Vietnamese Table Tennis and the Data Void: The Regional Gap Begins in an Empty Notebook

118 points and one notebook

Two players walked into the fifth game at 2-2. Eleven minutes later the game was over. In that window I logged 23 points: who served, what spin, at which stroke the rally ended, who won. By the end of game seven the count stood at 118.

Those 118 points only mean something next to thousands of others. Since 2026 I have hand-logged and then digitised 1,284 matches — the national championship, youth events, some regional matches, and WTT fixtures I could watch on stream. In total, 41,286 points.

This log is mine. It does not replace official federation or organiser data. The sample is small, skewed toward what I can watch, and dependent on whether I recorded things correctly. Data does not need me to believe in it. Data needs me to verify it.

What kept me coming back for six years was not the good matches. It was the sense that the gap between Vietnamese table tennis and the regional leaders is being explained wrongly — by feeling, by tradition, by phrases such as insufficient fighting spirit or too little international exposure. Those explanations are not baseless, but they cannot be measured. What cannot be measured cannot be fixed.

Football has xG. Table tennis does not.

I came to table tennis the long way around. Before that I spent years on football data, where a V.League match can be dissected with xG, passing maps, touches in the box, PPDA. I assumed every combat sport had built similar infrastructure. I was wrong.

Vietnamese table tennis publishes almost no point-level data. Ask how a national championship match unfolded and you will get a description: player A has a strong forehand loop, player B blocks well, player C is mentally fragile in deciders. Nobody hands you service points won. Nobody hands you rally-length distribution. Nobody hands you conversion at 10-10.

That is not entirely anyone's fault. Table tennis has the fastest tempo in the racket-sport family: an average point lasts under five seconds, and the recorder must decide in a split second whether a point ended with a loop or a failed push. Hand-logging 41,286 points taught me that the biggest error is not missing a point but misclassifying one.

The competition system itself is clear. Domestically there is the national championship, the national youth championship, open events and the sports festival, concentrated among a familiar group of units: Hanoi, Ho Chi Minh City, Hai Phong, Da Nang, the Ministry of Public Security, the Army, Vinh Long and a few private academies. Internationally, the WTT system tiers upward from Feeder to Contender, Star Contender, Champions and Grand Smashes, ending with the year-end Finals. The ITTF world ranking is published weekly.

The problem sits between those two layers. A Vietnamese player can compete at a WTT Feeder and hold a world ranking, yet nowhere records how often that player wins at 10-10. Meanwhile regional rivals — Singapore, Thailand, Malaysia, Indonesia — have had internal analysis units doing this work for years.

Four variables I track

Within the 41,286 points I isolate four families of indicators. They are not a complete indicator set. They are what one person in the stands with a notebook can record consistently over six years.

When the serve is not yet a weapon

In the log, 24,317 points had a Vietnamese-system player serving, and 9,842 had a regional player serving. Vietnamese-system players won 49.8% of their own service points. Regional players won 55.1%.

In table tennis the serve is an advantage. The server controls spin, height, placement and the tempo of the first contact; winning more than 50% of service points is normal at a high level. So the Vietnamese cohort in my sample sits exactly on the line: they do not lose on serve, but they do not turn the serve into a weapon either. If the serve does not create an advantage, the whole point structure afterwards has to run on footwork and reflexes — the two things that erode fastest across seven games.

I once treated this as a purely technical issue, a missing heavy sidespin serve. Rewatching footage changed my mind. The problem is usually the third ball: after serving, Vietnamese players often lack a clear attacking plan for the most common return, so they push the ball back and put themselves on the defensive.

10-10 is where the gap shows

This is the part of the log I trust least, because the sample is tiny. Across 1,284 matches I recorded 214 deuce points involving Vietnamese players and 134 involving regional players.

| Group | 10-10 points | Won | Rate | |---|---|---|---| | Vietnamese players | 214 | 88 | 41.1% | | Regional players | 134 | 79 | 59.0% |

An eighteen-point gap on such a small sample must be read carefully. But its direction is consistent with what I see in the hall: at 10-10 Vietnamese players tend to shorten their options, choose the safer serve, choose the lower-risk return. At deuce, safety is often the riskiest choice, because regional opponents have prepared for exactly that pattern.

I do not want to turn this into a psychological truth. There is a duller and possibly more accurate explanation: Vietnamese players rarely face the same opponent archetype a second time at 10-10, so they never get the chance to correct.

Long rallies: unmeasured ground

Average rally length in the log is 4.1 strokes. Split by length, the picture sharpens.

| Rally length | Points | Vietnamese win rate | |---|---|---| | Three strokes or fewer | 21,940 | 54.7% | | Four to six strokes | 13,902 | 48.3% | | Seven strokes or more | 5,444 | 46.2% |

This inverts a popular intuition. Vietnamese table tennis is often described as durable, resilient, good at extending rallies. My log does not support that description: the cohort wins most points in short exchanges and loses its edge as rallies lengthen.

A simple explanation exists. Short rallies are decided by basic technique; long rallies are decided by fitness, movement and the ability to rebuild a plan mid-point. If Vietnamese players mostly win short, that suggests the technical foundation is not bad. The problem lies in endurance and reorganisation once a point passes the seventh stroke.

I should add that this indicator is affected by a variable I do not measure: the opponent. A player who specialises in short rallies pulls the distribution short regardless of anyone's fitness. I do not systematically tag opponents by style, so every conclusion here is conditional.

The U15 cohort and the attrition problem

Tracking youth cohorts is the most time-consuming and unsettling part of this work. I selected eight players who reached the national youth quarter-finals in each age group between 2026 and 2026 — 40 players, 20 men and 20 women — and checked how many later appeared in the national top 16 at any point.

The result: 9 of 40, or 22.5%. Six of 20 among men, three of 20 among women.

That number is easy to misuse. It does not prove youth coaching is weak, nor that youth events are worthless. There are many reasons a player leaves the track: academic pressure, career options, injury, or simply that professional table tennis income cannot keep someone in the sport for another decade.

But it sets up a concrete comparison. If only about a fifth of the youth elite reaches national-team level, then investing in earlier talent identification is far less efficient than investing in retention at age 22. In table tennis, the highest-return investment is not finding more young talents; it is reducing how many players leave the system between 17 and 21.

The longest-surviving men's generation is the group around Dinh Quang Linh and Nguyen Anh Tu; on the women's side, the group around Nguyen Khoa Dieu Khanh, Tran Mai Ngoc and Nguyen Thi Nga. They lasted not only because of skill, but because they had stable competition slots and a place within their sponsoring units. That is an organisational variable, not a technical one.

Explanations the data does not support

Lack of international exposure

This is the most common explanation and the one I once used. Vietnamese players rarely compete abroad, so they fade in deciding points.

In the log I counted international matches played by young Vietnamese players: 6.4 per player per year in 2026–2026, against 11.2 in 2026–2026. A 75% increase in matches.

Deuce win rate across the two periods: 39.6% and 41.1%. The 1.5-point difference sits entirely inside the margin of error for a sample of a few hundred points. International matches rose sharply while the indicator said to measure match experience barely moved.

This does not mean international competition is useless. It means playing more is not automatically gaining experience. A player who plays eleven matches a year against eleven different opponent types, in eleven different contexts, never gets a second encounter to correct anything. Experience is not the number of matches played. Experience is the number of times you meet the same problem again and solve it.

In table tennis, where reading spin and predicting placement decide most outcomes, repetition beats variety. That is why strong teams run long training blocks against the same partners rather than chasing tournaments.

Home court is an advantage

When the Bundesliga played in empty stadiums, I realised home advantage is just a variable waiting to be deleted.

In table tennis, the crowd matters more than in football: the gap between spectators and table is a few metres, and applause lands directly in the server's ear. I used to treat this as a clear advantage.

In the log I isolated matches from the 2026 regional games hosted in Vietnam. In that group, the home players' service error rate was 8.9%, against 6.1% at neutral venues. Their deuce win rate fell to 33.3%, though on just 18 deuce points — far too small to conclude anything, and I note that in the spreadsheet.

Still, the direction matches another observation: for young players, a home crowd raises the psychological cost of a missed serve, and the safe response under pressure is to serve easier. An easier serve means losing the advantage from the first stroke. In that case home court stops being an advantage and becomes a variable that amplifies whatever state the player is already in.

Youth titles as a guarantee

In many table tennis nations, an age-group title is read as a signal for the future. In my log, the correlation between youth results and elite results exists but is far weaker than most people assume.

World Cup 2026 taught me one thing: the model did not collapse — I was the one who believed it absolutely. I once ran a model that assumed the stronger team on paper would go far, and it failed badly. The lesson was not to abandon models but to separate a model from belief. The table tennis equivalent: a youth champion reflects early specialisation and good training conditions at 13 to 15. It predicts nothing at 25, when everything depends on surviving the system, on competition slots, and on whether someone pays for six hours of daily practice.

The error section

My first V.League dataset had hundreds of errors, but it taught me cleanliness better than any course. I want to list the equivalent errors in this table tennis log, because this is the part readers need most in order to judge it.

I only record what I see. I watch northern matches far more easily than southern ones, so the sample leans toward units such as Hanoi, Hai Phong and the Ministry of Public Security.

I cannot record spin, ball speed or precise placement. Those are the three most important variables in modern table tennis, and all three are absent.

I classify ending strokes by eye. A missed loop and a push that gets attacked can look identical from the fifth row.

The deuce sample is too small. With 214 points, the standard error around a rate near 41% is roughly 3.4 percentage points, so the confidence interval is wide. I need at least 400 to 500 points before claiming anything about deuce ability.

I have confirmation bias. I began following table tennis with the hypothesis that Vietnam's problem is structural, and every subsequent observation of mine tends to reinforce it. The only defence is publishing the spreadsheet for others to audit.

Signals to watch

If the log has practical value, it lies in pointing out what should be counted over the next twelve months.

The number of times a Vietnamese player meets the same opponent archetype within a single year. This is the indicator I consider most important, and the one nobody tracks.

Service points won at regional level. If the 49.8% figure in the log reflects reality, this is the cheapest and fastest intervention point in the entire system.

A deuce sample large enough to say something — roughly 400 points or more collected over at least three seasons.

Publication of point-level data at the national championship. Nothing elaborate is required: a four-column record per match, maintained for three years, would create more value than any short training camp.

The 2026–2026 U15 cohort, tracked over three consecutive years. That is the only way to know whether 22.5% attrition is structural or just noise.

I read a player through thirty variables before listening to a commentator. But those thirty variables only matter if more than one person in the fifth row writes them down. What I want most in the coming year is not to predict who wins, but to turn the 41,286 points in my personal spreadsheet into an open dataset so anyone can find where I was wrong.