The Empty Dossier: Vietnamese Swimming and the Discipline of Saying “Not Enough Data”

11 p.m. in Hai Phong. The nine-layer analysis file opens on screen, every cell lit blue. I have just come back from the stands of a national swimming final, notebook full, ears still ringing with whistles and crowd noise. Twelve minutes later, all nine layers return the same line: not enough information to assess.

No split times. No water temperature, no pool depth, no subsurface current speed in the middle lanes. No official start list with a publication date. No note on which altitude camp the swimmer has just returned from, or how many events they have swum in four days. The results sheet hands me exactly one hard data line: the final time.

One line. The seventeen people in tomorrow morning's meeting can all talk about that line. Most of what they say sits outside it: character, hunger, “proven class”, a future about to open. I keep one professional vow to tie my own hands: numbers do not lie, but the people who read numbers do.

So I stay an extra hour instead of filing the story immediately.

Nine layers, and one layer of zero

The framework I use is built for a single purpose: to force every conclusion to point back at the data that produced it. The technical layer covers start and underwater phase, turn and finish technique, stroke efficiency, and venue adaptability. The performance layer places the swimmer on a coordinate system against the world record, the all-time list, and the current-season ranking. The competition-system layer asks where this meet sits in the four-year cycle and how entry is secured. The world-landscape layer maps who rules each event and where the talent supply chain flows from. The rules and anti-doping layer checks eligibility, equipment, and incident procedure. The athlete-career layer plots age, improvement slope, and the puberty threshold on a single chart. The risk layer scans injury, psychological load, and event volume. The public-narrative layer measures expectation heat against fundamental strength. The industry layer looks sideways at the coaching market, equipment, event business, and venue investment.

All nine run on one input, which I call Stage-1 data. Split times per 50 metres. A start list with a publication timestamp. Logged pool conditions. Separated heat, semi-final, and final results. Sourced injury status. A schedule with density.

At many domestic meets in Vietnam, Stage-1 data exists only in fragments. Results are published as final times, sometimes with a name and a club. Split times are almost never archived publicly. Short-course and long-course results appear mixed inside the same news line, so a national record can be compared with a mark set in a different system. Age-group records are logged without course notation, without a record date, without meet context.

Based on my experience following matches and regional multi-sport games, this gap is not the carelessness of one organising committee. It is the structural signature of a sport that has not yet built data infrastructure.

Split times: the biggest missing piece

Two swimmers finish the 1500m freestyle in the same 15 minutes 20 seconds. On the results sheet, they are one person. In the data, they are two entirely different athletes.

The first goes out in 7:35 for the opening 750m and returns in 7:45. The second goes out in 7:20 and comes home in 8:00. Same finish, two structures. The first has an energy-distribution engine inside controllable range; the second has a time bomb. Reading only the result line, a coach cannot tell which one is being developed.

In breaststroke the gap between structures is wider still. A swimmer can win through the underwater pull and the turn, or through stroke rate over the last 25 metres. Two routes to the same medal demand opposite training programmes.

The claim “weak finish” appears constantly in sports coverage. Most of those claims are written about the last 50 metres of a race nobody measured. The writer sees the order change in the closing metres and infers cause from effect. That is the worst form of sports inference, because it cannot be wrong in any testable way.

When only the final time exists, every technical judgement becomes a guess dressed in professional vocabulary.

My technical layer requires six indicators: progression across rounds, start and underwater phase, turn and finish, stroke efficiency, venue adaptability, and race-to-race stability. Without split times, all six return empty values. The dossier is not weak. The dossier is empty.

Fast pools, slow pools, and the illusion of progress

A national record broken on a night with no crowd, in a pool 2 metres deep instead of 3, with water warmer than standard, carries a different value from the same record broken in a full arena against comparable opposition. Flagging that difference does not dim the achievement; it places the achievement at the correct coordinate so the next comparison is possible.

In 2026, when the global sporting calendar stopped, I assembled 3,487 Bundesliga matches from 2026 to 2026 and compared them with 412 matches played without spectators after the league resumed. Home advantage fell 42 percent, from an average of 0.48 goals per match to 0.28. The result does not say football got worse. It says part of what we call home advantage comes from the stands, not from the grass.

In swimming, the equivalent variable is named crowd and water. A young swimmer racing a national final in front of a full arena produces a different race from the same swimmer on a quiet morning. When the world stopped turning, I built my own rotation of data; and that rotation taught me that most “breakthroughs” are an undeclared environmental variable.

Medals are not performance data

This is where composure is easiest to lose.

A gold medal is a fact of the competition system, not data about swimming ability. Two athletes on the same podium can be ten seconds apart. An event with four finalists differs in kind from one with eight finalists and three who have reached a continental final.

I call this adjustment the result-interpretation discount. It does not deny the achievement. It answers a question: if that same time were placed in a deeper, stronger field, where would it sit on the regression line?

I learned this lesson by nearly getting it wrong. At the 2026 World Cup, my editors urged me to write that Russia got lucky against Spain. A PPDA of 8.7 showed Russia actively pushing the opponent wide and closing the central lane. Russia's expected goals conceded reached 2.9, and their goalkeeper saved six shots. I refused to change the headline to the word “miracle”, and the piece drew heavy disagreement.

Four years later, before the 2026 World Cup, my model predicted Morocco reaching the semi-finals on a PPDA of 6.9 and transition speed among the fastest in the tournament. I said on air that Morocco was not a fairytale but a complete defensive system, and was called a dreamer in a data room. When Morocco did reach the semi-finals, my analysis video passed 4.5 million views.

Both times, I did nothing except refuse to call a structure a wonder. A miracle is only a data point that has not been regressed. Every shock already has a portrait inside older data.

In Vietnamese swimming, the result-interpretation discount is almost never applied. Each SEA Games cycle, the medal count is used as the yardstick for the sport's quality. That number measures delegation strength inside one specific competition system. It does not measure the average 50m speed of the youngest swimmers in the squad.

Shoulders, knees, and the puberty threshold

The athlete-career layer is the most misread layer in swimming, because it demands patience.

A 14-year-old breaking an age-group record usually produces a steep improvement curve. A linear model looks at that curve and extrapolates. The human body does not improve linearly. After puberty the slope changes direction, fat ratio and muscle mass shift, arm span changes, and the training programme has to be rewritten. This is where talent-valuation models tend to overstate young potential while understating unmeasurable variables such as team chemistry and the quality of the training environment.

Injury risk also sits in a grey zone. The shoulder for freestyle and butterfly, the knee for breaststroke, the back for heavy backstroke workloads — all carry long histories. But when a swimmer misses a meet, published information usually offers two options: injured or absent. There is no middle column for overload, recovery, or a preservation decision.

The locker room, a variable models cannot read

This is where I have to argue against my own profession.

Valuation models read age, height, performance slope, and improvement frequency very well. They read the stability of the training environment very poorly. A swimmer changing training base may change coach, training group, time zone, and nutrition regime at once. The model records the jump in race times and attributes it to talent. The true baseline of that jump lives in the locker room.

Another rarely written variable is agent noise. In the football transfer market, agents are the largest hidden cost, and the noise they generate distorts price. In swimming, intermediaries appear in different form: negotiating international meet slots, personal sponsorship deals, scholarship placements abroad. Every time a young swimmer changes sporting nationality or training base, domestic coverage frames it as an emotional event. In the data, it is a resource-migration signal, and it is often forecastable several months before the announcement.

Expectation heat and fundamental strength

The public-narrative layer is the most ignored layer, and also the most forecastable.

The mechanism is simple. I measure social expectation heat around a swimmer, then place it beside that swimmer's own data foundation. The ratio between the two is an overheating indicator. A rising swimmer can carry an expectation index five times their foundation, and the market will correct itself, usually at the least convenient moment for the athlete.

In Vietnam, the heat cycle is tied to the SEA Games. Every two years a new figure is pushed forward, the story is retold, expectation is charged, and the following two years are silence. For figures who have crossed multiple cycles, such as Nguyen Thi Anh Vien, who announced her retirement in 2026 after more than a decade at regional elite level, the public narrative rests on a thick data foundation. For the next generation, the foundation is far thinner.

Nguyen Huy Hoang won silver in the men's 800m freestyle at the 19th Asian Games held in Hangzhou in September 2026, according to the organisers' official results. That is a hard fact: a name, a date, an event, a competition. This type of fact is the most reliable signal Vietnamese swimming currently produces. It is still not enough to run the technical layer, because the technical layer needs split times, and the split times of that race only become useful when placed beside the same swimmer's splits in three previous races.

The ripple effect

Vietnamese swimming is expanding fast at market level: private pools, learn-to-swim classes, equipment, mass-participation races. Investment heat at this level runs far above investment heat in data infrastructure.

That gap has consequences. A coaching market that sells to parents through medals will naturally prefer the trophy photo over the split-time sheet. An event business that sells tickets through stars has no incentive to publish start lists with pool data. The result is a sport that generates a great deal of imagery and very little reusable data.

Where I can be wrong

A dossier of empty values can easily become a pose of moral superiority.

I need to say this plainly. Declaring “not enough data” in every case is laziness wearing the uniform of rigour. Missing data is missing data; it is not evidence of anything, including evidence that a sport is weak. An analyst can use the silence of data to build personal authority, and that is the trap I set for myself every time I sit down.

The principle that correlation does not imply causation also bites back at the writer. A swimmer improving by two seconds after a meet may be due to taper timing, a faster pool, absent rivals, a growth phase, or one technical change in the underwater phase. Writing that the swimmer “transformed technically” assigns a cause to an effect we have not yet isolated.

I also have to admit I am the person bringing data into the room and the person who wants the room to remember my name. An analyst's instinct is to be right. A better instinct is for the dossier to be right. Those are different things, and only one of them survives ten years.

What to watch in the next round

I do not believe in luck; I believe in the margin of error. And the margin of error in Vietnamese swimming narrows the moment three things are published consistently: 50m split times in heats and finals; start lists with publication dates and 25m or 50m course notation; and a pool-condition log for every session.

None of that requires a large budget. It requires one decision.

Data only dies when we stop asking questions. The question I leave for next season is concrete: if one national meet published full split times, how many swimmers would we discover are better than the medal table ever said, and how many things we believed were turning points would turn out to be one evening in a fast pool?