Fourteen Blank Pages in a Scouting File: The Data Limits of Badminton's Transfer Window

On the screen is a spreadsheet of forty-one tabs. I opened it on a Tuesday morning in Surabaya, when the light outside was still pale and the ceiling fan turned slowly overhead. It was a scouting file for a player a club had asked me to assess before they finalised their squad for the new season. Tab one held the name. Tab two held height, wingspan, weight. Tab three held the record. Tab twelve asked for average movement speed within a rally. Tab nineteen asked for the win rate after trailing in the second game. Tab twenty-seven asked for recovery time between games. Tab thirty-eight asked for projected transfer value over twenty-four months.

Forty-one tabs. Fourteen of them returned the same string of characters: N/A.

The player was not weak. The problem lay elsewhere: nobody had measured the things the spreadsheet was asking for. The club's scouting department had asked exactly the right football questions, but had placed them on a sport that lacks the corresponding data infrastructure. That gap does not sit inside the player. It sits inside the system, and it radiates across the entire transfer window.

That is why I stayed with the file longer than planned. In the middle of a transfer window, when every club is hunting for one number to believe in, the most interesting subject turns out to be the empty cells.

Badminton's transfer market does not operate like football's. There is no centralised window that opens and closes, no publicly listed transfer fee, no global governing body coordinating deals along the lines of a football federation. A player moving from one team to another happens mostly in the silence of internal contracts with a national federation, with a host club, with a personal sponsor, and with a local authority. What the public sees is only the visible part: a national squad list published at the start of the year, a coach changing seats, a player appearing in new colours at the first event of the season.

That market has three tiers, and each tier has a different data logic. The first tier is movement inside a federation: a young player promoted from a province to the national training centre, or sent back down after a cycle of missed targets. The second tier is the club market, where team leagues such as the Premier Badminton League in India, the Purple League in Malaysia, the S/J League in Japan or the Danish team championship hire foreign players on short contracts of a few weeks. The third tier is transfer of allegiance between national federations, constrained by World Badminton Federation rules on waiting periods and residency, and therefore almost impossible to complete inside a single season.

For Vietnamese badminton, the first and second tiers are where money and people actually move. One player sits on the national team roster, receives a training slot at the centre, and is dispatched to events in the BWF World Tour system according to internal criteria. Another player belongs to a strong provincial team, covers part of the cost personally, and hunts entries at lower-tier events to accumulate ranking points. The Vietnam Open, a Super 100 event, has long served as a shop window: it is where a young player can appear before regional scouts without flying to Europe.

But this is where I want to pause. Those three tiers operate on three datasets of very different resolution, and none of them supplies what a modern scouting file demands. The federation tier runs on coach reports and training logs, almost never digitised. The club tier runs on short contracts and the subjective assessment of a coaching staff across two weeks of competition. The allegiance tier runs on legal documents, not on performance data.

In other words, badminton's transfer window is a market operated by professional memory more than by models. And professional memory, as I learned in the most expensive way available, tends to remember the beautiful rally and forget the repeated one.

Based on my experience watching matches at World Tour events over many years, I can say that most scouting decisions in Southeast Asia are still made after watching three to five matches of a player, plus one conversation with someone who has coached that player. That is a valuable process, but it is not a verifiable one.

Start with what the BWF system actually measures. The hardware of elite badminton data has four layers. The first is results and scores, fully recorded in the World Badminton Federation's tournament software, together with a ranking system computed from events played and rounds reached. The second is draws and brackets, which allow a player's path through each tournament to be reconstructed. The third is video adjudication, operated by the World Badminton Federation since 2026 under the name Instant Review System, allowing players to request review of line calls or disputed situations. The fourth is broadcast data, including shuttle speed at some major events and slow-motion replays for viewers.

Those four layers are enough to answer who won, by how much, and at which event. They are not enough to answer what a scouting file needs: how many metres this player covers in a game, in what shape they run, at which scoreline the speed drops, and which spaces on court are left open when they are forced to their left.

The largest gap between badminton and football is not money or popularity, but the fact that football has commercialised the measurement of off-ball movement while badminton has not. In football, a player can be assessed by presses per ninety minutes, by average distance between lines, by a PPDA index measuring how many passes an opponent is allowed before an intervention. In badminton, the equivalent of PPDA would be the number of rallies an opponent sustains before being finished by an imposing stroke, or the metres a player is forced to pay for every point lost. Those indices are entirely measurable with existing technology. They are not measured at system scale, because nobody has paid for them to be.

I still remember how I wrote about Croatia at the 2026 World Cup. Their PPDA was only 9.2 in the group stage, not the highest-pressing team at the tournament, yet they had the highest rate of ball recoveries in the opponent's half at 12.4 per match, thanks to the pressing timing of Luka Modric and Ivan Rakitic. When that piece was shared widely, I understood something: what makes readers stop is not the number but the mechanism behind the number. Croatia did not win the trophy, but they showed me a truth hidden inside a number. Badminton is missing exactly that mechanism at the public layer.

This has a concrete consequence in the transfer window. Without movement data, a player's value is priced by the three easiest things to observe: world ranking, results at major events, and age. Those three variables correlate with quality, but they are confounded by schedule, by draw luck, and by the federation's own selection policy. A world number 30 sent to eighteen events a year can accumulate more points than a world number 22 sent to nine because of budget. The ranking records outcomes, not the conditions that produced them.

I have made exactly this mistake. In 2026, while working as a data advisor for a Liga 2 club in Indonesia, I used an expected-goals model to advise the coach to push the defensive line higher in a promotion play-off. The model predicted 1.8 expected goals. In reality we lost 0-2 because the opponent sat deep and countered, turning every shot into a harmless effort from outside the box. I had ignored the pressing index and the origin position of each shot. I looked only at the total. The model was not wrong; I was wrong when I made it speak in place of my eyes.

In badminton, that trap takes a different shape. A player can hold a 68 percent win rate across a year, but if you split that figure by opponent, by court side, by how many days of rest they had between events, the picture can invert completely. I have seen players with excellent head-to-head records against Asian opponents who can barely beat European players built for long rallies, and the reverse. The aggregate hides precisely the part the scout needs.

So when the scouting file returned fourteen blank pages, my first reaction was not to fill them with guesswork. My reaction was to read those blanks as an independent dataset. The blank on tab twelve says the club has never hired a motion-tracking provider. The blank on tab nineteen says nobody has logged the scoreline progression rally by rally. The blank on tab thirty-eight says no valuation model has ever been applied to this sport in this market. Each blank is a statement about institutional capacity, not about the player.

There is a striking paradox here. Badminton has the densest scoring rhythm of any widely played combat sport. A three-game match can contain more than a hundred rallies, each a chain of decisions across five to fifteen seconds. Structurally, this sport generates more data per minute of play than football. But data only has value when someone records, standardises and publishes it. Structural wealth without recording infrastructure produces memory, not knowledge.

Look at basketball, where every possession is logged by action type, and the distance becomes obvious. Basketball has data on the origin of every shot, on the efficiency of every lineup, on every player's average time on the ball. Badminton has equivalent potential but lacks the commercial ecosystem large enough to sustain a dedicated data industry. And without that industry, clubs are forced to build measurement capacity in-house, something very few badminton clubs in Southeast Asia have the resources to do.

In that landscape, the Vietnamese national team and its regional peers start from the same data baseline, whatever their competitive records. Players such as Nguyen Thuy Linh or Le Duc Phat enter major events with a thinner international file than opponents of similar level, and the thinness lies not only in the number of matches but in the number of information dimensions a foreign scout can look up. A European player competing in a fully broadcast domestic league generates more data touchpoints than a Southeast Asian player who appears at three Asian events a year.

This produces a quantifiable information asymmetry. I once compared the number of fully recorded matches for two groups of players of similar ranking, one competing mainly in Europe and one mainly in Asia. The difference in available footage usually runs at two to three times. For a scout, that means the first group is assessed on evidence and the second on inference. In the same transfer window, two players of equal ability enter negotiations with very different levels of transparency.

There is another factor usually ignored in transfer discussions: when the rules change, old data loses value. From March 2026, the World Badminton Federation applied a service rule requiring the shuttle to be struck below 1.15 metres from the court surface, replacing the previous rule based on the server's waist. The change landed unevenly: it reduced the advantage of tall servers and forced the rest to rebuild their service technique. Any model trained on pre-2026 data without isolating the rule change will produce distorted forecasts for the period after.

I raise that detail because it illustrates a principle I have followed for years: data is the prayer beads, but intuition is the candle, and I light both whenever I read a match. The beads tell you what to read. The candle tells you when the beads have gone stale.

Back to the spreadsheet. After reading all forty-one tabs, I wrote a short note to the club, and in it I set out three scenarios rather than one conclusion. Scenario one: the player fits the current system and will adapt quickly, provided they are fielded in a familiar competitive position. Scenario two: the player has a sound technical base but lacks data on durability across three consecutive events, so injury risk cannot yet be quantified. Scenario three: the player may be a good investment if the club accepts the cost of building in-house measurement capacity during the first six months.

Those three scenarios are not evasion. They are the only way to be honest with the data. When the world stopped in 2026, I learned something I still write down whenever my models fail: the pandemic taught me that data also knows fear, and when the world stops, numbers mean nothing. A model built on fifteen rounds played in front of crowds collapsed the moment stadiums emptied and teams switched to fiercer pressing. The crowd variable was not in the equation, and that is why the equation was wrong.

There is a counterintuitive angle I want to place beside all of the above. People assume more data improves transfer decisions. That is true, but only when the organisation can digest it. A club with no analyst, no footage cross-checking process and no habit of writing post-match notes, suddenly handed a detailed data pack, will often make worse decisions than if they had relied on the coach's eye alone. A data pack creates a feeling of certainty, and certainty without verification capacity is a category of risk.

I have watched this repeat across the region. A club buys an international data package, sticks labels on players, and makes decisions based on indices designed for a different sport, in a different competition system, at a different match density. Correlation appears. Causation does not.

The second counterintuitive angle concerns adjudication technology. Badminton's video review system, like referee assistance technology in football, has not reduced controversy. It has moved controversy from the court into the review room and into the grey zones of the law. In badminton, disputes rarely concern whether the shuttle landed in or out; they concern whether a player had the right to challenge at the decisive moment, and whether a rally counts as deliberate distraction. Technology clarifies the easy part and illuminates the hard part.

That has a direct implication for the transfer market. As decisive rallies are increasingly recorded and analysed, a player's value lies not only in winning points but in managing risk in scrutinised moments. A player who knows when to use a review, who can stay calm after an adverse call, holds a skill that is measurable and priceable. Right now, almost nobody measures it.

Another angle worth discussing is competitive integrity. In emerging sports markets, betting money moves faster than supervisory capacity. In badminton and esports, the gap between the speed of wagering and the speed of regulation is creating grey zones that public data cannot illuminate. A player competing at a lower-tier event, where no full recording system exists, creates a blind spot that neither scouts nor regulators can see. That is a second, institutional reason to invest in data infrastructure at small events.

I want to return to a line I keep in mind when assessing any player: a player's true value lies where he runs and when he stops. Where he runs tells you his system. When he stops tells you how deeply he reads the game. Both are spatial data, and both sit among the things my spreadsheet could not answer.

For Vietnamese badminton, I believe the opportunity lies precisely in this gap. A small federation cannot compete with major powers on coaching budgets, but it can compete on the quality of its record-keeping. Systematically recording every internal match, every sparring session, every domestic tournament, using one shared set of index definitions, is labour-intensive but cheap. It creates an asset no major federation can buy back: high-resolution data about itself, across many years.

That also changes how we read a young player's record. Instead of asking how many matches this player won, we ask how this player won, against opponents with which characteristics, and at which stage of the match. The second question is harder to answer, and precisely for that reason it creates competitive advantage.

I am not claiming data will solve Southeast Asian badminton's transfer problems. I am only saying that most decisions today rest on far thinner evidence than existing technology permits. That gap is an opportunity and a risk. An opportunity for the organisation that invests first. A risk for the organisation that keeps buying players on instinct and selling them on instinct.

When I closed the spreadsheet, daylight had fully arrived. The fourteen blank pages were still there. I did not fill them with guesses. I noted in the margin that this was a list of things to measure over the next six months, and attached a small proposal: hire one full-time videographer for every internal club match across a season.

The cost of that proposal is far lower than the cost of one wrong contract. But it demands something clubs rarely possess: patience with things that do not pay back immediately.

The coming transfer window will give us a clear signal. If clubs in the region begin publishing basic movement indices alongside results, that marks a shift from buying players to building systems. If they still publish only rankings and win counts, the story will keep being written from memory, and the blank cells will keep appearing on tab twelve.

The question I leave for myself, and for anyone doing this work in Southeast Asia: if tomorrow a club handed you a fully populated spreadsheet, with no empty cells left, would you have the capacity to read it differently than when it was blank? Because complete data only has value when the reader has already learned to doubt himself first.