Rally Length and the Data Gap in Vietnamese Badminton

In a second-round match on the BWF World Tour, on the 47th rally of the deciding game, Nguyen Thuy Linh retreated to the left corner and released a cross-court smash. The arena erupted. I sat in the press row, pen in hand, and the pen stopped mid-line. Not because the smash was beautiful — across more than four decades of watching this sport I have seen thousands of such smashes, and most of them passed without leaving a memory. The pen stopped because of the number.

That rally was the 47th. Across the third game, the average rally length between her and her opponent hovered around 14 shots — nearly 40 percent higher than in the first game. Yet her scoring rate inside the first five shots of the third game fell sharply. The two data points contradicted each other. One said the match was being dragged long, that both players were being ground down. The other said her ability to close rallies early was fading. I looked at the numbers, then looked up at the court, and asked myself: where does this number stand in the flow of the match?

That is the question I have carried through my entire career. The data is not wrong. I had simply forgotten to ask where it stands.

To understand why a 47th rally made me stop writing, it helps to talk about the position of Vietnamese badminton on the world map, and about how data is produced in this sport.

Nguyen Tien Minh was the trailblazer. At his peak he put Vietnamese badminton on the world map with a style built on patience and stamina — a weapon many Asian players of his era had not fully exploited. After him, a next generation rose. Nguyen Thuy Linh in women's singles and Le Duc Phat in men's singles became regulars at BWF World Tour events. This is the result of a long process, not a sudden leap.

But there is a paradox I have observed for years. Badminton is a sport with an enormous amount of data. The Badminton World Federation publishes statistics on every rally, shuttle speed, distance covered, scoring rates by phase. Yet in Vietnam, most analysis still stops at win or lose, qualified or eliminated. Between that ocean of data and the fans, there is a large gap no one has filled.

I came to badminton by a roundabout road. I hold a degree in broadcasting, have hosted programs on the Table Tennis World Cup and the Sudirman Cup, and later worked in sports betting analysis. That trade taught me one cruel lesson: data is not neutral. It can be filtered to tell any story. A tipster only needs to offer a number that is technically correct but contextually wrong, and readers will believe it.

So when I sit in front of the statistics of a badminton match, I always start with a question different from the crowd's. Not what this number is, but under what conditions this number was measured. A player who runs six kilometres in a match may be showing stamina, or being pinned down relentlessly. The same number, two opposite stories.

Let us begin with rally length — the metric I consider the breathing rhythm of a badminton match.

In the match above, average rally length rose game by game. Around 10 shots in the first, around 12 in the second, around 14 in the third. Looking only at this number, concluding the match grew tighter and stamina more worn is reasonable. But the context changes completely once we separate the phases.

From shot 1 to shot 5 of the third game, Nguyen Thuy Linh's scoring rate dropped. She was winning fewer short rallies. In badminton, short rallies are where speed and initiative are expressed — whoever controls the opening rhythm usually wins. When that rate falls in the deciding game, it is not merely a sign of a long match. It is a sign that the opponent has read the service rhythm.

Yet rally length kept rising. If the opponent had read the rhythm, the match should have ended faster, not longer. This is the paradox of rally length — it can signal endurance, or deadlock, and the two cannot be told apart from a single number.

To break the paradox, I add another layer: distance covered per rally. When distance in short rallies falls while distance in long rallies rises, the picture sharpens. She was no longer winning quickly, but still controlled the long rallies by pulling her opponent side to side. This is an in-match tactical shift — from fast attack to counter-attack by attrition.

This is where the analyst's discipline is tested. It is easy for me to write that Nguyen Thuy Linh won thanks to superior stamina. But the data does not say that. The data says she changed tactics mid-match. Stamina is merely the condition that makes that change possible. Without the second data layer, I would have reached a conclusion right about the result but wrong about the cause.

Here it is worth adding a note on the tournament system, because it directly shapes how data should be read. In the BWF World Tour, events are split into several tiers, and ranking points are distributed differently. This means players do not enter every event with the same goal. Some events they use to accumulate points, some to test tactics, some to conserve energy for a bigger stage. If we pool all matches into one dataset, we are mixing things that should not be mixed.

I once made this mistake. Years ago, I built a dataset pooling all of one player's matches across a season, then drew conclusions about form. The results were skewed. The reason is simple: a match at a low-tier event, when the player had just returned from injury, cannot be treated the same as a final. Since then I always split data by event tier and by stage of the season.

Then there is the environmental factor. During a period when I spent months rewatching matches to study the effect of crowds, I noticed something interesting. When arenas had no spectators, some players tended to reduce their pressure intensity in the opening rallies. This directly affects rally length. A silent arena does not make a match easier; it makes it different. And raw data cannot capture that psychological pressure.

Now let us widen the view to Vietnamese badminton as a whole. When comparing Vietnamese players with the world's top players in the same discipline, a repeating pattern appears. In rallies under eight shots, the gap is not large — sometimes Vietnamese players even edge ahead in scoring rate thanks to shuttle speed. But from the ninth shot onward, the gap widens markedly. This is a metric I have tracked for years, and it is stable to a suspicious degree.

But — and this is the but I want you to hold on to — that stability may be an illusion. The sample of Vietnamese players facing the world's top players is small. They tend to meet at major events, where psychological pressure and playing conditions differ completely from regional events. If we sample only from major events, we are measuring something that might be called the gap on the big stage, not the gap in true strength. The two are not the same.

This is exactly why I write this section before writing about results. A number stripped of context is nothing but a lie dressed up. A number about long-rally win rate can be skewed by the fact that a player only met strong opponents in the semifinals and final, when energy was already spent. It does not measure stamina. It measures the schedule.

At this point I want to turn to an angle few who write about badminton touch: data itself.

In my years in betting analysis, I saw something I consider the darkest side effect of the digitisation of sport. Betting companies do not need match-fixing. They only need live data. When every rally is recorded in real time, odds can shift within seconds, and bettors are pushed into a game where they are always one beat behind. Data, created to make sport transparent, becomes a tool to exploit the very fans.

I do not write this to urge anyone to quit betting. I write it to say that when you look at a statistics table sponsored by a bookmaker, ask what story it was selected to tell. A metric can be entirely accurate and still entirely leading.

And there is one more thing. We tend to assume that more data means more understanding. That is true only when the reader can place a number in context. With Vietnamese badminton, most fans do not yet have that tool — not because they are weak, but because no one has taught them. This is the gap I try to fill by writing slowly, writing clearly, even writing things that seem obvious. The people who look competent actually need it too.

The rally-length metric, in the end, is only a stethoscope. It lets us hear the heartbeat, but it does not tell us what the patient is thinking. To understand that, the one holding the stethoscope must know how to be silent.

In my work I try to keep one discipline: no more than three hours a day on a single topic. I learned this after once letting curiosity pull me into a deep dive and missing major changes elsewhere. An honest analyst must acknowledge their own limits. Each of my articles, therefore, always has a section for what I missed.

I return to the 47th rally. That cross-court smash, seen only through the scoreboard, is a point. Seen through rally length, it is a sign of a tactic adjusted mid-match. Seen through context, it is an answer to a question the opponent posed from the 30th shot.

The question I took home after that match was not who won the third game. It was: if tomorrow ten more Vietnamese people start recording the rally length of their country's players, will we still be surprised when they beat a top-twenty player in the world? Or will we realise the signal was there all along, and no one had been willing to sit long enough to hear it?

When the court falls silent, I hear the whisper of the underlying data most clearly.