The Empty Split Column: Vietnamese Swimming's Data Problem After the Ánh Viên Era

The men's 1500m freestyle final at SEA Games 31, in May 2026, filled every seat at the Mỹ Đình Aquatics Palace. On the electronic board the finish time appeared clearly, but the split column — the time for each 100m — was empty. I sat in the sixth row, stopwatch in hand, recording every lap myself to rebuild the race. That night I understood something: the biggest problem in Vietnamese swimming is not inside the pool. It sits in the fact that almost nobody records what happened inside it.

A swimming nation can win medals and still have no data with which to read itself back.

From 2026 to 2026, Vietnamese swimming lived through the era of Nguyễn Thị Ánh Viên. Born in 2026 in Cần Thơ, she brought home 25 SEA Games gold medals, broke multiple Games records and competed at two Olympic Games. When she retired, the question people asked was who would fill the medal gap. But the larger gap lay elsewhere: a data system that had never been built for the next generation to learn from.

Across years of analysis work for training centres, I have seen one pattern repeat. An athlete shines, and the whole sport concentrates resources on that person. When they stop, no document is detailed enough for a successor to follow. Ánh Viên left behind 25 gold medals, but very few training data series that could be reused.

Nguyễn Huy Hoàng, born in 2026 in Quảng Bình, is the next pillar. He races the distance events, swam at the Tokyo 2026 Olympics and carries expectations at recent SEA Games. Trần Hưng Nguyên, born in 2026, is the coming generation in the medley events. Behind them are Nguyễn Hữu Kim Sơn, Phạm Thanh Bảo and Võ Thị Mỹ Tiên. The list is not short. The problem lies in how we read that list.

When I compiled Vietnam's swimming results across recent SEA Games for an internal report, I started with the split table. This is where everything shows itself, and also where Vietnamese data is thinnest. Organisers publish finish times, but the publication of 50m or 100m splits is not consistent between editions or between events. To get data, I had to press a stopwatch by hand, rewatch video and cross-check against the regional federation's result sheets. Three sources, for one number.

What I found: the final-50m speed of most Vietnamese swimmers in middle and distance events tends to fall below their average speed for the whole race, while Singaporean and Thai swimmers over the same distances usually hold or lift their pace in the closing stretch.

That number does not say everything, but it points in one direction. A structural conditioning issue, not an individual error. One swimmer slowing in the final 50m can be tactics. Ten swimmers doing the same is a training programme.

I once treated models as scripture. Now they are only a compass — but without one, we get lost.

To test it, I split the data by distance. Over 50m and 100m, the gap between Vietnam and the regional leaders sits mostly in start reaction and underwater technique. That is a zone improvable through technical coaching, and Vietnam has in fact closed some of the distance. Over 200m and 400m, the gap sits in pacing. Over 800m and 1500m, it sits in aerobic base and the ability to hold technique while tired.

Those three zones need three different programmes. For years, they were trained almost the same way.

Here I must be explicit about sample size, because this is where it is easiest to go wrong. What I present is a hand-recorded dataset, small in sample, covering only Vietnamese swimmers plus a regional comparison group at a few SEA Games. I call it a hypothesis, not a conclusion. Turning it into a conclusion would require a national database logging splits from the junior level, and at least five years of continuous records. A good model needs five years, not five meets.

Numbers do not lie, but people always find ways to lie with numbers. When a swimmer breaks a national record, the news posts the final figure. Nobody asks whether their first 50m was fast or slow, whether they held or lost the third 100m. Those numbers exist; nobody has bothered to record them.

At the same time, another structural issue appears when I rank Vietnam's events by medal density. Vietnamese swimming concentrates on distance freestyle and a few medley events. That is a sound strategy when you have one exceptional athlete, but a risky one when you have only one. Singapore spreads across events: freestyle, butterfly, backstroke, with several swimmers meeting standards. Thailand has depth in breaststroke and medley. The difference is not in peak talent, but in the number of contact points with medals.

That is a portfolio-management problem, not a coaching problem. A team strong in three events has a higher probability of winning than a team excellent in one, if total talent is equal. Vietnamese swimming is betting on a single door.

Another point few notice: the calendar. The SEA Games usually fall in the middle of a training cycle, not at peak form. Vietnamese swimmers tend to peak at the SEA Games, while Singaporean and Thai swimmers treat the SEA Games as a stepping stone to continental and world meets. In other words, we are measuring two different things on the same scoreboard.

When the pool empties, every model collapses. I rebuild from the burnt remains of the data.

There is one more factor analysts often skip: short course and long course. Many regional nations use 25m short-course meets to build technical foundations and increase race exposure, because each race involves more turns and underwaters. In short course, turn technique and post-turn push-off are trained more densely. Vietnam has few official short-course meets in the national system. That is a source of data and experience left unused.

On staffing, a purely technical problem: the number of coaches specialising by stroke remains thin. One coach often covers an entire event group. This directly affects the quality of technical data: nobody has time to record stroke rate and distance per stroke for every swimmer in every session.

Without those numbers, all technical comparison is just sensation. And sensation does not store.

On the market side, Vietnamese swimming has almost no sustainable individual sponsorship ecosystem. Athletes live mainly on state funding and medal bonuses. When they retire, no structure converts their knowledge into shared assets. Ánh Viên opened her own swimming academy; that is a direction worth watching, but it is the exception rather than the rule.

And there is a question about records I want to raise carefully. A record exists inside its context: swimsuit type, pool type, water conditions, the technology of the moment. Comparing a 2026 time with a 2026 time without noting context is a comparison that is dishonest in data terms. Many social-media arguments rest on numbers torn from their context.

People measure Vietnamese swimming by SEA Games medals. I measure it by how many times a race is recorded in enough detail to be reused. The two measures barely correlate.

This is the counterintuitive point. A SEA Games that succeeds on medals can be a failure for the system, if nobody learns from it afterwards. Conversely, an edition with few medals but full split data, conditioning profiles and technical records can be the foundation for the next cycle.

Correlation is not causation. We see medals rise and conclude the system is good. But medals can come from one extraordinary individual, not from the system. Ánh Viên is proof: she brought 25 golds, but those medals did not automatically produce the next Ánh Viên. One exceptional athlete can mask an incomplete system for an entire decade.

The unread portion of the data is also large. I do not know exactly what happens inside junior training plans, do not know each swimmer's lactate threshold, do not know how much they sleep or what they eat. Those are dark zones. Honestly speaking, I can only read the outer layer of the story, and that is why I call my conclusion a hypothesis.

The signal to watch in the next cycle is not medals. It is whether someone starts logging splits from the junior level, whether the federation publishes internal data, whether a race is stored well enough for a twelve-year-old to open it and learn.

Reputation is only a name. What remains is always how you read the race.