Vietnamese Swimming and the Discipline of Empty Data

Tuesday night, nine tabs, and the decision to write nothing

11:40 on a Tuesday night. I reopen the nine-dimension analysis sheet — the skeleton I built and have used for every deep swimming dossier for years. Nine tabs. Tab one is technique. Tab two is performance and data. Then the competition system and the qualification mechanism. Then the world swimming landscape and the event map. Rules and anti-doping governance. The athlete's career path. The risk profile. The public narrative and expectations. Finally, the ripple effects across the industry.

In every cell where a number should sit, I type the same line: "insufficient information to assess."

Not out of laziness, and not because I ran out of time. Because the input data is empty. No original title, no source name, no information points. No athlete, no meet, no coach. No splits, no times, no dates, no pool context.

The easiest thing to do would have been to open a results page, pick a few familiar names, bolt them onto a plausible story and call it analysis. Filling blanks with guesswork is always the fastest way to produce something readable. It is also the fastest way to destroy the only thing that brings readers back: the belief that the numbers in the piece are real.

So I left all nine tabs blank.

This article is not about a swimmer. It is about a professional decision: when a deep analysis must return a null result, and why in swimming that happens far more often than people assume.

Swimming, the sport of numbers nobody records

Football has hundreds of cameras, event-by-event data providers, and a three-source cross-check network for every pass. Swimming is different by nature. A 200-metre race lasts about two minutes, mostly underwater, and the things that decide the outcome sit where the naked eye cannot read: entry angle, dive depth, the count of underwater dolphin kicks, the instant of the wall touch and the push-off.

What the mainstream media sees is only the final number on the scoreboard. Yet that number is the most misleading thing of all, because it is the sum of many small decisions, and it tells no one which decision was wrong.

To read a swim, an analyst needs at least four layers of data. The first is splits — the time for each 50-metre segment. The second is technical indices: stroke rate, distance per stroke, reaction time off the block. The third is competitive context: a 50-metre long-course pool or a 25-metre short-course pool, indoor or outdoor, which round of the competition. The fourth is historical comparison: where this result sits relative to the same swimmer three months ago, a year ago.

Miss one of the four layers and the conclusion skews. Miss all four and there is no conclusion at all — only prose.

This is why I always tell young editors that a decent swimming analysis starts at the humblest point: check what you actually have in hand before you open your mouth. Based on my experience tracking swims since my years as a swimming reporter at a newsroom in southern Vietnam, one rule has never changed. No splits — do not talk about tactics. No reaction time — do not praise the start. No idea whether it was a long or short course — do not compare records. Nothing at all — do not write.

When the pool is empty, every model collapses. I rebuild from the burnt remains of data. But there is a kind of wreckage where even the builder must stand still: ash with no fragments left to pick up.

Nine layers of questions and the price of an empty cell

This is the technical core. I walk through each layer of the framework, spelling out what data it needs, why it needs it, and what happens to the quality of the conclusion when that data does not exist. This is how a null result becomes information instead of an apology.

Layer one — technique: where everything is decided under the surface

In swimming, technique is not a side note to performance. It is performance. A 100-metre freestyle race splits into four blocks: the start and dive, the first breakout, the turns in between, and the finish. Each block has its own set of indices.

The start block needs reaction time, measured from the signal to the feet leaving the block. At elite level this usually falls between six and seven tenths of a second. That is a very narrow band, yet it contains an entire coaching school. A swimmer leaving the block two tenths slower than six months ago is not simply weaker; it can be a signal of a changed foot placement, or of lost confidence after a shoulder injury.

The dive and the underwater phase are governed by the fifteen-metre rule. It states that after the start or a turn, the swimmer's head must surface before the fifteen-metre mark. The rule exists because the underwater dolphin kick is faster than any arm stroke. An entire modern training culture revolves around one question: how to hold the longest possible underwater glide within fifteen metres without being flagged by officials.

Without data on the breakout distance, an analyst can say nothing about start technique. Without underwater kick data, nothing about turn efficiency. And at that point, every compliment paid to a beautiful swim is just a feeling.

The turn block needs touch time and push-off time. In a 50-metre pool, a 200-metre race has only three turns. In a 25-metre pool, the same distance has seven. This is why short-course and long-course records must never sit side by side on one table without a note. A swimmer with strong turns gains enormously in short course, and a ranking that mixes both pool types is a meaningless ranking.

The finish block needs second-half pacing data. In the trade we speak of a "negative split" when the back half is faster than the front half. That is a sign of a swimmer in control of rhythm, or of someone who entered the race so cautiously that an opportunity slipped away. Telling the two apart requires 50-metre splits and round context. No splits, no distinction.

Layer two — performance and data: placing a number in historical space

A swim time on its own says nothing. It only means something when placed against three coordinates: the world record, the all-time list, and the current season ranking.

The first coordinate is the world record. But the world record in swimming carries a historical problem readers often overlook: the high-tech swimsuit era. Around the late 2000s, polyurethane suits let swimmers float higher and cut drag to an absurd degree, producing a wave of world records in a very short time. The global governing body later banned the suits, but the records set in that period remain on the books. It means that when comparing today's result with a record from that era, the analyst has a duty to state the context.

This is where honesty about data stops being a technical matter and becomes a professional ethic. Numbers do not lie, but people always find ways to lie with numbers.

The second coordinate is the all-time list. The third is the season ranking. Both require something Vietnamese swimming often lacks: standardised, continuously updated data. If a result is not logged in a structured database, it lives in the memory of a few people and vanishes when those people leave.

On qualification, people speak of A-cuts and B-cuts. An A-cut grants a swimmer a direct entry to a major meet; a B-cut depends on quota allocation by the federation. Yet even with a cut, the story is not over: the slot must pass through each country's selection mechanism. The United States runs its own trials with a top-two rule per event. Australia has its own trials system. Vietnam takes a different route, usually based on comprehensive evaluation and regional results.

To analyse a swimmer's qualification chances you need all three layers: personal performance, the meet's standards, and the national selection mechanism. Miss one and the conclusion errs by degrees. Miss all three and the question of whether that swimmer qualifies cannot be answered with data — only with rumour.

Layer three — competition system and participation mechanism

Swimming has a clear rhythm: the four-year Olympic cycle. Each year in that cycle plays a different role. The first is the foundation year. The second tests competitive volume. The third belongs to the world championships. The fourth is the Olympic year, when everything is pushed to the limit.

Results in each year must be read with a different coefficient. A poor result in year one of the cycle is normal. A poor result in an Olympic year is serious. If an analyst does not know where a meet sits in the cycle, they will apply the wrong coefficient to a correct result.

Then there is the schedule. Elite swimming lets one athlete race multiple events at the same meet, sometimes on the same day. That creates a distinctive risk: a packed schedule forces a choice between a signature event and one with medal potential. That choice is rarely announced, but its traces sit in the splits — a cautious heat swim, a final with a completely changed tactic.

Without a detailed schedule and splits, people will call a tactical decision a loss of form. This is the most common error in swimming analysis in the media.

Layer four — the world landscape and the event map

World swimming is organised into event groups: freestyle, backstroke, breaststroke, butterfly, individual medley, and relays. Each group has its own structure of dominance, and that structure changes slowly.

Reading a dominance map by stroke involves four tiers: the dominant tier, the challenger tier, the chasing tier, and the potential tier. One nation can dominate middle-distance freestyle while winning nothing in breaststroke. Another may have no outstanding individual yet be very strong in relays thanks to squad depth.

Squad depth is the concept I consider most underrated in regional swimming analysis. A team with one star but no credible second and third swimmer will always lose to a team with no star but six athletes of solid standard. At analysis level this can be tracked through squad-depth datasets, such as the player depth index regional analysts use to rank delegations.

To draw this map you need results from at least three consecutive seasons, event-by-event data, and information on junior cohorts. Without those three layers, any map is a rushed snapshot.

Layer five — rules and anti-doping governance

This is the most sensitive layer, and the one where carelessness does the most damage. In swimming, governance spans the world federation, the international anti-doping agency, the Olympic committee, and national bodies. Each has its own rulebook, and those rulebooks overlap in ways not always clear to outsiders.

There are four basic checks: anti-doping rules, competition and officiating rules, equipment rules, and eligibility. For each, the analyst must separate confirmed fact from allegation from speculation.

One principle I set myself years ago: in any doping-related story, without an official document from the competent authority, the piece may only be written as process, never as conclusion. Write process and no one is wronged. Write conclusion and someone might be.

And of course, when the input references no incident at all, this layer must stay blank too. No facts, no sanction simulation, no classification, no scenarios. This is the silence the profession demands, not the silence of fear.

Layer six — the athlete's career path

This is where swimming differs most sharply from football. Footballers typically peak in their mid-twenties. Swimmers, especially women, often peak far earlier.

In the technical literature this stage is called the puberty barrier. During puberty the body changes in ways that disrupt the ratio between strength and weight, between height and arm span, between fat and muscle. A female swimmer who was extremely fast at thirteen may stall at fifteen — not from laziness, but because her body is changing faster than her technique.

This is one of the stages worst handled by the media. A few poor results in this window are routinely turned into a story of decline, when it is simply a predictable physiological phase.

To assess it properly an analyst needs age, sex, height, arm span, and above all the performance curve over time. That curve must span at least three years. With it, one can distinguish a swimmer in transition from one who has hit a ceiling.

Alongside the puberty barrier, this layer carries three other risk groups. First, swimmer's shoulder — the most common injury in freestyle and butterfly, from the enormous stroke volume accumulated over years. Second, breaststroker's knee, from the specific kick. Third, big-meet psychology, which data cannot measure yet shows up very clearly in splits.

Vietnam has notable cases here. Nguyen Thi Anh Vien is a rare example of Vietnamese swimming with a development path recorded relatively continuously across multiple regional Games. Nguyen Huy Hoang is another case in the distance freestyle group. Tran Duy Khoi was once a prospect in the individual medley, while Pham Thanh Bao represents para-swimmers, where data is even thinner. Yet even for the most prominent names, our data remains far short of what deep analysis requires.

Layer seven — the risk profile

Risk in swimming falls into six groups. Competitive risk covers opponents, schedule and event selection. Career risk covers injury, the puberty barrier, and a short peak window. Doping risk covers violations and strict liability. Rules risk covers the chance of an officiating call, especially at the fifteen-metre underwater mark. Psychological and public-opinion risk covers expectation pressure. Systemic risk covers funding shortfalls, a lack of similarly matched rivals, and inadequate facilities.

The distinctive feature of Vietnamese swimming sits in that last group. A swimmer in a country with hundreds of equally matched training partners improves in a daily competitive environment. A swimmer where only a handful train at the same level improves more slowly — not for lack of talent, but for lack of rivals.

That is a systemic risk individual data cannot capture, yet it shapes every other number. Assessing it requires a picture of how many swimmers exist in each age group, each event, each locality. We usually do not have that picture.

Layer eight — public narrative and expectations

Sports media runs on labels. A young swimmer who breaks a national record is labelled a prodigy. An older swimmer returning is labelled the king's return. A night with several records is called a record night. A plateau is called decline.

Each label has a life cycle. First excitement. Then expectation. Then disappointment when expectation is unmet. Finally a new label, usually tragedy.

The data analyst's job is to ask: how much data is this label built on? If a swimmer is called a prodigy after one regional meet, where does that result sit in the world ranking for the same age group? If a label rests on three races, can it survive ten more?

In Vietnamese swimming the gap between public expectation and the underlying data is often wide. A regional medal is covered with the frequency of a global feat. That creates pressure on the athlete and a false reference frame for readers.

I do not think we should diminish regional achievements. I think we should state clearly where they sit on the world map, so that when the athlete steps onto a bigger stage, fans do not feel deceived.

Layer nine — ripple effects across the industry

Swimming extends well beyond elite athletes. It is a value chain: upstream is the youth coaching market and the talent supply; in the middle are athletes and meets; downstream are broadcasting, sponsorship, equipment, and derivative markets.

A decision upstream can ripple a long way. If a locality invests in pools and a mass learn-to-swim programme, the effect surfaces at elite level roughly eight to twelve years later. If a country cuts funding for youth training centres, the consequence does not appear immediately — it appears exactly when a generation should have entered its peak.

This delayed causality makes swimming a sport where today's decision can only be judged a decade on. It is also why analysing the swimming industry is harder than analysing a single race. You cannot replay the video of a policy decision.

In Vietnam this value chain has one very clear break point: data. Without standardised data on how many children can swim, how many pools operate regularly, how many coaches hold certification, every impact calculation is guesswork. The figures routinely quoted in the media on swimming literacy mostly come from small surveys, insufficient to build a model.

A null result is a product, not a failure

Now the part where many in the trade will object to me.

In sports data there is a quiet but powerful pressure: always deliver a result. Editors need copy. Readers need answers. Sponsors need numbers. In such an environment, an analyst who returns a null result is seen as someone who cannot do the job.

I think that is one of the most damaging misconceptions in the profession.

A null result does not mean ignorance. It is the output of a complete process in which the input failed the minimum conditions for a conclusion. If I run a blood test and the sample is contaminated, the professionally correct result is "sample invalid", not a diagnosis. Sports data analysis is the same.

But there is an important nuance. A null result does not mean sitting still. It means shifting from the question "what is the result" to "what is needed to obtain a result". That is a very concrete move, and it creates value.

When I left nine tabs blank, I did not end the work. I began it. I mapped the cells that could not be read, and for each one noted the type of data required, which source could provide it, and the maximum confidence that data could carry. That map, in turn, became a far more useful product than a wrong analysis.

This is the difference between a writer and an analyst. The writer needs words. The analyst needs to be right.

There is another temptation worth naming: turning correlation into causation. In swimming this error appears constantly. A swimmer changes coach and swims faster; the conclusion drawn is that the new coach is brilliant, when the real cause may be that the swimmer passed the puberty barrier at the same time. A national team trains abroad and wins a medal; the conclusion drawn is that overseas camps work, when the real cause may be that this particular cohort happened to be stronger.

Separating correlation from causation requires one of two things: a control group, or a long enough time series to rule out other variables. Vietnamese swimming rarely has either. That is why I always write in the conditional, use the word "hypothesis" when three sources have not been cross-checked, and state the sample size inside the piece. An anomalous figure from a sample of three races is a question, not a conclusion.

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

And with that compass, one rule may not be broken: when the needle points nowhere, do not draw a direction yourself and call it north.

Signals to track

Three things I will watch over the next twelve months, speaking of Vietnamese swimming.

First, the emergence of a structured swimming database. It does not need to be big. It only needs to log a result with distance, stroke, pool type, competition date, and splits where available. With that foundation, every analytical layer above becomes automatically more feasible. This is the most important signal because it is the precondition for everything else.

Second, how the media handles young athletes' plateaus. If articles begin referring to the puberty barrier instead of calling it decline, that is a sign the baseline of understanding is shifting. I will read the language closely, not just the results.

Third, the appearance of squad-depth datasets at regional level. When a nation starts being judged by the number of athletes in the chasing tier rather than only by gold medals, the way we see regional swimming will change.

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

As for that Tuesday night, I closed the sheet, left all nine tabs blank, and wrote one note to myself: if someone someday asks why this analysis has no conclusion, the answer is not that I did not know. The answer is that I knew exactly what was missing — and that, sometimes, is already half the answer.