Vietnamese Swimming and the Missing Data Layer: Consequences That Outlast a Season

On the electronic scoreboard of a domestic swimming meet in Vietnam, the only line of data that survives is the final time. No reaction time, no 15-metre split, no turn time, no closing speed. A 0.3-second gap between two medals — a margin that at world championships can be decomposed into four separate causes — gets compressed into a single line of text: name, club, result.

I started timing races by hand at the pool wall in 2026, when I was a swimming reporter for Thanh Nien. More than twenty years later, I sit in a room in Saigon building data models for football clubs, and I keep returning to the same question: if we only measure the final result, are we reading a result, or are we guessing at a cause?

At the international level, swimming data crossed that threshold long ago. World Aquatics publishes world championship results with reaction times — the interval from the starting signal to the moment the feet leave the block, captured automatically at 0.01-second resolution — alongside 50-metre splits for every event from 100 metres upward. Starting blocks and touchpads at major aquatic centres carry sensors, and the data reaches the coaching staff's laptops minutes after the hand hits the wall.

In Vietnam, that resolution appears unevenly. A few pools built to international competition standard have automatic timing, but most youth meets, regional meets and internal time trials still run on hand-held stopwatches and phone cameras. Installing, calibrating and operating a compliant sensor system costs thousands of US dollars per event. For a two-day meet with a few hundred swims, that line item is usually the first to be cut from the budget, after prize money and after officiating fees.

The consequence is not administrative. It is a performance picture shot through a very coarse lens. Nguyen Huy Hoang races the 1500-metre freestyle on the Olympic stage, where effort distribution across thirty laps decides everything, yet without published splits an analyst can only speculate whether he lost time in the opening laps or the final four hundred metres. Vo Thi My Tien competes in the 100-metre butterfly, an event where the last fifteen metres usually separate a finalist from a ninth-place finisher. Tran Hung Nguyen, who won gold in the 400-metre individual medley at the 2026 SEA Games in Hanoi, swims an event with eight clearly delineated segments — and in domestic archives those eight segments usually collapse into one number. Nguyen Thi Anh Vien left elite competition and with her went a gap in medals, and a data gap that has never been filled.

If I had to list what is missing, I would start with the start itself. Among elite international swimmers, reaction time sits between roughly 0.60 and 0.75 seconds; the spread among junior swimmers is far wider. In a 50-metre event, where the whole race lasts under half a minute, one tenth of a second at the block equals a visible distance on the water. Without measuring reaction, a coaching staff cannot tell whether an athlete lost because of a slow start or because of a low peak speed. Those two causes demand two completely different training plans, and picking the wrong plan costs a young swimmer at least one season.

The second layer is the underwater phase. The rules allow each swimmer up to 15 metres underwater after the start and after every turn. The dolphin kick is faster than surface freestyle, but it burns oxygen faster, so the optimal exit point depends on the event, on the day's condition and on who is in the next lane. Without a 15-metre split, the instruction to "stay under longer" is advice nobody can verify. I once watched a swimmer receive exactly that advice for two full seasons while nobody knew he had already been using the entire legal allowance from his very first trial.

The third layer is turns and finish. In a 25-metre short-course pool, a 100-metre race contains three turns; each turn plus the following underwater phase accounts for a substantial share of race time. A swimmer losing 0.15 seconds per turn loses 0.45 seconds across the event — roughly the gap between gold and bronze in many regional finals. The closing metres work the same way: stroke rate over the last five metres, head position and touch timing are three variables that can be fixed within four weeks of training, but only if they are recorded.

The fourth layer, and the most labour-intensive, is the velocity model. Swimming speed equals stroke rate multiplied by distance per stroke; two swimmers can produce the same time from two entirely different configurations, and those two configurations age at very different rates. A junior who swims on high stroke rate will dominate sprints and collapse over distance; one who swims on long distance per stroke tends to progress more slowly but has a higher ceiling. Counting stroke rate and measuring distance per stroke needs only an overhead camera and frame-tagging software — far cheaper than a touchpad system — yet it barely appears in the training files of smaller centres.

The consequence compounds across seasons, and I see it through a comparison from another sport. In 2026, analysing Vietnam's U20 team at the U20 World Cup in South Korea, I had expected-goals data for every attacking move, and with it I could show that the team's problem lay in conversion rather than in chance creation. Football already had a public data layer thick enough for an analyst outside the dressing room to ask the right question. Vietnamese swimming does not yet have that layer, so every argument about a performance ends where instinct begins. What cannot be measured cannot be fixed, and in swimming most of the margin sits in stretches the human eye cannot catch: the reaction, the underwater phase, and the turns.

This is where the counter-intuitive angle sits, and I want to state it plainly because it argues against my own trade. Having data does not automatically produce faster swimmers. Between two series that move together, I only claim causation when I can point to the mechanism that links them: data changes outcomes only when it returns to the coaching staff within forty-eight hours, in the form of a concrete adjustment to next week's plan. A split published and then filed away is a dead split. I have seen enough cases of data specialists walking into a dressing room with sensors in hand while their conclusions drifted away from the actual rhythm of training, and the result was mutual loss of trust.

The second risk is measuring too much. When a junior swims to hit a number rather than to beat the swimmer in the next lane, the effort distribution warps: she conserves energy mid-race to protect a pretty split, and loses in the final fifteen metres. The third risk, one I cannot quantify, is the variance that comes from the stands. A home final in front of a full crowd generates noise no model fully explains; I am forced to widen the confidence interval on any judgement made under those conditions. Every shock has its own probability. We call it a shock when we have not yet checked the table.

The signal I am watching over the next two to three seasons is not a national record. It is the format of the results sheet at domestic meets. If reaction time and 15-metre splits start appearing routinely on the scoreboard, I expect performance gains to concentrate in the turns and the back half of races rather than in raw speed. At that point, a fourteen-year-old who is slower than her peers over the first length but holds her rhythm over the last will be kept inside the system instead of being cut at selection. Most people watch the finish to understand a race. I watch the race to understand the years. A tactical era dies when nobody reads its data sheet.