The Empty Cells on Vietnamese Badminton's Data Sheet

The Empty Cells on Vietnamese Badminton's Data Sheet

Seven Blank Cells on a Single Sheet of Paper

The official statistics sheet from a men's singles quarterfinal at the Vietnam Open, a tournament on the BWF World Tour Super 100 circuit, has twelve rows. Seven of them are blank.

I sat in the sixth row, notebook open, recording every rally in the third game. In my notebook I had the length of each rally, the number of times the home player advanced to the net after a low serve, and the seconds he spent walking back to his receiving position between points. On the organiser's sheet, those three columns did not exist. Nobody was being lazy. The collection system was built for television, and television only needs what fits into a three-second slow-motion replay.

I am thirty-eight years old. I have sat in commentary booths, in analysis rooms, and now in the stands with a notebook. Over twenty-two years I have learned one thing about this sport: Vietnamese badminton does not lack players. It lacks filled-in cells.

Ten years ago, when I brought an xG report into a football club's meeting room, they pushed it aside with the phrase "football is not arithmetic." Today those same people call me to ask for numbers. The same principle is now waiting for badminton. The difference is that here, the paper is still blank.

A Sport Recorded by the Naked Eye

Football is roughly two decades ahead of badminton in data infrastructure. Opta began collecting match-level events in the mid-1990s. By around 2026, expected goals had become common language in European analysis rooms. PPDA, the number of passes a opponent is allowed before a defensive intervention, was born from the need to measure pressing intensity. Those things are now sold as goods.

Badminton started later. The BWF introduced the Hawk-Eye-based Instant Review System around 2026, according to federation announcements. High-speed cameras appeared on the show courts of Super 1000 events, World Championships and the Olympic Games. But the system's original purpose was to judge whether a shuttle landed in or out, not to generate a data mine.

The result is a paradox few people notice. At a top-tier event, the organiser can record the landing coordinates of a shuttle to within millimetres, yet cannot publish a distribution of rally lengths by game. The technology is strong enough to keep referees from erring, but not open enough for coaches to learn anything.

What the BWF currently publishes per match on the World Tour is fairly coarse: game scores, points won on serve, points won on receive, longest rally, fastest smash, match duration. These are good numbers for a broadcast bulletin. They are not enough to answer the question every coach needs answered: what exactly did this player lose to?

I have spent most of my career telling people that data never shouts, it simply stands still and waits for someone calm enough to read it. In football, I had to convince people to read. In badminton, I have to convince them to record. Those are different problems, and the second is much harder.

Rally Length: The Discarded Metric

Of everything measurable in a badminton match, rally length is the easiest to measure and the least used.

All you need is a person with a stopwatch. From the moment the shuttle leaves the serving racket to the moment it touches the floor, that is one rally. A top-level men's singles match usually contains sixty to ninety rallies per game. Add them up and you have a distribution.

That distribution tells you more than any single smash. In men's singles, average rally length typically falls between eight and twelve shots. In women's singles it is usually lower, hovering around seven to nine. Men's doubles has the shortest rallies, sometimes only three to five shots, because the pace at the front court is too high to sustain more.

But the real value lies in the tail of the distribution, not the mean.

Imagine two players with the same average rally length of nine shots. Player A has very few rallies exceeding twenty shots. Player B has fifteen percent of rallies exceeding twenty-five shots, and wins seventy percent of them. Same average, entirely different players.

In my tracking work, players in group B usually possess something the scoreboard does not display: the ability to turn a long rally into a tool that applies reverse pressure. The longer it goes, the more the opponent reveals their position. The finishing shot does not come from power; it comes from the other person's fatigue.

People often assume a long rally signals two evenly matched players. That is half true. The other half is the signature of one player actively controlling the tempo and another enduring it. Telling those two states apart by eye is difficult. Through a rally-length distribution, it appears in thirty seconds.

In Vietnam, I have never seen a domestic tournament publish this distribution. Not for lack of people. For lack of anyone assigned to the task.

The Silence Between Points

I have an odd habit when watching badminton: I time from the end of one point to the moment the next shuttle is served.

The rules allow a short interval between points and a longer one when a side reaches eleven points in a game. Those seconds are the grey zone television cuts away to show commercials, and the grey zone data ignores.

Yet behaviour in that grey zone is remarkably stable.

Some players walk straight to the receiving line, stand still, wait. Some circle the court once before returning. Some wipe their face, choose a new shuttle, adjust a shoelace, glance at the stands. Each behaviour consumes a different number of seconds, and that difference is not random.

In a dataset I collected myself at international events held in Vietnam and a few regional tournaments, players who had just lost a point took on average three to five seconds longer to return to their receiving position than after winning a point. The number is small. But it appeared in nearly every case I tracked, and it tended to grow as the match entered the deciding game.

How to read it is a coach's job. It might signal distraction. It might be a way of self-calming. It might simply be the aftermath of breathing. But it exists, it is measurable, and it is free.

An empty stadium is not silence; it is the answer to a thirty-year assumption. I wrote that line in 2026, when Europe's top leagues played without crowds. Home win rates in the Premier League and La Liga fell from roughly forty-six percent to roughly thirty-eight percent. Most fans concluded that crowds create home advantage. That conclusion is correct, but only partly.

The rest lies elsewhere. When the schedule compressed, teams lost preparation time, and preparation time is what generates most of the advantage. Crowds only amplify what is already there.

Vietnamese badminton has a natural opportunity to test this. Domestic tournaments draw large home crowds in some provinces and near-empty stands in others. If someone sat down and compared home players' win rates at those two types of venues, we would have a more valuable answer than any online argument. I have proposed this a few times. Nobody has taken it on.

Receiving Position and the Central Corridor

Football has a simple, brutal concept: if you let the opponent into the middle of the pitch, everything else becomes harder.

Badminton has an equivalent, and it is just as brutal.

When a player receives a low serve, where they stand determines the entire structure of the following rally. Stand deep and you have time to handle a deep serve but lose the front court. Stand close to the service line and you win the front court but expose the space behind you. There is no correct choice. There is only the choice that fits the type of opponent.

In an analysis I prepared for a group of youth coaches, I sorted receiving rallies into three categories based on the receiver's starting position: deep, mid, and tight. Then I counted the receiver's point-win rate in each category.

The results were not as uniform as I expected. The tight group produced the highest win rate among players with a defensive-counterattacking style, and the lowest among pure attackers. The deep group produced the opposite.

This sounds obvious once written down. It was not obvious in the meeting room. For years I watched tactical decisions being made on a shared belief that the best receiving position is the one a player feels most comfortable in. A player's comfort is data. But it is data about the player, not data about the match.

Nguyen Tien Minh and Twelve Years of Unrecorded Data

No Vietnamese player is better suited to illustrating the data gap.

Nguyen Tien Minh was born on 12 February 2026. He reached world number five in the BWF men's singles rankings in 2026, according to rankings published by the federation. He competed at four consecutive Olympic Games: Beijing 2026, London 2026, Rio 2026 and Tokyo 2026. He reached the quarterfinals of a World Championship.

That is a career at a level Vietnamese badminton has touched only once.

And almost none of that career has data.

When Tien Minh was at his peak, the match-level statistics the BWF published were even cruder than now. Analytical cameras were rare. Matches were filmed, but the footage was scattered, low quality, and largely unindexed.

I tried something I believed useful: gathering whatever footage remained accessible from the major matches he played, and manually coding a set of basic metrics. Number of rallies lasting more than fifteen shots. Rate of net approaches after a low serve. Frequency of the back-boundary push.

The work took weeks per match, and the results still lacked continuity because the footage was incomplete.

Even with incomplete data, however, one thing became clear. Tien Minh did not win with speed. He won by dragging opponents into long rallies and then changing tempo at exactly the right moment. That is a teachable model, and it was never written down as documentation.

In many other countries, a player of that calibre would become a living dataset. Every rally would be coded, classified, archived. The next generation would learn from it. In Vietnam, when he retired, most of what he left behind lived in the memories of people who had watched him play.

Memory is data. But memory cannot be copied.

Nguyen Thuy Linh and the Shift of Gravity

Nguyen Thuy Linh was born on 20 November 2026. She has entered the group of thirty top-ranked players in the world in women's singles, according to BWF rankings. She qualified for the Paris 2026 Olympic Games and has appeared in the main draw of numerous World Tour events.

Before Thuy Linh, Vietnamese badminton's centre of gravity for two decades was tied to a single name in men's singles. After she emerged, that centre shifted to women's singles.

The shift means more than a results sheet. It reveals something the data of Asian federations makes clear: in women's singles, the gap between the thirtieth and the tenth is not as wide as in men's singles. Consistency matters more than absolute power. A women's player with a solid physical base, few unforced errors and a sense of when to accelerate can go further than a men's player of the same ranking.

This is the kind of conclusion data can support. Unforced error rates in women's singles are usually higher than in men's singles, because rallies last longer and involve more shots. That sounds like bad news. In fact it is good news for a developing badminton nation, because it means improving consistency pays off directly on the scoreboard, faster than chasing smash speed.

Nobody measures consistency systematically. People count errors. But counting errors without recording the rally that produced them is bookkeeping, not analysis.

The Men's Doubles Assumption and a Calculation Nobody Has Done

In conversations about Vietnamese badminton, one phrase appears often: our strongest discipline is men's doubles.

I decided to test it by retrieving BWF ranking data from roughly the past decade across five disciplines: men's singles, women's singles, men's doubles, women's doubles and mixed doubles.

The results did not support the claim.

The highest ranking a Vietnamese player or pair has reached over the past two decades came in singles. Men's singles produced Nguyen Tien Minh at world number five. Women's singles produced Nguyen Thuy Linh inside the top thirty. In the doubles disciplines, Vietnam's best results have generally settled between world number one hundred and two hundred, with occasional deeper runs at continental events.

In mixed doubles, the pair of Do Tuan Duc and Pham Nhu Thao drew attention at some regional tournaments but never entered the world's leading group on a stable basis.

So why does the assumption persist?

There is a reasonable explanation. Doubles is the discipline in which smaller badminton nations can compete fastest, because it demands fewer years of individual development than singles. A country with few elite players can still assemble a workable pair in a short time. That creates a sense of progress. The sense is not emotionally wrong. It is only wrong in terms of data.

Vietnamese badminton is not weak in doubles. It is weak in the middle tier, the group of players ranked between one hundred and three hundred in the world, and that group exists mainly in singles.

That is where most athletes are, where data is thinnest, and where opportunity is scarcest. A singles player ranked two hundredth in the world has no dedicated analyst, no one coding footage, and often nobody to sit and review their matches with them.

I once spent an evening rewatching footage of four matches by four different young players. None of them had a single written record of what they had done wrong. They remembered the feeling. They did not remember the position.

The Development Pipeline and the Earliest Forgotten Data

There is a category of data I consider more important than match metrics: data about young athletes' bodies and development.

Between the ages of fourteen and seventeen, bodies change fast enough that competitive results become hard to interpret. A player who wins consistently for a year may simply be someone who matured earlier than their peers. Another who loses for two years may be in a growth spurt that temporarily costs them coordination.

No spreadsheet records any of that.

Over years of working with coaching groups, I noticed a repeating pattern. Athletes cut at fifteen were usually not the ones with the weakest technique. They were the ones with good technique whose bodies developed out of step with their cohort's standard. Selection criteria are based on match results, and match results at that age are strongly influenced by temporary physical factors.

If we recorded height, arm span, lateral movement speed and injury history for each athlete quarterly, we would make different decisions. This requires no advanced technology. It requires one patient person with a spreadsheet.

Load Management: A Romanticised Number

Over the past decade or so, load management has become a concept discussed with near-religious reverence.

In badminton, it shows up in talk about packed calendars. Top players enter many World Tour events each year, plus team events such as the Thomas Cup, Uber Cup and Sudirman Cup, plus SEA Games and Asian Games. Plus an Olympics every four years.

The story told is that players must rest to protect their bodies.

The reality is often different.

After years of reading calendars and cross-checking withdrawal notices, I have one simple observation. Most withdrawals attributed to physical condition do not occur before important tournaments. They occur after them. Which means athletes still take the court when it matters, then collapse when it no longer does.

What does that mean for data? It means match counts do not reflect real load. A player might play ten matches in six weeks at maximum intensity, or ten matches in six weeks with three thrown. These two cases look identical on paper and are entirely different inside a body.

To measure real load you need positional data per rally. The kind Hawk-Eye can generate but does not publish.

Hawk-Eye and the Wall Nobody Mentions

This is the least discussed part of any conversation about technology in sport.

Hawk-Eye is a commercial system. It is installed under contract with a tournament organiser. The data it produces belongs to the party paying for it. At top badminton events, that data is used for reviews and broadcast graphics. A very small portion becomes published statistics. The rest sits in the system and goes nowhere.

Football encountered the same problem and found a way around it. Data companies buy collection rights, code the data, then sell it back to clubs, bookmakers and media. A market formed.

Badminton has no such market. The audience is smaller, there are fewer events, and national teams' willingness to pay is lower. So the data stays with whoever owns the technology, and the people who need it most are the ones without access.

In Vietnam, we are not even in that game. Domestic tournaments have no Hawk-Eye. International events staged in Vietnam install it on a few courts, but the data is not shared with the home coaches.

People talk a lot about lacking machines. The problem lies elsewhere: even with the machines, we have nobody standing at the other end to read the data.

Video Is the Only Data Mine Everyone Already Owns

Let me return to my notebook.

Amid every infrastructure limitation, there is one data source anyone with a smartphone owns. Video.

A match filmed from a single fixed angle is enough to code most of what is needed. Rally length. Type of finishing stroke. Receiver's position. Number of net approaches. Time between points. All readable from a frame that need not be high quality.

The obstacle is time. Manually coding a three-game men's singles match takes two to three hours for an experienced person, depending on detail.

For years I did that work alone. I recoded matches I had watched live, purely to check where my memory of the match was wrong. The answer was almost always: wrong in many places.

A spectator's memory is dominated by dramatic points. The final thirty seconds of a game occupy most of the memory's storage, while the twenty minutes before are compressed into a vague feeling. Video has no feelings. It has only events.

If every youth team in Vietnam had one person spending twenty hours a month coding footage, we would have a data repository within two years. The cost is far lower than hiring a foreign expert for a month.

This is the kind of proposal that wins little applause, because it produces no beautiful images. It only produces results, and results take time.

Correlation Is Not Causation: The Beginner's Trap

Here I have to say something I believe is the most important point in this entire piece.

An argument is spreading through the sports world: teams with more data win more. It is illustrated with examples of European clubs investing heavily in analysis.

It is wrong for at least two reasons.

First, teams investing in data are usually also the richest teams, and money buys good players. Looking at their results and concluding that data produces success is a causal error.

Second, and more importantly, data does not act. It only describes. A report sitting in a drawer scores no points.

I once witnessed this in its purest form. In 2026, I presented a report showing that the home team allowed opponents nine central corridor entries into the box in one match, while we only shot from distance. The report was dismissed. The team kept losing along exactly that pattern for many matches afterwards.

The problem was not the data. The problem was that nobody had the authority to change how the team played based on it.

The same thing will happen to Vietnamese badminton if someone builds a perfect collection system. We will have a beautiful spreadsheet. Decisions will not change.

The mistake I made at twenty-nine was not that the numbers were wrong, but that I forgot people need time.

I learned that over the years that followed. When you present data to a coach, you are not just handing over numbers. You are handing over a different way of seeing their own profession. That takes patience and a companion, not just a spreadsheet.

What Transfers From Football to Badminton, and What Doesn't

When I moved from football analysis into badminton tracking, my first question was: which metrics carry over.

The short answer: very few.

Expected goals works in football because every shot can be assigned a probability based on location, shot type, number of defenders in front and goalkeeper position. Badminton has no equivalent structure. A smash cannot be assigned a scoring probability without knowing the opposite player's position, the contact height, and the state of the previous rally's breathing.

PPDA does not transfer either. In football it measures passes an opponent completes before an intervention. In badminton, the nearest equivalent is the number of net approaches per rally. But approaching the net in badminton does not imply applying pressure. Sometimes it is a forced action after a weak shot.

What transfers best is mindset. How you frame the question. How you refuse to conclude before the sample is sufficient. How you test a popular belief against a simple table.

Counting an Unforced Error Correctly

There is a small technical problem I believe has a large effect, and almost nobody notices it.

In badminton, when a rally ends, the scorer must decide who won the point. The question sounds simple. It is not.

There are situations where the shuttle lands out after a stroke by player A. The point goes to player B. But who created the situation? If player B had just played a shot that forced player A to cover the full width of the court and then return from a position of lost balance, then the error belongs to B's construction, not to A's incompetence.

Current recording systems, at most tournaments, simply credit B with the point and A with an error. There is no column for "points created."

Football solved an analogous problem by separating goals from expected goals. Badminton has not.

The consequence is that today's badminton statistics sheets systematically undervalue players who build points through sustained pressure, and overvalue players who finish with a powerful stroke that only appears at the end of the rally.

This has a practical consequence in talent selection. Constructive players are often judged as lacking power, because the numbers do not see their work.

Media sells excitement; I sell probability. Fans deserve both.

I do not mean to oppose the media. I am in this profession because I love the moments. A rally saving three match points is what makes this sport worth watching. But behind that moment there is always a chain of decisions, and that chain becomes visible if we are willing to record it.

Money, Sponsorship and the Value of One Data Cell

There is a question I often get from sponsors: what is this data for.

The commercial answer is simple. Sponsors buy presence. Presence is measured in broadcast time. Data does not increase that time.

But there is another layer of value rarely exploited.

When a tournament has detailed data, it can tell its own story. A match with the longest rally in the tournament's history. A player with the highest point-win rate after a low serve. A pair with the most net approaches in a single game. Those details generate content, and content generates views.

In Vietnamese badminton tournaments today, media content relies mainly on results and quotes. Results are what everyone has. Quotes are easily forgotten.

A properly filled data cell can become a headline. The cost of filling it is close to zero.

A Contrarian View: The Gap Is Not the Biggest Problem

Here I want to offer a point contrary to most of this piece.

I have spent many pages talking about empty cells. But if someone asked me what Vietnamese badminton's biggest problem is, I would not answer that it lacks data.

The bigger problem is that the data we do have is recorded incorrectly.

Every national tournament has results. Every match has a score. But the way errors are classified, the way points are credited to whoever created the situation, the way effective net approaches are counted, all of it is done by habit rather than by any standard. The result is a database that looks complete but cannot be compared across tournaments, across years, across coaches.

An incomparable database is worse than an empty one, because it creates the illusion of understanding.

This is why I often tell younger colleagues: do not build a system. Write the standard first.

A two-page definition document specifying what counts as an unforced error, what counts as a created point, what counts as an effective net approach, is worth more than any software. Software can be bought. Standards must be agreed by people, and that demands patience few are willing to spend.

There is another paradox worth putting on the table. Vietnam has no infrastructure advantage, but it does have a scale advantage. We have few enough players that a small group can sit down and agree on recording methods. Countries with thousands of athletes cannot do that as easily.

Being small is a disadvantage in competition and an advantage in organisation. Fewer people to convince.

Signals to Track in the Coming Cycle

Over the next two years I will track four things, and I suggest anyone interested in Vietnamese badminton track them too.

First, how domestic tournaments publish statistics. If one tournament starts publishing a rally-length distribution, that signals someone in the organising body has understood the problem. It is the easiest metric to produce and the most valuable.

Second, the appearance of an official analyst on the payroll of a national team or training centre. Not a foreign expert hired per project, but a Vietnamese person in a fixed position, with a contract and responsibility.

Third, the number of young athletes with quarterly physical tracking records. If that number rises, we will see results in about five years, and those results will appear in the one-hundred-to-three-hundred world ranking group.

Fourth, the question of women's singles. Whether Nguyen Thuy Linh's emergence produces a following generation in the same discipline, or remains an isolated phenomenon. The answer lies in how many Vietnamese women enter the world's top three hundred over the next three years.

A Closing Thought

I return to the sheet of paper with seven blank cells.

The organisers that day did exactly what the system allowed. Nobody made a mistake. Nobody was lazy. There was simply one decision nobody had ever made: that Vietnamese badminton would record more than a news bulletin requires.

Data does not emerge from technology on its own. It emerges from a human decision, usually a small one, usually unannounced, usually made by someone whose name nobody knows.

The intuition of a million data points never sleeps. But to have a million points, someone has to be willing to record the first one.

I still keep my notebook. In it are small lines recording what the sheet of paper left blank. Perhaps one day a young coach will ask me how to record them.

On that day, Vietnamese badminton will begin to calculate what we can currently only feel. And when you can calculate it, you can teach it. And when you can teach it, you no longer need to wait for a talent to be born. You can produce ten.