Rome 2026 and 43 World Records: When Swimming Data Had to Learn to Forget

Over eight days of competition at the World Aquatics Championships in Rome in 2026, 43 world records were shattered. A normal championships produces just a handful. The number 43 is so large that it is no longer a roll of honour to admire, but a big question mark placed before the entire swimming-data analysis industry. Fifteen years later, every time I reopen those old spreadsheets, I still feel I am reading a confession rather than a medal table. The number does not lie, but it knows how to hide something. In 2026 and 2026, world swimming entered what experts call the "high-tech suit" era — the age of polyurethane swimsuits that covered the whole body, lifting swimmers higher in the water and cutting drag to an almost physically impossible degree. The peak came at Rome 2026, when record after record was wiped out within days. World Aquatics, then still called FINA, had to act. From 2026, polyurethane suits were completely banned in international competition, and swimming returned to the "textile" era. That was when the concept of "era screening" was born in analytical circles. When assessing a performance, the first question is not "how fast", but "what year was it swum, in what suit, and under which standard". A world record set in 2026 is not of the same nature as one set in 2026, even though both carry the same title. This leads to a very specific technical consequence. In swimming, every performance is recorded in two systems: the 50-metre long course and the 25-metre short course. Short-course times are generally much faster, because swimmers get twice as many turns to push off the wall, and every turn is an extra launch. The two record systems are therefore entirely separate. Mixing them is the most basic mistake a newcomer can make. A swimmer may hold a short-course record yet never have come close to the long-course mark — and vice versa. Ignore the pool, and every comparison becomes meaningless. Then there are the tiers of records. World record, continental record, national record and championship record are four tiers of very different difficulty. A national record broken in a small country does not mean the same thing as a continental record broken. Above all, a record in the 50-metre butterfly is of a wholly different nature from one in the 1,500-metre freestyle — where endurance and pacing decide everything. This is where the data starts to tell a story. A distance swimmer in the 800m or 1,500m is judged not only by total time, but by split structure. The race is divided into 50-metre segments, each timed. Out of this arises a concept I pay particular attention to: the negative split — swimming the second half faster than the first. In distance events, a negative split shows a swimmer who paces well, saving energy and unleashing it in the closing half. In sprint events, especially the 50m and 100m, the structure is the reverse — the first half is usually faster, because speed and reaction off the start dominate the result. Another metric viewers often overlook is the 15-metre rule. After the start or a turn, a swimmer may stay underwater for a maximum of 15 metres. This is the zone where underwater speed is usually higher than surface speed, thanks to the leg kick and a hydrodynamic posture. Elite swimmers, especially in butterfly and freestyle, turn those 15 metres into a genuine weapon: they glide almost the full permitted distance underwater before surfacing. So if you look only at the total time and ignore the 15-metre segments at the start and after each turn, you miss most of the technical story. In breaststroke, the rule is even stricter. A swimmer may perform only one dolphin kick immediately after the start and after each turn, before switching to the breaststroke kick. Any extra kick is a violation. These legal boundaries create a layer of risk that any data analyst must always attach to their numbers: a performance may look beautiful on the sheet, yet carry the risk of disqualification if an official catches a technical fault. Back to Rome 2026. When records fell en masse, the truly frightening thing was not the number, but this: a generation of records was born under conditions that could never be repeated. After 2026, when high-tech suits were banned, many old world records stood for years — not because no one was better, but because no one was allowed to wear that suit again. That was when "era screening" became mandatory. When assessing a contemporary swimmer, the right question is not "how does his mark compare with the world record", but "how does his mark compare with the standard of his own era". To understand why swimming data is so complicated, you have to look at the sport's map of power. The United States remains the number-one power through systemic depth, with the NCAA collegiate system acting as a giant production line of athletes. Australia keeps its tradition in middle- and long-distance freestyle. China has risen systematically across many events. Europe produces single-point breakthroughs, while Canada has made its mark especially in women's swimming. Each swimming nation runs a different development model: some rely on the school system, others on national training centres. Understanding a country's development model is the key to forecasting the next wave of swimmers. Signals of the future usually do not lie at the world championships, but at junior events — age-group world championships, age-group records. This is where data gives us the earliest view of the next generation. But it is also here that the biggest trap lies: ignoring the "puberty barrier", when a changing body can cause performance to stall or decline, most sharply among teenage female swimmers. A fifteen-year-old prodigy breaking age-group records may not become a world champion at twenty. So every time I read an article hailing the next "prodigy", I remind myself to compare them with their own age-group peers in history. The same holds for selection systems. In many countries, especially the United States, the system called "Trials" takes only the top two finishers in each event on finals day. That means a swimmer may be ranked third in the world and still miss the Olympics, simply because two compatriots are faster. In other systems, the A-cut and B-cut standards apply: hitting the A-cut means automatic qualification, while the B-cut means waiting for quota allocation. Misunderstand this mechanism, and an analyst will easily make wrong predictions about who will appear on the big stage. In swimming analysis, there are also systematic biases. The public tends to undervalue swimmers from countries without a strong swimming tradition, and automatically overvalues familiar names. That is a cognitive distortion data can help correct: look at head-to-head results, at seasonal improvement rates, at split structure — rather than at reputation alone. Beyond the technical side, swimming is a complex governance environment. World Aquatics runs competition rules and the international record system; WADA and national anti-doping organisations monitor prohibited substances; and the Court of Arbitration for Sport handles appeals. Whenever a doping allegation arises, the most important thing is to separate four tiers clearly: confirmed violation, contamination dispute, procedural violation such as a missed test, and a purely public-opinion allegation. Blurring these four is the gravest mistake a reporter can make. Another little-noticed dimension is the sporting nationality switch. A swimmer may change the federation they represent, but must go through a release procedure and a fixed waiting period. For analysts, this is a signal worth tracking, because it can change the balance of an event without any record being broken. Behind every lane lies an entire industry. The swim-coaching market, the equipment and swimwear sector, event business, the athlete-representation ecosystem, facility investment, and derivative markets are all affected by a major event. The star effect is real: one gold medal can trigger a wave of swimming enrolments in an entire country. But to measure that wave, you need a concrete event as a marker — a record, a title, or a policy change. I remember that Saigon summer, sitting by an old computer, hand-building a spreadsheet named "Chance-Counting Data". I learned that data, too, needs to be watered. A dry number standing alone will die. But place it beside the context of its era, the pool, the rules, and the athlete's injury history, and it begins to live and to tell a story. Maintaining a dataset never ends; it is like wiping the glass of a swimming pool — always needed, and invisible to all, until the water turns murky. In the opposite direction, there is one thing I want to stress: most of the swimming data we consume daily comes from results reports — where every number is clean and clear. But it is precisely the articles with no numbers, or with numbers but no context, that often carry the more important stories: a policy change, an injury, a quiet transfer. Ignoring them because they are "hard to analyse" is a form of data bias. And here is my counter-intuitive view. We tend to believe that more data means firmer conclusions. In swimming, that is not true. A world-record table is packed with impressive numbers, yet most of them belong to an era that cannot be repeated. A dataset not screened by era, by pool, by rules, will lead us to confidently wrong conclusions. The paradox is this: sometimes the most honest way to read a number is to admit we cannot yet conclude anything from it. Emptiness is not failure — emptiness, recorded properly, is itself a kind of information. Every record is a data point, but not every data point is a record. I once sat for hours before a screen only to conclude that my data was insufficient to say anything. At first it annoyed me. Later I understood: that is precisely the moment an analyst grows up. A disciplined data worker does not fear blank space; what he fears most is filling blank space with unsourced stories. The lesson from Rome 2026, and from every swimming dataset I have ever built, is not which record is the most beautiful. It is this: to read a number correctly, you must first learn to ask under what conditions that number was born. Once you can answer that, every record — old and new — begins to converse. And that is when data, the driest of things, becomes the most vivid thing on the water.