Dissecting the Lane: Why Vietnamese Swimming Needs Its Own Data Set

In December 2026, I sat in the fourth row of an indoor pool, a worn stopwatch in hand, eyes fixed on the water. Each time a swimmer touched the wall, I pressed the button, wrote the number in a small notebook, then pressed again for the next lap. After the session, I checked my notebook against the electronic scoreboard. Average error: 0.3 seconds — enough to change who took a place in the final.

Twenty years later, the stopwatch lies still in a drawer. In its place: touchpad sensors under the wall, tracking cameras and software that splits every swimmer's race down to the hundredth of a second. The paradox is this: data has multiplied a hundredfold, while our ability to read it has barely moved.

That is where the question I want to dissect begins: is Vietnamese swimming short of a talented athlete, or short of a data set clean enough to see that talent? The question has no tidy answer, and that is exactly why it deserves dissection.

Context: a swimming nation running on memory

For more than a decade, Vietnamese swimming has been tied to one name and one medal table. Nguyễn Thị Ánh Viên — born in 2026 in Cần Thơ — has been the centre of every story: multiple Olympic appearances, dozens of SEA Games gold medals, and domestic feats no regional rival matched in the same period. She was the flag-bearer of a sport once treated as a sideshow.

A sport built around one individual always pays a price eventually. When that star is absent or slowing down, the whole system shows its gaps. In those gaps, people reach for the most visible causes: too few athletes, too little funding, too few pools. All three are true, but none is sufficient.

What is rarely mentioned is that Vietnamese swimming runs almost entirely on memory and instinct. A good coach is trusted because this one can see it. A swimmer is rated highly because his stroke looks beautiful. Those judgments are not wrong, but they cannot be copied, passed on, or verified. We are running a sport measured in hundredths of a second on a system of evaluation measured in impressions.

Regional rivals have moved ahead. Singapore built a training centre with physiological data and video analysis years ago. Thailand, Indonesia and Malaysia have been folding tracking and split analysis into their programmes. The gap at the SEA Games is not simply physical; it sits in the information system behind each lane.

There is another point worth noting: Vietnamese swimming is usually viewed through the lens of medals. A Games is judged by golds, rarely by the data it generates. That habit turns every training cycle into a gamble: win, and we do not know why; lose, and we do not know why either. A sport that does not know why it wins cannot repeat a win.

The geography of Vietnamese swimming raises its own data problem. The main centres cluster in Hà Nội, Hồ Chí Minh City, Đà Nẵng and a few places with tradition such as Cần Thơ and Quảng Bình. Each has different pool conditions, climate and training plans. Without a shared data standard, a time swum in Cần Thơ cannot be compared with one swum in Hà Nội, even in the same 100m freestyle. Lacking a common standard, national selection leans on a feel for the quality of the lane rather than on a normalized number.

Reading the lane through data: from splits to stroke rate

When I moved into sports data analysis, I carried the habit of a man who once timed by hand: trust the number, but check it three times. In swimming, there are five layers of data any serious swimming nation must be able to read.

The first layer is start reaction. Reaction time off the blocks at elite level generally runs between 0.6 and 0.75 seconds in short events. A gap of 0.1 seconds here can equal months of training elsewhere. It is easy data to collect and compare, yet it is rarely published at domestic meets.

The second layer is split structure. A 200m race is not four identical 50m segments. How a swimmer distributes speed between the first and second half tells a story about fitness, tactics and even psychology. Negative splitting — a faster second half — signals a durable base. A large positive split signals an over-fast start. At many domestic meets we have final results but no saved splits. Without splits, there is no diagnosis.

The third layer is stroke rate and distance per stroke. The two trade off against each other. A higher stroke rate can lift instantaneous speed but cut efficiency; more distance per stroke saves energy but may cost peak speed. The optimum is not fixed — it depends on body length, arm span and event demands. This is where video analysis meets sensor data, and where a coach's instinct can be quantified rather than replaced.

The fourth layer is the turn and underwater technique. Since the rules allowed full exploitation of the underwater dolphin kick, the distance and number of kicks after the start and after each turn have become a battlefield of their own. In short events, a good turn can save several tenths of a second — more than a perfect stroke cycle. In events of 200m and up, the total time saved across seven turns can decide the placings.

The fifth layer, the most neglected, is physiological and training-load data. Heart rate, lactate, weekly volume — these numbers decide whether an athlete peaks at the right moment or collapses when it matters most. Without them, a training cycle is a gamble.

These five layers form a chain. Read the whole chain and the picture appears. Read one layer alone and you can fool yourself.

There is a sixth layer rarely mentioned: relay data. In the 4x100m and 4x200m, the start time of each leg, the takeover distance and the touch time of the swimmer coming in can all be measured. A strong relay is not just four strong individuals; it is a takeover timed to the tenth of a second. This demands high recording discipline, and it is also where smaller teams can gain an edge if they commit to analysis.

The difference between short-course (25m) and long-course (50m) pools is another important data variable. The same swimmer can post markedly different times in the two, because the number of turns and starts differs. A swimming nation that does not separate the two when recording will create flawed comparisons, especially in short events. When I compare Vietnamese athletes with regional rivals, the first step is to check the pool type and competition conditions before comparing any number.

When data tells a different story from instinct

At one SEA Games I followed, a young Vietnamese swimmer finished behind a regional rival by a margin the stands read as comprehensively outclassed. Splitting the race reversed half the picture: the Vietnamese swimmer won the middle and lost only the final 15m. The problem was not the aerobic base or stroke technique; it was the finish and the ability to hold stroke rate in a state of extreme fatigue.

Read only the final result and we draw the wrong conclusion and fix what does not need fixing. Read the splits and we know what to do: train the ability to hold stroke rate at the end, train the touch, train the psychology of the closing metres. A small change in the right place is worth more than a big overhaul in the wrong one.

This is where I recall a lesson from the xG model I once built for football. A team can win by luck, but a long enough data series pulls results back to where they belong. Swimming has less luck than football, but it is not perfectly linear either. A swimmer can explode in one race and vanish. A lane can look beautiful and be inefficient. The frightening part is this: the data is telling the truth, and we are not patient enough to hear the whole story.

I once treated the model as scripture. Now it is only a compass — but without it, we are lost.

The counter-view: import the method, not the whole model

There is a strong temptation in sports analysis: take a model from the swimming powers wholesale and apply it directly to Vietnam. That is the fastest route to beautiful reports and wrong conclusions.

Western models are built on different body types, nutrition, competition density and facilities. The lactate threshold of an athlete raised in that system cannot be compared directly with one raised in the Mekong Delta on an entirely different training and nutrition schedule. Copying the number while ignoring the context manufactures error.

The second danger is small samples. A swimmer who posts an unusual time at one domestic meet may be a one-off. With only a few dozen races a year, any statistical conclusion is fragile. So every number I cite comes with its sample size, and every prediction that runs past the data is labelled a hypothesis.

The third danger, subtler, is turning the sport's irrationality into a joke. Swimming has nights when everything logical fails: the best-trained athlete finishes behind, the underrated one sets a personal best. The worst response is to shrug and say that's sport. The right response is to treat that irrationality as a legitimate variable, to be built into the model rather than shut out of it.

There is one more trap rarely discussed: false causality. A team increases training sessions and wins a medal — that does not mean the extra sessions caused it. Perhaps the athlete simply reached maturity, or regional rivals declined. Data gives us correlation; causality needs a comparison design and time. Vietnamese swimming does not yet have enough years of data to claim causality with confidence, and the honest move is to say so rather than pretend certainty.

I have fallen into exactly that trap: I watched a swimmer with a beautiful stroke and concluded he would go far, only for later split data to show the physical base could not keep up with the technique.

What stands after a bad night

During the global football shutdown, I had a chance to study how systems behave when every old assumption is stripped away. When the stadium is empty, every model collapses. I rebuilt from the burnt data. That lesson applies to swimming: when a star is absent, when a training centre closes, the only thing that stands is a data structure recorded carefully beforehand.

A swimming nation that wants to last needs three things that can be copied: split records of every athlete over time, technical data (stroke rate, distance per stroke, turns) updated regularly, and training-load data tied to the competition cycle. None of these requires expensive technology. They require recording discipline.

I learned this from my years with a stopwatch. That 0.3-second error taught me that a number is only trustworthy when you know how it was measured, by whom, and when. Numbers do not lie, but people always find a way to lie with numbers.

Closing: signals for the next cycle

Vietnamese swimming is not short of talent. It is short of a recording system dense enough that talent is not wasted, and clean enough that decision-makers trust it more than they trust an impression. In the coming cycle, the signal I am watching is not a medal, but whether someone starts publishing the splits of each athlete regularly — even when the results are not pretty.

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