Thuy Linh's 24-22 game against Chen Yu Fei: mapping 110 rallies and the real limit of Asiad's third round

On September 26, the electronic scoreboard for the third round of the women's singles showed three lines: 10-21, 24-22, 12-21. Chen Yu Fei won. Nguyen Thuy Linh lost. The same day, on another court in the same arena, a second scoreboard: 16-21, 14-21, Nguyen Hai Dang lost to Kunlavut Vitidsarn. I sat with those two scoreboards until nearly dawn in my office in Nagoya, copying each line of numbers down the whiteboard and counting. No video, no shuttle speed, no unforced-error sheet, no rally durations. The original diary left exactly six lines of numbers and a handful of names. Six lines of numbers, read the way a tactical analyst reads, can still build a map. Before building the map, I have to state the limits of the source. The original piece calls the event "Asiad 20", gives no year, no edition, no assessment of source quality. The results are recorded exactly as written, but any inference about ranking, about Olympic qualification points, about historical standing has to be suspended pending verification. That is a principle I have held for years: read the numbers, leave the blanks blank, do not fill them with feeling. One more thing needs saying clearly, because it is the key to this entire piece. Badminton uses rally scoring: every rally won is a point, best of three games, games to 21, at 20-20 you must lead by two, capped at 30. The consequence few people notice: the total points in a match equal the total rallies in that match. No rally is wasted, no point is free. Every time the shuttle hits the floor the scoreboard moves, and every time the scoreboard moves two people pay in legs, in breath, in a decision. From that I built a simple table. For each game I recorded each player's points, summed them into total rallies, then divided Thuy Linh's points by total rallies to get a scoring rate. The method is crude. It does not say where the shuttle went, who was controlling, which rally turned the match. But it has a quality most badminton numbers in circulation lack: it cannot be beautified. A player who scores 46 of 110 rallies has a rate of 41.8 percent, no matter what anyone writes about fighting spirit. I do not use shot-quality estimation models here, because at the level of a single match at a multi-sport Games there is no input data to run one. No landing coordinates, no shot classification, no racket speed. Feeding an under-specified model into an article only produces an impression of precision, which is more dangerous than admitting you are estimating. A scoring rate is something I can defend in front of any critic. The result for Thuy Linh against Chen Yu Fei is as follows. | Game | Thuy Linh | Chen Yu Fei | Total rallies | Thuy Linh scoring rate | |---|---|---|---|---| | Game 1 | 10 | 21 | 31 | 32.3% | | Game 2 | 24 | 22 | 46 | 52.2% | | Game 3 | 12 | 21 | 33 | 36.4% | | Match | 46 | 64 | 110 | 41.8% | The table says three things, and the third is the one worth discussing. First: Thuy Linh's baseline in this match sat between 32 and 36 percent. Game 1 was 32.3. Game 3 was 36.4. The two numbers are close enough that coincidence is hard to argue. In game 1 she entered the match and was pushed out of rhythm; in game 3, after winning a tight game, she returned to exactly that zone. A player who touches the same level twice in three games is usually