Vietnamese Swimming: The Blank at the Data-Input Stage

On the electronic board of an indoor pool in Saigon, the men's 200m individual medley final has just ended. The board shows exactly three things: name, lane, touch time. No reaction time off the blocks. No 50m splits. No stroke rate. I open the organisers' technical file to the section where metrics are recorded, and that section is entirely blank, from the first line to the last. A real race, eight real lanes, and not a single unit of data retained.

For someone whose job is reading numbers, this is a measurable kind of loss. We know who touched first. We do not know whether that swimmer won it in the first 50 metres or the last 15, whether the third turn was fast or slow, whether the stroke rate held from the 150m mark into the 200m or collapsed. The result survives. The process disappears.

Every shock has its own probability. We call it a shock only because we have not checked the table yet. In Vietnamese swimming, the table is empty from the very first column.

The data point: the smallest unit of a verifiable fact

The analytical process I use has two layers. Layer one extracts what I call a data point — an atomic fact, too small to break down further, verifiable through a photograph of the scoreboard, a referee's report or an official result file. Layer two takes those points as its foundation and analyses them. Without layer one, layer two has only two options: stay silent, or invent.

In a swimming race, the minimum data set has six groups. Reaction time off the block, measured in hundredths of a second, tells you how long a swimmer took to leave the platform. Four to eight splits at 50m intervals tell you how speed was distributed along the race. Stroke rate is the number of arm cycles per minute. Distance per stroke is how far the body travels with each cycle. Turn time covers the touch and the push-off. And underwater distance covers the glide after the start and after every turn.

These six groups exist for a concrete mechanical reason. A higher stroke rate does not automatically produce a faster time; it only helps when distance per stroke is preserved, because propulsion is generated in the catch phase. A swimmer who lifts from 38 to 44 cycles per minute while distance per stroke falls from 2.1 metres to 1.8 metres is moving the arms faster and travelling slower. Given only a final time, those two ways of swimming look identical.

The world standard and the home standard

At events run by World Aquatics — the new name for FINA since December 2026 — every final is published with 50m splits and reaction times; many events add stroke rate and distance per stroke. The 2026 World Aquatics Championships in Singapore placed all of that data on an open results system, downloadable in a couple of clicks.

At domestic meets, the standard stops at the final time. I have gone back through published results from many domestic competitions and recent SEA Games: mostly placing and time, occasionally a note about a record. Race structure does not appear.

When I reviewed the published results of the SEA Games held in Thailand in late 2026, I still could not find splits for most finals. Regional teams compete under the same form, submit into the same system, yet the process data remains blank on the receiving side.

The gap lies somewhere else, not in the equipment. Electronic timing systems in certified pools can output splits, because the touch pads at each end send separate signals to the processor. The problem is that nobody requires splits to be recorded, stored and published. The hardware has been there for years. The habit of recording has not.

Reading a race with numbers

An ordinary spectator asks immediately: did this swimmer win because they swam faster, or because the opponent slowed at the end? A results table cannot answer that. 50m splits can, and the answer usually sits in the second half of the race.

Take an illustrative example, not data from any actual meet: a swimmer finishes the men's 200m individual medley in 2 minutes 14 seconds, with 28.2 in the butterfly leg, 33.6 in backstroke, 38.9 in breaststroke and 33.3 in freestyle. The runner-up swims 32.8 in freestyle but 37.5 in breaststroke. On the results sheet, the two are nearly indistinguishable. On the splits, the runner-up swam the best freestyle leg of the field and lost the race in breaststroke. A coach who can read that immediately knows which muscle group needs extra work, and more importantly, knows the athlete's baseline speed is already enough to win.

The chain of evidence

Nguyễn Thị Ánh Viên won silver in the women's 400m individual medley at the 2026 Asian Games in Incheon, a fact recorded in the official result files of the Olympic Council of Asia. For a data analyst, the value lies elsewhere: we can still dissect that race years later, because every 100m split survives. That is how we can see she was stronger in butterfly than in breaststroke and made it up in freestyle.

Nguyễn Huy Hoàng, born in 2026, won the men's 1,500m freestyle at the 30th SEA Games in the Philippines in 2026. Trần Hưng Nguyên won the men's 400m individual medley at the 31st SEA Games on home soil in Hanoi in 2026. Two gold medals, two long-distance events, two different speed distributions. In both cases, the data reaching domestic specialists remained a results table.

The accumulated consequence is the expensive part. A swimmer trains four years with one coach, then moves up to the national team. The new coach has no dataset on how that athlete distributes speed over 200m, what their baseline stroke rate is, or how much slower the fourth turn is than the third. Everything is measured from scratch. In swimming nations with open data repositories, that is inherited assets. Here, the assets are deleted every time the personnel change.

At provincial pools, a 14-year-old with an underwater phase longer than the age-group standard and a strong distance per stroke, but without a fast final time, will not appear on any selection list. Nobody chooses wrongly. Nobody has anything to choose from.

A training era dies when nobody reads its data tables any more.

When the stands go quiet

2026 left behind a natural experiment I still use when talking with coaches. Pools closed, stands silent, meets postponed. When competition returned with limited crowds, home-pool advantage almost vanished. When the stands go quiet, home advantage dissolves into a number near zero. The variables still separating swimmers were individual training load and recovery — precisely the two things the domestic data set does not measure.

During that period I reviewed load data for a group of athletes and found high-intensity running volume rising before soft-tissue injuries appeared. The conclusion is not that training more causes injury. A sudden jump between two consecutive weeks is the warning sign. To see the warning sign, you need weekly data. Without weekly data, the warning sign only shows up as an injury that has already happened.

The counter-intuitive angle

The easiest thing to misread here is a reverse trap. Many will conclude that Vietnamese swimming needs to buy a data platform. My experience in other sports says the opposite: platforms do not create readers. I have seen beautiful dashboards, fully coloured, refreshing every second, sitting idle on a department computer because no coach was ever taught how to ask them a question.

The second trap is subtler: mistaking correlation for causation. In swimming data, two series often move together. A lower stroke rate usually accompanies a better long-course time. The cause is not the reduction in stroke rate. The cause is that the swimmer held propulsion per cycle while the body was tiring, and therefore had to lower the rate to preserve distance per stroke. A lower rate is a symptom of a strong aerobic base. To tell the difference, you need stroke rate and distance per stroke at the same time. With only one of the two, you will send an athlete in the wrong direction for an entire training cycle.

Here I have to state my own limits clearly. When I tried running a nine-dimension deep analysis on a recent swimming file — technique, performance, competition system, world landscape, rules and anti-doping, athlete pathway, risk profile, media narrative, industry ripple — all nine returned the same conclusion: insufficient information to assess. That is the correct answer. A data analyst has to be able to say "I don't know" in front of an empty table. Anything else is signing a cheque on an account with no money in it.

The shot happens once. Its trajectory lasts for years. In swimming, a race happens once, and its trajectory lasts exactly as many years as its data is kept.

Signals for the next cycle

I am not proposing a large system. I am proposing a minimum standard, small enough for a provincial meet to adopt: reaction time, splits every 50m, stroke rate recorded at each 50m, and turn time for every turn. Four data fields, saved with the result file, published openly as a machine-readable table. The cost of those four fields is thirty minutes of one person's time after a session, not an equipment budget.

If that standard exists, three things happen on their own without any master plan. Coaches start comparing an athlete with that same athlete at the previous meet. Outside analysts gain a foundation to verify or reject each other's claims, instead of arguing by feel. And swimmers at provincial pools become visible through race structure, not only through placing.

I sit far from the pool deck in order to see the race more clearly than the timekeeper does. But even the person sitting furthest away still needs a file to read. What I leave for next season is not a prediction. It is a question: which domestic swimming meet will be the first to publish 50m splits, and who will it publish them for?