Vietnamese Table Tennis Is Missing Match Data: Where Should Analysis Begin?

In a national table tennis championship final, the electronic scoreboard clicks to 11-9. The stands applaud, the two players wipe sweat, the umpire signals a change of service. If I turned to the person beside me right then and asked how many points the winner had taken directly off the serve in that game, how many ended on the third ball, how many rallies ran past seven contacts, I would almost certainly get a shrug. Everyone knows the score. Nobody records how the score was made.

I open my notebook and rule four columns: serve, receive, third ball, long rally. Then I sit and look at four empty columns. After more than thirty years of reading sports data, the most familiar situation remains this one: a real match, a real result, and not a single line of data to analyse.

Vietnamese table tennis has an obvious paradox. We remember medals very clearly, we remember the names of champions, we remember wins on the regional stage, but we preserve almost nothing at rally level. A name like Nguyen Anh Tu or Tran Tuan Quynh is invoked mainly through titles, not through point structure. The previous generation — Doan Kien Quoc, Nguyen Khoa Dieu Khanh, Mai Hoang My Trang — left their mark through audience memory, and memory cannot be queried.

In the WTT system, everything works the other way. Every serve, every point, every rally is logged point by point, enough to reconstruct a match in chronological order. A Grand Smash awards its champion several times the points of a Contender, and the ranking operates on a rolling 52-week deduction: points won a year ago expire automatically. That mechanism creates what I call points-defence pressure. A player can slide down the list not because they lost to anyone, but because there is nothing to replace the points that just fell away. To see that pressure, you need a points history. We do not have one.

The rule milestones of this sport are all moments when the data was rewritten. In 2026, the ball went from 38mm to 40mm. In 2026, scoring moved from 21 points per game to 11. In 2026, the hidden-service rule took effect. In 2026, speed glue was banned. In 2026, celluloid was replaced by plastic. Each time, data from the previous era became hard to compare with the next, and analysts had to start over. Without a baseline dataset, every rule change wipes the slate.

To analyse, you first need something to analyse

A serious analytical chain needs five minimum conditions. Identity first: who played whom, on what date, at which event. Then point structure: who scored, and how, at which beat of the rally. Next comes event context: what tier, and where in the Olympic cycle. After that, squad state: which player is at which point of a career, and how many points they must defend. Finally, the transmission chain: which doors this result opens or closes for entry slots, for sponsorship, for the youth pipeline.

When all five are empty, the only honest outcome is a table full of the words "insufficient information". That sounds dull. But I have seen the price of filling gaps with inference. In 2026, I bet on xG. The V-League answered with a shock. I built a model on aggregate data, concluded that one team played better and would advance, then watched that team go out in the next round. The mistake was not in the data. The mistake was that I refused to say I lacked data. The 2026 World Cup taught me a second lesson: data is never a single layer. A tournament has a group-stage layer, a knockout layer, a refereeing layer, a scheduling layer. Read one layer and conclude, and you misread the whole picture.

With table tennis, the blank space is larger still. We do not even have the first data layer. Tiering data is how I keep calm during a crazy transfer window, and here it is the only way not to fool myself.

The four columns in my notebook are not trivia. The share of points won on serve shows whether a player owns an active weapon, and whether that weapon still works in the fifth game or has run dry. The share won on the third ball shows the ability to convert service into attack — the quality of the first forehand loop. The distribution of rally length shows who is controlling the tempo. Added together, those three measures answer a question the score never answers: was this game won because someone took the initiative, or because someone made fewer errors?

At national-team level, the blank space spreads further. Without rally-level data, picking a doubles pairing rests on only two grounds: recent results and the coaching staff's feel. Both have value, and both are indirect evidence. A player with a strong receive rate who scores heavily on the fourth ball may suit a partner who attacks hard on the third ball better than a partner with a better record. Without data, doubles compatibility is only discovered after a loss.

One hypothetical shows the cost of compensation. A player wins three straight games after trailing 5-10. The media will call it nerve. An analyst short of data can go further, attributing it to conditioning or pressure tolerance. But the cause may simply be that the opponent changed service patterns too early, or that the umpire called a service fault at a sensitive moment. Without a logged point sequence, the cause is misattributed. Misattribution leads to misdirected training, and misdirected training costs a whole cycle.

The worry is not missing numbers — it is the habit of compensation

The default response to missing data is emotional storytelling. A player who comes from behind is explained with the word "nerve". A player who loses three straight games is explained with "mentality". Those words sound reasonable and are nearly impossible to verify. Placement heat maps walk the same road: they look scientific, they are colourful, and they hide a player's real role inside a tactical system. A pretty placement map does not necessarily correspond to a deliberate service structure.

The deeper problem is correlation read as causation. A player who wins many long rallies is usually described as having a strong physical base. It may be simpler: their opponent plays a safe pushing game, so long rallies appear more often, not because one side is fitter. Without opponent-classification data, the conclusion is wrong at the root.

I do not believe in a future where everything is measured. I believe in a smaller principle: where something has not been measured, say it has not been measured. After seven years, I trust the silence between two numbers, because that silence tells you how much your conclusion is worth.

A signal worth tracking

If one day the national championship organisers agree to log points rally by rally — just four columns, like my notebook — Vietnamese table tennis analysis will have its first foundation. A young player entering the national team will no longer be judged by a few remembered matches, but by a data series long enough to reveal a trend.

One question I cannot yet answer remains. If we have never recorded how a game was built, are we commenting on table tennis, or commenting on our own memory?