The Empty Cell in Table Tennis Data and the Cost of One Missed Check

Frankfurt, November 2026. In the press tribune at a WTT Champions quarterfinal, the small screen in front of me showed the live statistics panel. The column headed "receive-point win rate" returned three characters: N/A. Game one, N/A. Game two, N/A. Game three, still N/A. Down on the floor the two players changed ends, towels over their shoulders, the umpire calling the score with the steady rhythm of a clock that never tires. Nobody in the press room mentioned the empty cell. The broadcast graphic cut to a commercial, then returned with a clean scoreboard in which no column was blank. A colleague turned and asked me for the service-point win rate of the player who had just taken the third game. I opened my notebook, turned three pages, and said the sentence I hate most: "I don't have that number."

Seven years earlier I had stood in a similar position and also lacked what I needed. In June 2026, during the World Cup qualifier between Japan and Saudi Arabia, I mispronounced the name of midfielder Omar Hawsawi three times in the first half. Not once and corrected. Three times, consistently, as though I had convinced myself that my pronunciation was right. The broadcaster's phone rang right after the interval. I learned the lesson every young reporter must learn, except that I learned it late and painfully: a misspelled player name is where everything else starts going wrong. After the match I stayed until nearly midnight, writing out correct phonetic guides for 60 Asian players, recording my own voice and listening back, segment by segment. Since then the triple-check rule has been part of the job. But that rule only taught me to verify what I already knew needed verifying. It never taught me to look at the empty cell.

Table tennis runs on two separate layers of data, and they rarely meet. The first is official match data: game scores, point scores, supplied by the International Table Tennis Federation (ITTF) system for events in the WTT Series. The second is interpretive data: service-point win rate, receive-point win rate, first-three-ball win rate, average rally length, the interval between points. That second layer is marked up by third-party statistics providers or by the host broadcaster's own staff, working in an arena with multiple tables and multiple players competing at the same time in the qualifying rounds. One marker, two eyes, four tables, six hours. An empty cell is not an accident. An empty cell is the inevitable output of an operating model.

Japanese media's reliance on that second layer has grown extremely fast over five years. Tokyo 2026, held without spectators, forced broadcasters to compensate with graphics and statistics. Paris 2026 pushed the trend one step further, with receive-rate panels appearing after every game and becoming something viewers waited for the way they wait for the score. The WTT Series calendar has also thickened: a player inside the world's top 20 can play fourteen to eighteen tournament weeks in a year, each week producing a fresh statistics sheet. There are not enough people to mark up enough data for that many matches. So empty cells appear, and we learn to walk past them.

Based on my experience tracking matches, I began doing something I initially considered pointless: writing down by hand everything the screen left blank. During the 2026 season I attended 63 matches across three events in person, including a WTT Champions stop in Europe, a WTT Star Contender in Asia, and a domestic round in Japan. For each match I recorded five columns: rally length, first-three-ball outcome, service-point win rate, receive-point win rate, and the dead time between points. I used the stopwatch on my phone, resting on my thigh, thumb on the button. My margin of error is about half a second per press. That is the error I accept, and I write it at the top of every page.

The first result had nothing to do with table tennis and everything to do with time. In the qualifying rounds, the average gap between two points landed between 13 and 15 seconds. In the knock-out rounds, it rose to 17 to 19 seconds. And in the deciding game of knock-out matches, it stretched to 21 to 24 seconds, with one match reaching 31 seconds after a long rally in which the player who lost the point bent down to retie his shoelace. Dead ball is not the rest period of a match. Dead ball is where the match gets rewritten. I wrote that line in 2026 and I keep it unchanged, adding only one number: dead time accounts for 42 to 48 percent of the total duration of an elite game. Nearly half a match happens while the ball is not touching the table.

The second result concerns the empty cells. Across the 63 matches I tracked, 12 had at least one statistical column return an empty value for at least one game. Roughly one in five. What stands out is that matches with empty cells were not the dull ones. They were, in fact, usually the matches with the longest dead time in my notebook. One hypothesis I lack enough data to prove: when a match tightens, the person doing the marking tightens too, and data-entry steps get dropped precisely when the data becomes most valuable.

The third result concerns what table tennis does not have. Football has possession, and possession sometimes lies blatantly. I stood in Kazan in 2026 and watched Carlos Queiroz's Iran hold the ball for only 28 percent of their match against Spain, lose to a Diego Costa goal in the 54th minute, and nearly equalise in stoppage time. Iran's defence taught me to read a match with a different pair of eyes: having no ball is still a structured way of playing. Table tennis has no possession metric, and the closest thing to it is the first-three-ball win rate. In my notebook, players who win matches with a first-three-ball rate below 45 percent are usually the ones who accept longer rallies and drag the match into territory where the opponent is uncomfortable. They do not control the ball. They control the tempo.

To be more concrete, take the two matches I rewatched most in 2026. The first was Tomokazu Harimoto against Fan Zhendong in the men's singles quarterfinal at the Paris 2026 Olympics, a match Harimoto led and ultimately lost 3-4. In my notebook, the first three games had an average dead time of just 12 seconds, while the last three jumped to nearly 26 seconds. That shift did not come from the rules, did not come from the umpire, and did not come purely from fitness. It came from both players needing more time to decide what they would do with the next serve. The better a player understands the opponent, the longer the silence.

The second was the women's singles bronze-medal match between Hina Hayata and Shin Yubin, also at Paris 2026, which Hayata won 4-2 with her left arm heavily taped. That was Japan's second consecutive Olympic women's singles medal, after Ito Mima's bronze at Tokyo 2026, and I suspect it will be recalled far less often than it deserves. What I recorded was not the forehand loops but the number of times Hayata set her racket down on the table and picked it up again during dead time. Nineteen times. That number appears on no statistics panel I have ever seen, and I have never heard anyone ask for it.

At this point I have to say something the sports-data analysis world rarely wants to hear. Chasing a complete data table is a category error, not a technical one. We assume every match must generate the same twelve columns, and every match deserves the same table format. Table tennis does not work that way. A coach does not read the opponent's receive rate off a sheet before the match begins. He remembers the spin direction of the last three serves of the previous game, and he remembers it through his student's shoulder, not through a percentage. When an empty cell appears on screen, it does not break the system. It reveals that the system never needed that value.

There is a split I think about often. A table-tennis specialist would fill the empty cell. He would restart his stopwatch, recount the points, call a contact at the organising committee, and by evening he would have enough numbers to file. A multi-sport reporter like me asks a different question: why does that empty cell exist, and why did nobody see it. Both approaches have value, but they answer different questions. The person who fills the cell serves the audience that wants a number. The person who explains the cell serves the audience that wants to understand the sport. The trap of specialisation is that filling cells becomes the purpose, until one day the table is full and nobody remembers why it was needed.

I once misread a name so I would remember that no detail is small. Yet the smallest detail in this profession may not be a correctly spelled name, but an empty cell that everyone walked past. When the stands fall silent, I hear the data speaking for tens of thousands of people. That night in Frankfurt, the data was silent, and I recorded that silence in my notebook as a proper line of data.

Sport is a common language, and every language contains silences that cannot be translated. My job is not to fill those silences but to point precisely to where they exist, so that next time someone looks at a screen in the third minute of a deciding game, instead of walking past the empty cell, they pause for one second.