Istora, 8.4 Seconds, and the First Half-Meter: Data From 4,100 Serves

Istora, 8.4 seconds, and the first half-meter

Third game, 18-17. The player stands behind the service line, left hand holding the shuttle at hip height, thumb and index finger closing and opening twice. Istora Senayan is loud enough that the umpire has to call the score a second time. Then the noise stops — nobody signals it, seven thousand people just go quiet.

I start the stopwatch: 8.4 seconds from the umpire's call to the racket meeting the shuttle. In the first game, same player, it was 5.1 seconds. That three-and-a-half-second gap appears in no BWF statistics table. It exists only in my notebook, written in blue pen, in the margin of a match where I logged every single serve.

When every tournament stops, I finally hear my own pulse. That pause is also where badminton data runs out.

From the 1.15-metre ruler to the notebook

Since 2026, the BWF has enforced a fixed service height of 1.15 metres from the court surface, replacing the old lowest-rib rule. It is the biggest technical change to the sport in two decades, and it was designed to erase the advantage of tall players.

The geometry is simple enough. For a player 1.94 metres tall such as Viktor Axelsen, the 1.15-metre contact point sits around chest height, far below the shoulder line. The trajectory flattens, flight time to the short service line — 1.98 metres from the net in singles — shrinks, and the receiver, standing less than two metres away, gains a fraction of a second.

Over the past two seasons I hand-logged 4,100 serves across 62 men's singles matches at the Indonesia Open, the All England and the Japan Open. I sorted them into four buckets — short, high, flick and drive — then recorded the outcome of each service sequence. No software helped. Every number below may be a mistake of mine.

The evidence chain from the first half-meter

Short serves make up 71% of the sample. High serves 11%, drives 6%, flicks 12%. On the surface, that is evidence of a sport folded into the first half-meter of the court. Split it by game and the picture changes colour.

In game one, short serves are 76%. In game three the figure drops to 63% and flicks nearly double, to 19%. Same player, same opponent, two different sets of behaviour. What changes is not technique but the price of error: in a decider, a short serve that gets read means losing control of the rally at the most important score of the match.

On outcomes, points won after a short serve in my sample run at 47%. After a flick, 58%. Read only those two lines and the obvious conclusion is to flick more. It is not that simple: the rate of losing the point on the second or third shot after a flick is one and a half times higher than after a short serve. The flick is a bet with positive expectation and large variance, and in a 25-minute decider, variance costs more than expectation.

The receiving side has its own numbers. Returning a short serve, the players in the sample played a net shot 54% of the time, a flat push to mid-court 27%, and a lift 19%. Anthony Sinisuka Ginting is the receiver I tracked most closely, and his net-shot share in my notes runs above the sample average. Lift frequency rises noticeably when the receiver is ahead and falls when they are behind. Players do not pick the technically best option; they pick the psychologically best option at that moment.

Rally length differs by serve type too. After a short serve, the average rally is 7.3 strokes. After a flick, 5.1. After a high serve, 9.4. Each serve opens a different match: the high serve is physical torture, the flick is a single scalpel cut, the short serve is a chess game at the front court, where durable defenders such as Kunlavut Vitidsarn force opponents to play three more strokes before a point appears.

The counter-intuitive angle

None of the lines above prove that serving decides matches. The correlation is strong — short serves travel with a higher win rate — but the causal mechanism is murky. A player who serves short often may simply handle the net better than average, and that skill wins the point, not the serve itself.

The flaw is not in the source code; it is in the eyes of the person reading the source code. The brain files away the flick that won a point outright and deletes the ten short serves that quietly ended in four strokes. A shuttle clipping the net tape and falling on the opponent's side is not fate — it is a microscopic deviation between expectation and probability, and at a sample size of 62 matches I am in no position to call it a trend.

There is something else fans rarely see. This sport has no public metric equivalent to football's xG, no open database that lets anyone verify a single serve. Most granular tracking data at top-tier events is collected by private firms and resold to the betting market, where transmission speed is worth more than analytical quality. That is the least-discussed dark side of sports digitisation: live data does not reach the viewer, it reaches the bookmaker first. The more precise the number, the wider the distance between the people and the match.

Instant review — badminton's answer to Hawk-Eye — follows the same logic. It does not make controversy disappear; it moves controversy from the court to the screen, where people argue over a contact point reconstructed by animation.

In nine years of following this sport I have been loudly wrong. I said publicly that the 1.15-metre rule would strip tall players of their biggest weapon and that the next generation would belong to the short, quick and skilful. I reviewed 40 matches to test it and the conclusion landed elsewhere: they did not lose the right to serve, they lost the right to serve the old way. They compensated with flatter trajectories, with drives that kiss the line, and by dragging matches into short rallies where their reach still pays. Data does not lie. I just asked the wrong question.

The next cycle's signal

What I am watching over the coming months is not the serve but the format. If the proposal to shorten each game to 15 points is adopted, the number of high-leverage points in a match falls by about a third and pressure shifts onto the opening service sequences of each game — exactly the window where the high serve still survives. If that happens, I expect flick usage inside the first two points of a game to pass 25%, and that is a number I am ready to be proven wrong about.

The 8.4-second beat at Istora may never make an official record. The match still happens there, in the time before the shuttle leaves the hand, while every probability is still open.