Vietnamese Badminton Measures Smash Speed Every Day, But Nobody Records the Silence Between Points

In late July 2026, at the men's singles semifinal of an international badminton tournament held in Da Nang, the home player led 18-14 in the deciding game. Forty minutes later he walked off with his head down, beaten 19-21. The stands rose to applaud. Someone in the front row said the familiar sentence: he lacked stamina.

I stayed behind when half the arena lights had already gone dark. My handwritten notebook held 187 points from that match. For each point I recorded the time the player took to walk from where the shuttle last touched the floor back to the service line, the number of footsteps, the number of times he wiped sweat, the seconds he stood still before serving. Not a single line recorded shuttle speed.

Eighteen years ago, when I started breaking down badminton video for a coaching group in Da Nang, I also recorded shuttle speed. It took me a long time to understand: the thing measured most in this sport is the thing that decides least.

A badminton nation with many matches and little data

The domestic competition system is not short of tournaments. The national championship, the junior circuit, the Da Nang Open, the Ciputra Hanoi International Challenge, the Vietnam International Series, plus dozens of provincial amateur events run all year. More matches are staged in Vietnam annually than anyone could possibly sit through.

But the number of matches with detailed data is close to zero. The Badminton World Federation publishes deep statistics mainly at major events, where camera and sensor systems support electronic line calls. Down at International Challenge and International Series level, what remains is usually a score, a winner's name and a duration. Domestic national events sometimes do not even keep the score anywhere anyone will reopen.

For Nguyen Thuy Linh or Le Duc Phat, the number of matches per year that fall into the detailed-data tier can be counted on one hand. Most of their competitive life vanishes from every record. A player who competes thirty times a year leaves behind less analytical trace than a futsal player in a university tournament with a fixed camera.

Nguyen Tien Minh left the international stage, and with him went the longest reference archive Vietnamese badminton has ever had. Nobody digitised it before it ended.

The consequence is not a lack of inspiration. The consequence is that every argument about fitness, psychology and tactics in Vietnamese badminton stands on the feeling of whoever sat closest to the court. A coach says the player ran out of air. A commentator says the player lost composure. Both may be right. Neither can prove it.

Three layers of data, and the third one does not exist

The first layer is what gets measured. Shuttle speed after the smash, smash count, points won by smash. This is the easiest layer to sell because it produces round numbers for the big screen. A 400 km/h smash will always look better than a three-touch net exchange.

The problem is that the smash accounts for roughly one fifth of all strokes in an elite men's singles match. The other four fifths are lifts, drops, clears, drives, net shots and pushes that no spectator remembers by name. We are spending our entire measurement capacity on one fifth of the match.

The second layer is technical data that could be recorded if someone sat down to record it. Depth of the lift, height of the shuttle after being pressured, contact position on court, error taxonomy. These need a person with a notebook, not sensors.

The third layer is dead-time data. The silence between two points. The walk back to the service line. The seconds standing still. The number of sweat wipes. This layer barely exists in any professional badminton system, even though it occupies nearly half the real duration of a match.

Over two years of tracking Vietnamese national team matches on the domestic circuit, I hand-coded point by point from spectator video and coaching-staff footage. My sample is small, and I always state that. But a few patterns repeat often enough to be worth questioning.

Rallies lasting more than twelve shots make up about a quarter of all rallies. In the third game, that group accounts for the majority of points that changed the lead. The deciding game, in other words, is not settled by the prettiest rallies but by the longest ones.

Across the run from 18-14 to 19-21 in that semifinal, the average rest interval between points fell from roughly 11 seconds to roughly 7. The home player walked back to the service line faster, stood still less, wiped sweat less. He accelerated the speed of the silence.

Average footsteps per point barely moved. The legs did not fade. The rhythm of rest did.

And here is the number that made me rewatch three times: the fastest smash of the match belonged to the man who lost.

One more detail that no statistical system records. I coded forty lifts in the third game separately. Average depth of the shuttle declined across each sequence. By the final seven points, the lifts were landing noticeably shorter than at the start. A shorter lift means the opponent attacks earlier, from a higher position, with less defensive time.

Nobody records that number. It has no name in any report. Yet it is the direct path from one lost point to the next.

The "stamina" cliché and what it hides

"Stamina" is the most used and least verified word in Vietnamese badminton. When a player loses seven points in a row, stamina is the cheapest answer. It needs no data, invites no rebuttal, and forces no one to change anything tomorrow.

The data I have, small as it is, does not support that explanation in any simple way. The feet did not visibly slow. The intent did. The number of times the player chose an attacking shot from an unbalanced position rose across the final seven points. That is a signature of deciding faster than necessary, not of a drained battery.

Load management is another romanticised concept. A national-team player's calendar is not decided by an ideal spreadsheet. It is decided by prize-money events, friendly matches serving local media, and appearances that an administering body needs to show up for. When someone says a player is being "well load-managed", the next question should be: managed by whom, and protecting which number?

Ten years ago people threw away my xG report; today they pay me to read it. The distance between those two moments is not a distance of technology. It is the distance of somebody sitting down after the match ends.

Correlation is not causation, and here I must state my limits clearly. Shortening the rest interval does not cause lost points. Both may spring from a third variable I cannot measure: the belief that you are losing while you are ahead. Those last seven points are evidence for a hypothesis, not a proof. But a hypothesis with numbers behind it still beats a prejudice with nothing behind it.

The winning side that day did not play faster. They kept their own rhythm of rest while their opponent accelerated the speed of the silence. Data never shouts, it simply stands still and waits for people to be calm enough. Every point is a probability chain few bother to look at. But I look.

What deserves a column next season

The question I carried home from Da Nang was not how to measure more. It was: if every programme were allowed to add one column to its notebook next season, which column would it be?

If it were mine, that column would be the walk-back time to the service line. It is free, it needs one person with a stopwatch, and it will forecast the third game better than any shuttle-speed figure.

The intuition of a million data points never sleeps. Our job is to give it something to sleep with.