When Badminton Data Goes Empty: The Discipline of Verification in Analysis

On the night of March 12, I opened my computer at 11:40 p.m., after a men's singles semifinal at a Super 1000 event had ended in two games lasting 78 minutes. My habit had become a routine: sit down, open the tracking file, rebuild every rally, cross-check landing points against the height log, and only then write. But that night, the data file opened empty. No landing coordinates. No shuttle-speed records. Not even the movement log. All that remained was a still frame and the commentary ringing in my headphones. I stared at the screen for twenty minutes, fingers resting on the keyboard, and for the first time in years, I could not type a single word.

Outsiders tend to think the thing an analyst fears most is a lack of ideas. That is wrong. What we fear most is an empty data page while the pressure to publish remains. Because when there is nothing to measure, every sentence becomes a guess, and guesses are something I promised myself I would always label. That night I did not write. I turned off the machine, left a blank space on the screen, and told myself: a gap never disappears; it is only that we have not been patient enough to see it.

A profession growing faster than its data

Over the past ten years, badminton analysis in Vietnam has changed at a dizzying pace. When I began writing seriously, the only tools I had were a notebook and a television. I recorded every rally with symbols I invented myself: an arrow for a smash, a circle for a drop shot, a cross for a failed rally. Each match, I logged roughly sixty to seventy points, enough to redraw the rhythm of a match by hand.

Today, a Super 1000 event such as the All England or the China Open gives us thousands of data points per match. The Hawk-Eye system records landing points to the centimeter. High-speed footage lets us break a smash of more than 400 km/h into twenty separate frames. In theory, we are living in a golden age of badminton data.

But there is a paradox few people mention: the more data there is, the more clearly the gaps show. When you have only sixty hand-recorded points, you know exactly what you are missing. When you have ten thousand data points, you easily believe you have everything, and that is when mistakes begin.

I once fell into that trap. In 2026, in an analysis of a famous Danish men's singles player, I used tracking data to argue that he won because of superior court coverage. I drew heat maps, highlighted the zones where he moved most, and concluded with great confidence. A week later, a knowledgeable reader pointed out that his court coverage was in fact lower than his opponent's in that very match. He did not win by running more, but by running to the right places. I had misread the data because I wanted it to tell the story I had already decided on.

That lesson shaped everything I wrote afterward. Numbers do not tell stories on their own. The writer tells the story, and the writer always has bias. The task of an analyst is not to find data that supports a claim, but to find data that can break it.

Three layers of verification and the night of empty data

My process has three layers. The first is collecting raw data: landing points, speed, height, rally duration, distance covered. The second is identifying repeating patterns, things that appear at least three times in a match and at least fifteen matches in a season. The third is cross-checking against results: did that pattern actually lead to points, or was it mere coincidence?

My unbreakable rule is that before publishing a claim, I need a minimum of fifteen matches of data as a foundation. That is the number I arrived at after years of trial and error. Fifteen matches are enough for a trend to emerge, yet few enough that I do not fool myself into thinking I understand everything. Below fifteen matches, every conclusion is only a hypothesis, and I must say so clearly.

On the night of March 12, I had exactly zero matches of data. The file was empty. By my own process, I should have stopped. But I stayed two more hours, not to write, but to understand why the file was empty.

It turned out to be simpler than I thought. One of the two players on court had a sensor system that failed from the first game. The organizers only discovered it after the match ended. That means for 78 minutes, half the court was never recorded. I had complete data on the winner, and almost nothing on the loser.

What would a hurried analyst do? They would write about the winner. They would build charts from the remaining half of the data, and present it as though it were the whole story. That is the greatest temptation of this profession: telling half a truth in the voice of a complete one.

I chose a third way. I wrote about that very gap.

Reading the gaps on a badminton court

For years I was known for analyses of the space behind a defender in football. When I moved to badminton, I realized the principle had not changed at all. A singles badminton court is only 5.18 meters wide and 13.4 meters long, yet within that small space, every match is decided by the zones the shuttle never travels to.

Picture a player standing at the center, slightly toward the back. He waits for the shuttle. The opponent stands across the net, also waiting. In that instant, there are four corners and countless gaps between them. A good player does not hit toward where the opponent stands. He hits toward where the opponent will not arrive in time.

But reading a gap is not only reading physical space. It is reading space in time. A weak drop shot, placed at the moment the opponent has just lost momentum going backward, will be more effective than a 400 km/h smash into a player standing firm. The most dangerous gap on a badminton court is not at the corners; it is in the interval between an opponent's two footsteps.

That is why I always measure a player's recovery time after each shot. From the moment the foot lands after a smash to the moment the ready stance is restored, a world-class player needs about 0.4 to 0.6 seconds. In that interval, if the shuttle is returned to the right spot, the player is nearly frozen. I call it the freeze window, and it is the most important unit of measurement in my analysis.

Watching Viktor Axelsen in his prime, I noted that his smash was not only powerful but placed exactly into the opponent's freeze window. Watching Kento Momota in his golden era, I saw the opposite: he rarely smashed at full power, yet he controlled his own freeze window so well that it was almost never exploited. Two opposite styles, one shared principle: whoever controls recovery time controls the match.

The numbers that never appear in the stats sheet

When a match ends, organizers usually publish flashy figures: fastest smash speed, number of direct winners, serve-point conversion rate. Those figures are pretty, easy to read, and nearly useless to anyone who wants to understand the match.

What I care about lies elsewhere. I count how many times a player is forced to hit while off balance. I count the rallies in which the receiver pushes the opponent out of the center position. I measure the distance between a player's actual standing position and the ideal position they should have occupied.

A concrete example. In a women's singles quarterfinal I followed, the winner had a lower smash-winner rate than her opponent. Looking only at the stats sheet, one would think she won by luck. But when I rebuilt every rally, I saw the opposite. She smashed less, but each smash targeted a position that forced the opponent to cover maximum ground. After three such shots, the opponent lost her footing, and the fourth shot needed only a light placement. The smash-winner rate does not reflect that. The rhythm of the match does.

This is also how I used to follow Tai Tzu-ying. She has a style that renders stat sheets meaningless, because she wins with rallies that are not counted as direct winners yet place the opponent under constant strain. By contrast, An Se-young in her early rise stood out through physical foundation and consistency, and her most important metric was not winners but her win rate in extended rallies. Two different ways of reading, for two different players.

In Vietnam, Nguyen Thuy Linh is the case I have followed for years. What stands out is not her smash speed but her ability to hold the center position after defensive rallies. That is a metric that never appears on the scoreboard, yet it explains why she can pull even matches toward her side.

The serve and two decisive seconds

If I had to choose one rally to analyze an entire match, I would always choose the serve and the return of serve. Those are the two seconds in which everything is set. In those two seconds, the server declares intent, and the receiver declares an answer.

In singles, a high, deep serve pushes the opponent to the back court, opening a gap at the net. A low, tight serve locks the opponent near the net, creating a gap at the back. No serve is neutral. Every choice opens one gap and closes another.

I once spent six weeks recording the serve options of a top player and cross-checking them against point outcomes. The result surprised me: he did not have the technically best serve, but he could read the opponent before serving. He looked at the opponent's standing position, at which leg carried the weight, and chose his serve in under a second. That is not technique. That is match intelligence.

By the same principle, Lee Zii Jia's return of serve in his inspired matches always put the opponent on the back foot from the very first beat. He did not need a long rally to score. He needed exactly one beat to claim position. When the return of serve wins position, the rest of the rally is almost already decided.

When a diagram becomes a controlled lie

I always tell my students that every diagram is a lie. A diagram can never reproduce the full truth of a match. It is only an approximation, accurate enough to use, yet limited enough to demand humility.

Every diagram is a lie, but a lie accurate enough is called a tactic. What matters is not whether a diagram is right, but how useful it is and where it is wrong. An honest analyst always points out the place their own model cannot explain.

In the analysis of the night of empty data, I had to admit that my model had nothing to say. I had half the data, and half the data cannot produce a complete model. If I drew a diagram from that half, I would deceive my readers, and worse, deceive myself.

There is another memory I have carried for years. In the pandemic season of 2026, when every tournament was postponed, I spent six weeks excavating old matches, building models from what already existed. Those days taught me that a good model must survive outside the real court. If it is only correct when fresh data exists, it is not a model but merely a transcript.

A counterintuitive angle: when silence is a tactical decision

Here I must state plainly something the analysis world often avoids. In this profession, silence is treated as failure. An analyst who does not post for a week is seen as falling behind. Platform algorithms do not reward patience. Readers do not wait.

But there is an uncomfortable fact: most sports analysis online today is written in a state of empty data, and the writer never admits it. They fill the gap with emotion, with flowery words, with confident conclusions that have no basis. And readers, having nothing to compare against, believe it.

That is the biggest execution blind spot in the analysis profession. It is not that we lack data. It is that we pretend we have enough, because admitting a lack of data makes us look weak.

I once made the opposite mistake. I analyzed a big match and made a confident prediction about a matchup, and the result went entirely the other way. The online community called me a blind fortune-teller. But what I learned was not to stop predicting. It was to stop predicting without layering risk. A prediction without risk tiers is a promise, and I have no right to promise on behalf of the court.

If I am wrong, the thing I most want to be wrong about is this: perhaps readers do not need perfection. Perhaps they only need honesty. And perhaps one article saying I do not yet know is more credible than ten articles claiming I know everything.

Closing: a question for the next match

The pandemic taught me one thing: the pitch froze, but the data did not. On days without matches, I still wrote, still analyzed, still searched for gaps in old matches. The recent night of empty data taught me one more thing: there are times when the data freezes too, and in those moments, the most honest act is to wait for it to thaw.

The next match will come. The sensor system will be fixed. The data file will fill up again. But the question I carry will not change: when I publish a claim, am I telling the truth, or am I filling a gap with my own words? The answer to that question matters more than any number I could ever collect.

Data limitation: This article is based on my personal observation and analysis, along with public facts about the BWF World Tour competition system. Some details about the sensor failure that night are described from a personal log and have not been independently verified. Figures on court dimensions and shuttle speed are referenced from the official regulations of the Badminton World Federation. All tactical claims in this article should be viewed as hypotheses with layered risk, not absolute conclusions.