Badminton's Data Gap: When the Analysis Sheet Is Empty and the Eye Must Lead

That night in Surabaya I opened three windows on my screen at once: a live scoreboard, a slow-motion video feed, and a spreadsheet. The spreadsheet was empty. The match had ended two hours earlier. I had watched it in full, hand-noted nearly forty rallies, sketched a few markers for where the two players stood. But when I sat down to build the professional report, every cell I needed returned the same line: insufficient information to assess. No smash speed. No average rally length. No net-point win rate. Nothing at all. I stared at that emptiness for a long time and realized that most of the world's badminton is played without ever being measured properly. That silence of data is the story worth writing this week.

I have lived in Indonesia for eight years and work as a data consultant for football clubs, but my greater love is badminton, a sport I was assigned to commentate from 2026, when I hosted broadcasts of events such as the Table Tennis World Cup and the Sudirman Cup. In this country, badminton is not just a sport. It is identity, pride, the thing people argue about in coffee shops at eleven at night. But here is the paradox: a nation that produces world-leading players often records matches with tools far cruder than those used for a third-tier football league in Europe.

I am used to football datasets with hundreds of metrics, from xG and PPDA to average distance between lines. In badminton, even at some events on the World Tour system, data stops at the score and the match duration. The things that decide matches, such as the timing of a jump, the shuttle direction on the third shot of a cross-court exchange, or the wasted travel distance when forced into the left corner, mostly vanish once the match ends. Television cameras record for viewers, not for analysis. And so, when I sit down to build a report on a match I watched in full, I often have to admit I am writing a diary more than doing analysis.

The fault belongs to no single party. This is a structural problem. Football has money, cameras, and companies that sell data to both clubs and bookmakers. Badminton in many Southeast Asian markets still runs on passion and a couple of broadcast cameras. The result is an expensive paradox: the sport with the fastest decision-making among combat sports is the one recorded most thinly among popular sports.

A data gap does not equal a knowledge gap. It only means I must read the match with a different set of tools. I began reconstructing what I saw from disciplined memory and slow-motion video, dividing the match into spatial zones rather than columns of numbers. The core of the problem is this: when standard data is missing, the analyst is forced to shift from counting to locating. I cannot measure a player's smash speed, but I can see where he chooses to stand after serving, and how that distance changes when he falls behind. I have no net-point win rate, but I can see that on the third shot he almost always pushes the shuttle toward his opponent's left shoulder rather than the forehand side. Those fragments, added together, give me a picture clear enough to write about.

Let me tell you about a specific match. At a regional team event I followed in Jakarta, the host team's men's doubles lost the first game by a fairly clear margin. The scoreboard only said they lost. If I had only the scoreboard, I would write that they underperformed. But when I split the match by court zone, a pattern emerged: in the first game, most of their lost points came from the rear center area, where the two players left a V-shaped gap. They stood too close together. The average distance between them, which I estimated by eye from video, was only about one and a half meters, far too narrow for elite doubles defense.

In the second game they spread out. The number of points lost in the center area dropped sharply, while the number of rallies that forced the opponent to lift the shuttle increased. They won the second game and then the third. The final scoreboard still shows only aggregate values, but the real story lay in the distance between two people on court, something no standard badminton stats table measures in common practice.

This reminds me of a time I was wrong. In 2026, working as a data consultant for a Liga 2 Indonesia club, I used an xG model to advise the coach to push the defensive line high in a promotion play-off. The model said my team would generate 1.8 xG. We lost 0-2. The opponent deliberately played counter-attacking defense, sitting deep, turning every shot of ours into harmless long-range efforts. I had ignored PPDA and the location from which shots were taken. The model was not wrong. I was wrong when I made it speak instead of my own eyes.

That lesson applies even more sharply to badminton. If I only look at the score, I will never understand why a player who wins 21-15 is actually playing worse than one who wins 21-19. There are games where the winner is only better across the final six points, while for the first fifteen points the loser was the one controlling the rhythm. The scoreboard does not tell that. Only the eye can.

I learned to build my own metrics, simple but useful, for when standard data does not exist. I record how many times a player steps to the net within three seconds of the opponent lifting the shuttle, and I call it net reaction. I count consecutive shuttle-direction changes within one rally without a change of pace, a sign of rhythm control. I measure wasted travel distance in a lost game, because it says more about tactical disorientation than any aggregate value. These metrics are homemade, unverified, in no database. But they give me a language to retell the match. In a sport where official data is still so thin, that language is worth gold.

I remember once analyzing a young men's singles player at a domestic event. On the scoreboard he won two games to love, looking like an easy victory. But when I rewound, I counted that in the first game he lost balance seven times after jumping to smash, a sign of an unstable physical base. A weak opponent could not exploit it, but a stronger one certainly would. A 21-8 score does not reflect that risk. Only counting with the eye does.

In Indonesia, where I live and work, demand for badminton data is rising faster than supply. Domestic sports-data platforms are starting to build their own indices, such as a squad-depth index to measure a nation's personnel redundancy in each discipline. These indices are useful, but they are only trustworthy when the underlying data is thick enough. An index calculated from three matches cannot represent a whole competitive cycle, and I have seen far too many wrong conclusions built on such a thin sample.

I also look at how other sports solve this problem, to learn the method rather than copy the conclusion. In 2026, analyzing a team at the Euros, I dropped the habit of looking only at PPDA and moved to measuring the average distance between positions on the pitch. That approach showed me the team controlled matches by compressing space horizontally rather than by constant pressing. If I apply similar thinking to badminton, I can measure a men's pair by the distance between the two players in each rally, instead of only counting the points they win. A pair standing far apart may be defending well, while a pair standing close may be getting squeezed. The scoreboard cannot distinguish the two cases.

I have seen something similar in another field I follow: esports. There, every metric is available, every action recorded down to the millisecond. But precisely because the data is so complete, its value is eroded by another problem, competitive integrity, as betting regulations lag behind the speed of the market. Badminton sits at the opposite extreme: data too thin, so we struggle to understand. Those two extremes remind me that data, full or sparse, still needs a sober reader in the middle.

As a consultant, I usually ask the coach a question before presenting any metric. The first question is always: what do you want to change in the next match? A metric only has value when it answers a specific tactical question. If I tell a coach that his player has high xG, that information is useless. If I tell him his player loses 60 percent of points on the left half of the court when forced into the fourth shot, he has work to do in the next training session. Badminton is the same. Recording data only matters when it leads to a change that can be repeated on the training court.

Here I must be careful with myself. The greatest temptation of a data person is to turn a beautiful model into a truth. I used to think that if I measured enough, I would understand everything. The year 2026 taught me the opposite. When the pandemic hit and every league stopped, a club kept me on during lockdown to predict form when football returned. I used the first fifteen rounds of the season to build a model and advised the team to keep its possession game. When the league resumed, the team lost three in a row, because opponents exploited the empty stadiums to press harder, forcing us to lose the ball in our own half.

I had to admit my model was missing two variables no one measures: the crowd and social distancing on the pitch. The pandemic taught me that data too can be afraid, when the world stops, numbers are meaningless. By the same logic, a badminton metric measured in a packed arena will not match one measured in an empty hall. Cheering affects heart rate, heart rate affects decisions, decisions affect the score. No model of mine captures that whole chain.

So, for every conclusion I offer about badminton, I try to present at least two scenarios. An optimistic one: if regional federations invest in data-recording systems, we will have a genuine badminton analysis generation within ten years, with spatial metrics standardized and comparable across events. A cautious one: if data keeps serving only television and sponsors, much of this sport's tactical depth will remain word of mouth, and every analyst will keep building private metrics in the dark, no one able to verify anyone else.

There is another temptation I want to name plainly. Working between two badminton powers, China and Indonesia, I am often pushed into comparison. But comparing raw numbers between the two is a trap, because they define success differently. One culture may prize medal counts, the other prizes squad depth and the ability to survive cycles. Placing two values side by side without verifying origin and context is the fastest way to lie with data. I have seen enough such comparison tables to know they harm more than help.

I learned one thing from the Croatians, though that is football, not badminton. A team that does not win the title can still show me a truth hidden in small values, like the number of passes before winning the ball back. The value of data is not in the championship, it is in the process. In badminton, the real signal sometimes lies not with the tournament winner but with a player who exits in the second round with a rare break-point conversion rate. People only record the winner. I try to record the loser too, because there is more to learn there.

So what do I take from an empty spreadsheet? I learned that a player's true value lies where he runs and when he stops. Those unmeasured moments are where the match is actually decided. A sport without complete data is not a lesser sport. It is only forcing us to look more closely, and sometimes, to admit there are things we cannot yet know.

The question I leave for myself, and for those who do this work as I do: are we analyzing badminton, or analyzing its shadow on a screen? And if a major match takes place tomorrow that no one recorded properly, will we dare to say we do not know, instead of inventing a number to look as though we understand?