point to the same thing: there is no real data in that place. And I left it exactly as it was. I filled in nothing. My data café is busiest when the field is empty. I have said that half-jokingly, half-seriously for years. People assume analysis means sitting in front of a mountain of numbers and drawing clever conclusions. The bare truth is duller: most of the time, I sit verifying whether I actually have data or merely a template. I am 52 years old, working as a data consultant for clubs and several table tennis training centres in Saigon. I came to this work very early, back when "table tennis analysis" in Vietnam meant the gut feeling of a veteran coach. People would watch a player's forehand loop and say, "That shot is heavy". Nobody asked how heavy, compared to whom, under what conditions. That was the foundation I grew up in, and also the reason I spent nearly half my life learning to turn feeling into numbers. But just as I learned how to build numbers, I learned the harder thing: when not to build anything at all. Based on my experience tracking matches, most analytical errors do not come from miscalculation. They come from filling a gap with a plausible-sounding story. In 2026, I once wrote a 1,500-word piece on a winning streak of a club I consulted for. The whole city said the team won through fighting spirit. I pulled the opponent's PPDA figure and showed it was just 8.2 — meaning they deliberately dropped their press to counter-attack, not that they were overwhelmed. I also cross-checked actual xG running 4.7 below goals scored. That pretty streak was largely luck. The piece was shared widely, and I began to believe I could see truth where others could not. Then World Cup 2026 arrived, and I learned the opposite lesson. I was invited on television to commentate with data. In the opening match, I got so excited — my loud personality — that I misread the Russian striker's name three times in the first half. Viewers criticised me. I was ashamed, but I did not quit. For the whole following month I sat rewatching footage, noting each team's pressing figures minute by minute. An old recording is a mirror; only those who dare to look at it see themselves. I realised I had misread a name because I was reading from memory, from habit, not from the data on the screen. Since then I write more slowly, verify names before speaking, and always add a layer of movement data — showing the gap behind the full-back, the distance run, not merely retelling the event. So when that empty file reached me, I felt more grateful than annoyed. It forced me back to the most fundamental question of the trade: what does a number in table tennis actually mean? In table tennis, the data structure is far thinner than in football. Football has xG, PPDA, dense passing networks. Table tennis in Vietnam — most national-level events — still lacks a full automatic scoring system, let alone positional data per second. The WTT system captures live point-by-point data, but at provincial and youth level, data usually lives in the referee's notebook and the coach's memory. So I build my index set very carefully, and I always state the reliability of each layer. The first layer is service points. At elite level, the share of points won on serve typically ranges around 55 to 65 percent depending on standard, and a gap of a few percentage points is already enough to decide the outcome. The second layer is rally win rate in the closing phase — from point 8 onward. At that level, a player may win 70 percent of rallies across the match yet lose all three decisive ones. The third layer, which machines can barely measure, is the breath between serves. Table tennis players train reflexes, and reflexes reveal themselves when they slow down. A pause longer than half a second before a serve is a sign the hands are starting to hesitate. But with an empty file, all three layers stay exposed. I have no athlete names, so I cannot cross-check head-to-head records. I have no score, so I cannot compute the service-point rate. I have no dates, so I cannot place it in a form cycle. The only thing the file tells me is this: somewhere, the data-collection stage has snapped. A fragment like that says nothing about a specific table tennis match. It speaks about the pipeline itself. This is where people in the trade often get confused. The crowd looks at the scoreline; I look at the forgotten pass. The majority is the same, only at a different level: they see an empty table and automatically think "no information yet". I see an empty table and think of a broken extraction system returning blank values in every field. Those are two entirely different conclusions, and the responses are entirely different too. One demands more data. The other demands fixing the tool. Every number is a piece of a puzzle, but I do not assemble it out of habit. If I assembled it out of habit, I would fill in the name of a familiar athlete, assign a plausible score, and construct a quarter-final that looks utterly real. And I could do it very successfully, because I have watched thousands of matches, I know how a quarter-final usually unfolds, I know what an elite forehand loop looks like. My storytelling instinct is strong enough to fill a blank table within minutes, and readers would suspect nothing. But in doing so, I would have swapped the trade of analysis for the trade of fiction without ever changing the sign above the door. Because the line between those two trades sits exactly here: the analyst says "not enough data", the fiction writer says "I think". The majority in sports analysis, not only in Vietnam but everywhere, rewards confidence. Go on air with a firm prediction in a steady voice and people remember your name. Go on air and say "I have no numbers to conclude", and you are seen as lacking nerve. That is a distorted incentive system, and I believe it is the root cause of most rumours in sport. People read one figure, hear one source, then retell it as a complete story with no room for silence. But numbers know how to hold their breath, and I wait for them to exhale. Even during the transfer window, when noise drowns out signal, the principle does not change. People rumour that a player will move clubs, they talk about fees, about contracts. But what truly decides a table tennis player's future is not the rumour. It is the structure of the tournaments they are targeting, the international calendar, the ranking-points events the club wants them in. A contract says a player will wear new colours, but the release clause and the new wage structure are the real story. They are not attractive, nobody writes headlines about them. But they have data behind them, and rumour does not. I once feared the microphone; now I let the data speak for me. But data can only speak when it exists. When it does not exist, I must speak on its behalf and say that it does not exist. That is the hardest skill, and the one almost nobody in the trade teaches each other. I remember a session at a youth training centre. A 15-year-old student asked me why he won a match while his rally win rate was lower than his opponent's. I took out the match sheet and checked every serve. It turned out he won because the opponent made unforced errors in the decisive phase, not because of his forehand loop. The boy went quiet for a moment, then asked: "So did I win by luck?". I said: "You won because your opponent erred, and in sport that is still a win. But if you think your forehand loop won that match, then next time you will become complacent with yourself." That is exactly what the empty file was telling me. It denies no match. It simply refuses to call itself an analysis when there is nothing inside it. So I replied to my young colleague with a short message: re-run the extraction stage, check whether the document reader is broken, and do not send me a blank table again. I also suggested adding one line at the top of every analysis: the data source, the publication date, and the confidence level of each layer. Those three lines look procedural, but they are what keeps this trade standing. I do not think there will never be anything to write about. I only think that in a sporting culture where data is still sparse, honesty about the gap matters more than the glamour of a hasty conclusion. A mature analytical culture is measured not by how many conclusions it produces, but by how many times it dares to say "I do not have the numbers yet". If the young generation of analysts in Vietnam learns that before learning advanced metrics, then in ten years we will have table tennis data dense enough that the whole world has to read it. If they learn the metrics first, learn storytelling first, then we will have a great many analyses that sound wonderful and are not right at all.