Vietnamese Swimming and the Data Void: When Every Analysis Starts Again From Zero

On the electronic board of the aquatics arena, the clock reads 15:34.21. Beside it are the athlete's name, the lane number, and a tiny "PB" line. No 100-metre split. No 200-metre split. No reaction time measured in hundredths of a second. No stroke rate per minute. No distance per stroke. No data on the final 25-metre surge. Only a single final number, standing alone like the full stop of a sentence whose beginning has been forgotten.

I sat in the seventh row, notebook open, and across four hours I did not record a single line that could go into a model. Not because I was lazy. Because there was nothing to record. Every domestic season, the same scene repeats: a race drifts past my eyes, and when I open the spreadsheet, I get an empty result. An empty result and a bad result belong to two entirely different categories. A bad result still tells me something. An empty result tells me nothing at all.

Numbers never lie, but they know how to hide. Here they are not hiding. They are simply absent.

I. The empty vault of Vietnamese swimming

In leading swimming nations, a 200-metre freestyle race is not a number. It is a dense cluster of data: reaction time off the blocks, 50-metre splits, stroke rate per minute, stroke count per lap, distance per stroke, turn time, underwater time after the start and after each turn, and the rate of speed decay over the final 50 metres. With that cluster, an analyst can reconstruct the entire race without watching the video.

In Vietnam, we have the final number. Occasionally we have "PB". That is all. And I will say it plainly: most of the swimming analysis readers encounter at every SEA Games or national championship is built on exactly such an empty foundation.

This is not a minor matter. It is the root of every distortion that follows.

I have a professional habit formed in 2026, when I was still in Nha Trang, teaching myself to build an expected-goals metric from video of the first 12 rounds of the V.League using a spreadsheet. That habit is simple: before making any claim, I must list how many data points I hold, where they come from, and whether they are thick enough to support a conclusion. Applying that habit to swimming, I find myself frequently stopping at the second step. Where does the data come from? Largely, it comes from nowhere.

1.1. Technique: no splits, no technique

Technical analysis in swimming requires four minimum data groups: start and underwater phase, turn segments, split structure, and stroke efficiency. None of these is publicly released in Vietnam.

Take a concrete example. A male swimmer racing the 1500-metre freestyle. On the international stage, analysts divide the race into 30 segments of 50 metres and track the standard deviation of each. If the first 20 segments vary within 0.8 seconds and the final 10 spike to a spread of about 3 seconds, that signals speed decay caused by poor pacing — a fixable technical problem. But with only the final number, we do not know whether the swimmer won by exploding early or by holding an even rhythm. We do not know where the sprint came from. We know nothing.

Worse, some technical elements never appear on a scoreboard. In breaststroke, the rules permit a single kick after each stroke cycle when surfacing. In freestyle, the 15-metre rule limits how long a swimmer may stay underwater after the start and after each turn. For elite swimmers, the margin of victory lies in maximising those 15 metres — and assessing that requires underwater-time data, which does not exist in the domestic dataset.

The result: every technical judgement I have ever read about Vietnamese swimming falls into phrases like "beautiful stroke", "impressive surge", "good rhythm". That is the language of feeling, not of technique. A technical analysis without splits is like a diagnosis delivered by a doctor who only looks at the patient through a window.

I once tried to apply the same approach I use when analysing free kicks or long-range shots in football — where I computed an expected-goal probability of just 0.03 for a strike from nearly 50 metres. With swimming, the equivalent calculation needs segment-by-segment speed data. Without it, every technical comparison between swimmers is merely a comparison of impressions.

1.2. Performance: an empty coordinate system

When I analyse a race, the first task is to place it on a coordinate system: against the world record, against the all-time list, against the current-season ranking. Those three coordinates tell me where a swimmer stands.

With Vietnamese swimming, I usually have only half a coordinate system. I can look up the world record — data from the global aquatics governing body is publicly available. I can look up the Asian record. But I lack data on the value of a performance within its specific context: long course or short course, water temperature, salinity, competitive pressure.

This is a point outsiders often miss. The same swimmer, over the same distance, records significantly different times in a 25-metre pool and a 50-metre pool, and short course is not automatically faster at every distance. Converting performance value between the two pool types requires a correction model, and that model needs a sufficiently large sample. In Vietnam, the number of times a swimmer competes internationally in a long-course pool each year can often be counted on one hand. The sample is too small to say anything statistically meaningful.

I once tried to build a prediction model for the women's 200-metre individual medley, and I had to stop halfway for lack of sample. A model built on four data points is not a model. It is a straight line drawn through a few raindrops.

There is one further detail often overlooked: when a swimmer breaks a national record, people compare it to the previous record. But the previous record may have been set more than a decade ago, under completely different conditions. Comparing two numbers ten years apart without adjusting for context is a form of coordinate error. It is like comparing the speed of two ships without accounting for the ocean current.

1.3. Competition system: an invisible cycle

Swimming has a clearly tiered competition system: the long-course world championships, short-course world championships, World Cup, regional multi-sport games, national championships, and junior meets. Each has a different function. Some are for sharpening, some for Olympic qualifying standards, some for peaking.

Looking at a Vietnamese swimmer's competition calendar, I cannot read that logic. The number of meets is small, the gaps between them are uneven, and information about the objective of each meet is almost never published. When a swimmer performs below expectation at a meet, I do not know whether it is because they are in a heavy training block, nursing a minor injury, or simply not yet peaking. Those three causes lead to three entirely different conclusions, but from the outside they look identical.

Within an Olympic qualifying cycle, the window for hitting A and B standards is decisive. Which meets a swimmer chooses to swim in that period, and which they skip, says a great deal about strategy. Without that information, any analysis of "form" is guesswork. This is the kind of information a transfer-market administrator like me always seeks first, because it determines the entire interpretation of the results that follow.

1.4. The world map: who dominates, and where we stand

At world level, swimming is divided into dominant clusters by distance and stroke. The leaders in men's sprint freestyle, the dominant group in women's butterfly, the strongest in individual medley. These clusters are not fixed; they shift with the Olympic cycle, and the shift usually begins in the youth development system, not with a few freak individuals.

For Vietnamese swimming, the story sits in a very narrow point: a few individuals capable of reaching Olympic qualifying standards in a few events. This is the "single-point breakthrough" model — dependent on a handful of key swimmers, not on a system that reliably produces talent.

This model carries enormous structural risk. When the key swimmer retires or is injured, there is no comparable successor to fill the gap. I have witnessed this across many sports, not only swimming. A system with only a peak and no body will collapse when the peak disappears. And even while the peak remains, it carries the overload of supporting the entire system.

Compare this with swimming nations that possess depth. They do not depend on any single individual. Every position on the national team has at least two contenders of similar strength. That creates not only internal competition but also a source of cross-checking data: when several swimmers reach the same performance level, we know that level is the genuine output of the system, not the luck of one person.

1.5. Rules and governance: an unchecked grey zone

Swimming is a sport with detailed rules and a strict anti-doping framework. Equipment regulations, eligibility, and testing procedures have all been standardised internationally.

Most of this governance content never appears in domestic swimming coverage. That means when an incident occurs — a technical dispute, a question of eligibility — readers lack the foundation to understand it. They receive only conclusions, never the process.

I do not mean to invite suspicion here. I mean only to say that a swimming system whose public cannot access basic governance documents will always exist in a grey zone. When transparency is lacking, rumour replaces data. And rumour, unlike data, cannot be verified.

For example, the doping testing process has specific steps: sample collection, storage, transport, analysis, and the right of appeal. A swimmer can be subject to out-of-competition testing. But the public usually learns only the final outcome, in the form of a verdict already delivered. No one explains the process, the thresholds, or the rights of the parties involved.

1.6. Athlete career: the age curve

Swimming is a sport where the performance curve is tightly bound to age and physical development. Female swimmers typically peak earlier than males, and puberty can create a temporary "wall" that flattens or reduces performance before recovery.

To position a swimmer on that curve, I need year-by-year performance history, training background, and injury information. With most Vietnamese swimmers, I have only a few scattered numbers, not enough to draw the curve.

This leads to a specific consequence: we tend to judge a young swimmer on a single race, not on a trend. One fast swim says nothing. The trend of ten swims over two years is the story.

There is a professional anecdote I have told many times, about a player I tracked and found had lost nearly 38% of his acceleration compared to the previous season, even while still scoring regularly. I warned that he would fade after the 70th minute and advised against extending his contract. When the league returned, exactly as predicted, he lost his starting place. The lesson is not that I am good at prophecy. The lesson is that when you have a data curve, you see the future; when you do not, you see only the present.

In swimming, that curve barely exists. Every season, we start from scratch, as though the previous season never happened.

1.7. Risk profile

With complete data, one can build a risk matrix: injury risk, form-decline risk, big-meet psychological risk, risk of being overtaken in heats. Each has its own probability and impact.

In Vietnam, most of these risks cannot be quantified. I have no public injury data. I have no data on performance at major meets. I have no data on competition load. When risk cannot be quantified, people tend to ignore it — until it happens.

In swimming, two characteristic injuries are swimmer's shoulder and breaststroker's knee. Both have early warning signs that can be measured: joint range of motion, cumulative training load, and changes in technique under fatigue. No one measures those indicators at national level. As a result, injury always appears as a sudden event, when in reality it was usually forecast long in advance.

1.8. The public narrative

At every major games, the media builds stories: the prodigy, the miracle, the golden dream. These stories have appeal and are useful to a degree, because they generate interest. But when they are not built on data, they create false expectations and lead to unnecessary disappointment.

I have seen this across many sports. A young swimmer has one good race, the media calls them a "prodigy", and two years later, when that swimmer fails to get out of the heats, the same media calls it a "failure". Both times were wrong, because neither was based on data.

There is a way to measure the mismatch between expectation and reality. It is called the expectation gap. When the media pushes a swimmer very high, the market and the public place emotional bets on a certain performance level. If that level is not reached, the disappointment is far greater than if nothing had been expected. The wider the gap, the heavier the loss.

I do not wish to dampen fans' enthusiasm. I want to say that enthusiasm can come with data. A swimmer can still be loved, but loved on the basis of understanding rather than illusion. That difference determines whether we can hold on to an athlete through the difficult phases.

1.9. Industry ripple effects

Finally, a strong swimming nation pulls along an entire value chain: the training market, the equipment sector, event business, agency systems, facility investment. But this chain only functions when data exists to direct capital.

Without data, the investment market gropes in the dark. Pools are built according to trends rather than real demand. Training programmes open according to instinct rather than skill gaps. People look at the valuation table; I look at the curve. Many investments in swimming die before they are ever announced, because no one checked the curve.

Conversely, when a country has good data, the value chain organises itself. Training centres know which skills they need to coach. Equipment makers know which segment they need to serve. Media outlets know which stories they need to tell. All of it starts with recording the first number.

II. The twist: an empty result is not a bad one

At this point I must challenge myself.

There is a strong temptation in this profession: when data is thin, an analyst is tempted to fill the gap with speculation. A little experience here, a little intuition there, and a conclusion is born. It sounds reasonable. It reads smoothly. But it has no basis.

I have been wrong in exactly that way. Years ago, I made a prediction about an athlete based on two matches. I was confident. I was persuasive. And I was wrong. Wrong not because my intuition was poor, but because I treated statistical noise as signal. That lesson shaped how I have worked ever since: set the significance threshold in advance, and verify afterwards.

Luck is something I do not have. I have probability and sufficiently thick data. When the data is not thick enough, the correct answer is not a bold prediction. The correct answer is an empty result — and saying plainly that no conclusion can yet be drawn.

This is the difference between an analyst and a prophet. The prophet always has an answer. The analyst has the right to say "not enough data". Swimming needs more analysts, and fewer prophets.

The second twist: an empty result, if published correctly, has diagnostic value. It points precisely to where the system is lacking. It turns silence into a blueprint for the next round of data collection. A pipeline that returns an empty result is not a failure. It is a signal that the layer above it needs rebuilding.

In many cases, an empty result is the most honest result we can offer. And an honest result, however lacking in excitement, has more lasting value than a wrong prediction that was cheered.

I know this is not easy to hear. An analysis system returning an empty result reads like a confession. But that is precisely why it is necessary. Football analysts learned this lesson when their models collapsed before unexpected results. Sometimes what kills a national team is not the opponent's magic, but the fact that they stopped moving, stopped pressing, and their metrics stopped connecting to one another. In swimming, the same happens at the data level: a system does not collapse in one night; it collapses when there are no longer any metrics to connect.

III. Recovering what is missing

So what is to be done? I harbour no illusion that one article can change the data infrastructure of a sport. But I know where to begin.

First, publish splits. An electronic timing system already has the capacity to record times at every wall touch. Publishing 50-metre splits for each race requires no new technology; it requires a decision.

Second, record reaction times off the blocks. This is the most basic metric in any international race, and it is almost never published at domestic meets.

Third, build a historical database by athlete. I did this on a small scale during the COVID shutdown, when I reopened the directory and built a multi-season dataset. No league is meaningless. Every result, however small, is a data point, and enough data points will draw a curve.

Fourth, publish information about the objective of each competition phase. When spectators know a meet is for sharpening rather than peaking, they will read the results differently. This is the least costly step and delivers the most value.

Fifth, and most importantly: accept that some questions currently have no answer. That is the condition for being able to answer them in the future. A sport that refuses to accept emptiness will always fill it with stories, and stories can never replace data.

My years of watching competition have taught me one thing: the most frightening thing is not a prediction model that is wrong. The most frightening thing is a model with nothing to predict. When the data foundation is empty, an analyst's talent cannot be brought to bear, and a swimmer's talent cannot be properly recognised.

Takeaway

The signal I am tracking in the next cycle is not a new record. It is how many splits are published at the next national championship. If that number rises from zero to anything above zero, Vietnamese swimming will have a foundation for genuine analysis. A national team does not collapse in one night; it collapses when its metrics stop connecting to one another. And conversely, it rises when the first metrics are recorded.