The Empty Spreadsheet and Table Tennis's Hunger for Evidence

On an afternoon in the press area of a WTT event in Beijing, a young colleague sent me a file with a short message: "Take a look at this analysis for me, our system runs nine dimensions." The file opened exactly like a professional report. Full title. A tidy nine-dimension framework: technique and tactics, equipment, player data and head-to-head, event system and scoring rules, the competitive landscape between China and the rest of the world, rules and governance, coaching staff and talent pipeline, risk surface, public narrative and expectation, industry transmission. Eleven major sections. Not one of them contained real content. Every field was filled with the same sentence: "Insufficient information, cannot assess." Thirty-four years covering table tennis, I have received hundreds of reports. Every one of them was full of numbers. Every one of them had a conclusion. This time I received a report so honest it was empty, and it taught me more than all the full ones combined. The sport has moved into the digital era at breakneck speed over the past eight years. WTT arrived, optical tracking systems were mounted around the table, spin-rate sensors appeared, heatmap software for placement patterns became standard, algorithms calculated win rates by table zone. A high-level match now generates thousands of data points: average ball speed, revolutions per second, direct-service winner rate, third-ball winner rate, placement distribution along both axes, average rally duration, attack-to-defence conversion rate. National federations hire their own analytics units. European academies build data labs. Clubs in Japan's T-League publish head-to-head dossiers before every round. I have nothing against data. I make my living from data. But I have seen enough beautiful reports to know where they come from. When the input is empty, the system does not stop. It fills the gap with something else. An analysis model handed an article it cannot classify will emit the template skeleton without populating a single field. An analytics team handed incomplete data will infer the missing part. An editor handed a blank draft will rewrite it from memory. Nobody in that chain is lying. But the final product is still a flawless-looking article with not one piece of evidence inside it. Based on my experience following matches, I can point to exactly where data gets invented. The placement heatmap is the clearest example. A player can have a heatmap glowing in the left corner, looking as if he dominates that zone. But the heatmap does not tell you how many times he played there while leading, how many times while trailing, and how many times simply because the opponent had already read the intention and deliberately pushed the ball there to wait for a counter. Same bright spot, two completely opposite tactical stories. To tell them apart you have to rewatch the footage rally by rally, cross-reference the score at that moment, and note the opponent's standing position. Without that manual work, a heatmap is a pretty picture with no meaning. The third-ball winner rate works the same way. It is routinely cited as the measure of an attacker's class. But it depends entirely on the quality of the serve and the quality of the opponent's receive. Against a weak receiver, the rate spikes without the player performing any better. Against a strong reader of serves, the rate collapses while his level stays exactly the same. Quote that number without the opponent context and you are describing luck while calling it skill. I sat in the arena in Paris in 2026, in the press rows, and watched Truls Moregard eliminate Wang Chuqin in the round of thirty-two of the men's singles. Before that match, every statistical table ranked Wang Chuqin number one in the world, with a win rate above ninety percent for the season. After the match, analyses appeared in droves, explaining which weakness in the opponent's system the Swede had exploited. Some came with charts accurate to the percentage point. The problem: those charts were drawn after the result was known. That is fortune-telling in a data costume. The lesson lies elsewhere. Moregard won because he served with a rhythm nobody could predict, because he accepted high-risk shots at the exact moment the opponent's system had not yet adapted, and because he was not afraid of trailing. None of those three factors appear in any pre-match table. They live in the footage, in how he moved between rallies, in which shot he chose at which score. To see them you have to sit and watch. No software sits and watches for you. The same logic explains why the comeback in Budapest in 2026 resists analysis. Ma Long beat Mattias Falck in the final four-one, a scoreline that looks comfortable. The rhythm of the match was something else entirely. The Swede applied pressure with heavy forehand drives, and Ma Long had to change his serving pattern mid-match to break the tempo. Read only the scoreline and you conclude it was an easy win. Watch the footage and you see a tactical adjustment executed across roughly three minutes between the second and third games. The scoreline does not record those three minutes. And those three minutes were the entire match. What bothered me most about that empty report was how the team treated it. Nobody praised it for being honest. Several people called it useless. Several suggested rerunning the system with different data. And then someone proposed the familiar solution: just fill in whatever is reasonable, as long as the report looks complete. That was the moment this industry quietly surrendered its credibility, in a meeting nobody took minutes for. The counterintuitive point I want to put on the table: an analysis that says "insufficient information" is worth more than an analysis that is confidently wrong. In this profession we have grown used to judging reports by thickness. The more pages, the more trustworthy. But thickness only measures word count, not evidence count. A three-page report where every conclusion traces back to a specific rally, a specific score, a specific minute, is worth more than three hundred pages of claims nobody can verify. In Vietnam we have a strange advantage in this story. Our table tennis data is thin. There is no optical tracking mounted around every domestic table. There is no dedicated analytics lab. Because we lack it, people in the trade are forced to watch footage, to take handwritten notes, to sit in the arena and count rallies. That manual approach is slow, but it does not let you invent. When you have to record every point by hand, you know exactly what you are missing. That is a form of discipline the data-rich table tennis nations are gradually losing. I once sat with a youth team coach in Hai Duong who recorded every match his students played in a hardcover notebook. He wrote the score, the type of serve, the direction of spin, and whether his student hesitated before that ball. That notebook had no charts. But it contained exactly the kind of information a nine-dimension model needs and does not have. He never said "insufficient information." He said: "It is missing, I will go watch again tomorrow." The difference between the two approaches is not technology. It is who takes responsibility for what gets written down. A model takes no responsibility. An analytics team chasing a deadline takes no responsibility. The person whose name sits under the article takes responsibility, and that person must have the right to say: I have not verified this. Table tennis is a sport where everything happens inside four square metres and lasts under three seconds per rally. Spin rates reach hundreds of revolutions per second. Placement shifts within tenths of a second. That is why data has value. And it is also why data is easiest to fake, because nobody can verify a number without sitting down and rewatching. Thirty-four years in this trade taught me one simple thing. An empty spreadsheet is not a verdict. It is a reminder that the ball only speaks when we are willing to listen. Table tennis forgives nothing to those who guess, but it always rewards the one who goes back and counts again. And the memory of the stands may deceive us about feeling, but it never lies about what we actually saw. That report, I still keep on my machine. Not to mock anyone. But to remind myself that every time I sit down to write, I must answer one question before the first keystroke: did I actually see this, or do I just need it to be true?