The Supersuit Era Is Still Mispricing Swimming's Record Board

Men's 100m freestyle final, Rome, 30 July 2026. César Cielo touched the wall in 46.91 seconds. The scoreboard carried that number, and the swimming world believed it would stand for at least a decade. It stood for thirteen years. On 13 August 2026, again in Rome, David Popovici swam 46.86 and erased it from the books. What made me stop at 46.91 was not its speed but its durability against every other figure on the same page. Numbers have no gender, but the people who read them do. Anyone reading swimming's record board over the past fifteen years has been reading a table contaminated between 2026 and 2026, when full-body polyurethane suits let the human body float like a buoy. I read tables for a living, and this is the dirtiest table in any sport I have ever analysed. To read it properly, you need to know how swimming data is generated. Unlike football — where I work with xG, PPDA and distance covered — swimming is a sport of a single variable: time. But raw time says nothing without splits. Every 100m freestyle race divides into two 50m splits; 200m into four; 400m into eight. From that chain of splits come the concepts analysts use to price a swimmer: start and underwater work, turn efficiency, stroke rate, distance per stroke, and split structure — negative split or front-loaded. A 25m short-course pool and a 50m long-course pool produce two data systems that cannot be directly swapped, because short course doubles the number of turns, and every turn is a push off the wall. World Aquatics recognises two separate record sets for exactly that reason. Anyone comparing a short-course mark to a long-course mark without naming the reference frame is doing the job wrong. There is another layer: qualifying standards. To reach the Olympics or a world championship, a swimmer must hit the qualifying time inside a window set by the federation, and each country gets a limited number of entries per event. A swimmer can beat the standard and still be left out because two compatriots swam faster. This is the flashpoint of every selection controversy, and it is also where data gets distorted most. World Aquatics — then still called FINA — approved the ban on high-tech suits in July 2026, effective 1 January 2026. The new rule forced suits to be made of textile and capped thickness and buoyancy. It was the right decision. But it left a consequence nobody handled: the world record board kept its old entries. At the 2026 world championships in Rome, 43 world records fell in eight days. Set that figure beside an ordinary championship with a handful of records, and the distortion is plain. But the trap lies elsewhere: the contamination is uneven. In events where the suit delivered an enormous advantage — longer distances, where buoyancy held the body horizontal and cut drag — the 2026 records still stand today. In events where the edge was smaller, the records fell long ago. The result is a hybrid record board. Same page, same font size, but half the lines were produced under different rules. When a commentator says "this record has stood for fifteen years", he is comparing two things that do not share a reference frame. That is a foundational error, and it repeats every Olympic cycle. The clearest case is the men's 100m freestyle. Cielo's 46.91 survived three Olympic Games. It only fell in 2026, and the man who beat it wore a fully textile suit. Read 46.91 as a pure marker of human capacity and you misprice the entire generation in between. Swimmers like Cameron McEvoy, who raced around 47 seconds for nearly a decade, were undervalued simply because they came after the ban. In the women's 200m freestyle the story is even clearer. Federica Pellegrini swam 1:52.98 in Rome in 2026. That number held for fifteen years, across three Olympic Games, until Ariarne Titmus beat it in 2026. For all those fifteen years, every female swimmer in the event was measured against a yardstick her era had no right to touch. This is what I call systemic mispricing — nobody cheated, but the table still lied. Now to the data that is genuinely usable. Splits. Based on my experience tracking races across five seasons in Brisbane and Melbourne, what I record is not the final time but the gap between the two halves of a race. A male 200m freestyler who swims the first half faster than the second tends to collapse over the final 50m. Conversely, a clear negative split — a faster second half — signals a solid aerobic base and trustworthy energy distribution. In my data, that pattern repeats often enough to count as a signal, not a coincidence. Technically, the freestyle catch is where split data becomes most useful. A swimmer who raises stroke rate while losing distance per stroke is usually muscling the water rather than swimming it, and that shows up in the third split of a 200m. Conversely, someone who holds distance per stroke while lifting rate over the last 50m has mastered energy distribution. I use this to separate a fast result that is durable from a fast result that is temporary. But splits are not the whole story. One variable escapes the table: the underwater segment after the start and after every turn. Since the rules allowed up to 15 metres underwater, elite swimmers have turned that segment into a weapon. In a 25m pool, a strong underwater swimmer can gain two metres per turn over a weak one, and with eight turns in a 200m, that accumulates into half a body length. This is why I keep telling bookmakers that short-course results cannot be used to infer long-course results without a conversion factor. Then there is the talent supply chain. Swimming carries a paradox of gender and age: in women's events, peak performance often arrives early, sometimes before 20; in men's events, it usually arrives later, after 24. The puberty barrier is a real variable. Female swimmers who are blisteringly fast at 15 can lose momentum as their bodies change, and many talents in that group vanish from the tables before they turn 20. This is a data zone where I refuse to draw conclusions, only to take notes. A head coach once told me at a technical meeting in Brisbane: "You can measure a child's time, but you cannot measure whether they still want to swim next year." I wrote that line down, and it became one of the most important entries in my file. At market level, this distortion has a price. Bookmakers build record markets by anchoring to all-time lists. If that list contains lines from 2026, every anchor is dragged optimistic or pessimistic depending on the event. When I consulted for a betting company in Brisbane, the first thing I did each season was audit every line of the record board and flag which belonged to the high-tech suit era. It is manual, time-consuming work, and almost nobody does it. The most recent Olympic Games were a test. In many events, leading swimmers closed in on old records without breaking them, and the media called it a stagnation of world swimming. That reading is wrong. They have not stagnated; they are hitting the ceiling of a clean era. The distance between those two readings is the entire value of an analysis. One more signal: sporting nationality switches. Over the past decade, the number of swimmers changing the nation they represent has risen, dragging disputes over residency periods and eligibility behind it. For betting and media markets, this is a variable that can shift an entire medal map inside a single four-year cycle. Above all of it sits the anti-doping system: every record lives under an unanswered question about ten-year stored samples. That is why I never use the word "certain" about a landmark. Here I have to argue against myself. The popular reading today is to discard the 2026–2026 records entirely as garbage. I do not fully agree. The swimmers who broke records in 2026 were not products of a suit — they were already at the top. Cielo won Olympic gold in 2026 when the suit rules had not fully opened up. Pellegrini had broken world records since 2026. Discarding all their landmarks is a form of data cleansing, and data cleansing is also a way of lying. The problem is not the number. The problem is the interpreter. A time of 1:52.98 does not say by itself whether it is fast or slow; it says only that it was faster than anything previously swum, in a suit that is no longer legal. When an analyst places it beside a 2026 performance without context, he has created a new bias to replace the old one. And this is the second limit, the one I learned in Kazan: correlation is not causation. A swimmer with the finest splits in my data can still finish third. A swimmer with worse numbers across the board can still touch first. Pacing, psychology, a coach's wrong call on energy distribution — none of that fits neatly into a table. Kazan was the day I learned that a 99% probability can still die on the betting table. It did not teach me that data is wrong. It taught me that data is only right within its own margin.