The Pool and Its Columns of Numbers: When Data Puts a Record Back on Trial

At La Défense Arena on July 31, 2026, when Leon Marchand touched the wall in the 200m individual medley final, I wasn't looking at the number on the scoreboard. I was looking at the split file I had built over three weeks before the race, updated daily with twenty variables per lap. Four legs — butterfly, backstroke, breaststroke, freestyle. Marchand swam the breaststroke leg an average of 1.3 seconds faster than his rivals, but what made me sit up straight wasn't that speed. It was his backstroke leg, the one analysts have long treated as the permanent weakness of an attacking medley swimmer. He swam it nearly 0.8 seconds slower than the world-record split, yet compensated by absolutely conserving energy for the two decisive legs. The number didn't shout. But its structure did. I've sat long enough in this job to know that a world record is just the end point of a curve the media never bothers to draw. They screamed when they saw Pan Zhanle's 46.40 in the men's 100m freestyle. They didn't see that 46.40 was built on a start and underwater phase that, three years earlier, no coach would have dreamed of for an Asian swimmer whose arm span is roughly five centimeters shorter than the average Western rival. They called it genius. I call it a decoded training model. The race ends, but the data keeps talking. I came to the pool the way nobody chooses. In 2026, at eighteen, I signed up as a statistics volunteer at the U19 Asian Championship held in Shanghai, and while my classmates were busy recording football scores, I sat beside the pool of a different arena, building tracking sheets for every lane. I counted strokes per cycle, distance per cycle, underwater time after each turn, and breathing-rhythm consistency. I had no coach to teach me. I had a notebook and a naive belief that if someone swims a certain way, it will repeat a certain way, and whatever repeats can be measured. My job now is reading swimming data. I don't race at the elite level, but I swim every morning as a way of keeping discipline, and my own arms know the feeling no spreadsheet can describe: the burn in the shoulders at the seventieth meter of a 100, the moment every prediction model calls "speed decay" and every swimmer calls "hitting the wall." That's why I never believe a number is right by itself. I believe a number is right within its context. And context — the context swimming media usually skips because it isn't sexy — has changed this sport more than any athlete. I remember the summer of 2026, when I first realized data could stand against an entire legend. That was the World Cup summer everyone talked about as football, while I sat writing commentary and building Excel sheets on the side. I still remember that feeling — the feeling of a young man convinced that if he counted enough, truth would reveal itself. I once thought data was the answer. 2026 gave me a better question. Fifty-one seconds, then forty-six. In the pool, history can be compressed into a few digits, and that is precisely the trap. When a world record falls, the media's first reflex is to compare it to the previous record, sometimes three years apart, sometimes three decades, as if the gap in time didn't exist. They draw a straight line from one year to the next and call it "the sport's progress." But swimming doesn't progress in a straight line. It progresses in steps, and each step is usually built by something that isn't in the water. To understand why, we must return to the most haunted summer in modern swimming history: Rome 2026. The World Championships in Rome produced forty-three world records. Not four, not fourteen. Forty-three. Within one week, an entire tier of records believed to stand for decades was nearly wiped out, and the cause was not that humanity suddenly learned to swim faster. The cause was full-body polyurethane suits — literally plastic armor that helped the body float higher, compressed muscle, and repelled water better than human skin. In 2026, the international swimming federation banned full-body polyurethane suits. And so an entire generation of records was permanently marked with an asterisk. Here, what I want you to understand isn't the question of cheating. Wearing what the rules allow isn't cheating; it's racing within a game the rules permit — and that's exactly why it's more dangerous than cheating, because it's legal. And when a record is born under the old rule, and that rule is then abolished, every later world ranking in this sport becomes a comparison between fresh apples and apples wrapped in plastic. I spent nearly a year doing just one thing: rebuilding what I call the "era-adjustment table." I took the history of men's 100m freestyle from the mid-2000s to now, tagged each result by suit type and applicable rule, then calculated the average gap between the polyurethane era and the textile era. The number I got was fairly stable: about 1.5 to 2.5 percent for freestyle events, and even higher for butterfly — the discipline where the suit's vertical lift helps most. What does that mean? It means that if you see a men's 100m freestyle record set in 2026, that number must be reduced by one to two seconds before you can compare it to a record set in 2026 after textile-suit rules were tightened. And when you do that subtraction, you start seeing a strange truth: humanity has swum far faster than we thought. It's just that the asterisk of synthetic plastic had hidden their real speed for fifteen years. Because of that asterisk. The asterisk isn't proof of anyone's guilt. It's proof of the laziness of comparative thinking. Once you strip it away, once you adjust for era, you begin to see real curves. And the real curve of modern swimming isn't the smooth straight line the news wants you to believe. It's a series of steps, each tied to a coaching revolution or an institutional revolution. Look at the splits. This is where swimming data tells its best story, and where the media reads it wrong most often. A 100m freestyle race isn't a single block. It's four different events stitched together: start and underwater, surface swimming, turn, and finish. Each part has its own energy philosophy, and every elite swimmer is a different solution to that energy-allocation problem. Take Pan Zhanle and his 46.40 record in Paris. If you only look at the final number, you see a Chinese swimmer swimming the fastest 100m freestyle in history — something I, raised in an Asian swimming culture, never thought I'd witness. But if you look at the splits, you see something more interesting: Pan didn't win with strength. Pan won with water. His start was modest — about 0.1 seconds behind American record-holder Caeleb Dressel in reaction time. But in the underwater phase after the start, Pan covered a distance I measured as more than two meters longer than most rivals, and more importantly, he held underwater speed longer. In modern swimming, underwater is a forbidden zone for most athletes. Because you can't breathe underwater. And not breathing puts the body into oxygen debt — what training manuals call "drowning." So how does a man with a 1.97-meter arm span swim an underwater phase that long? Because Pan has something my data measures but the media doesn't see: an extremely high post-warm-up breathing elastic index. That is to say, his body tolerates oxygen debt without collapsing. That isn't technique. That's physiology, and physiology is the most neglected element in swimming analysis. Marchand is the exact opposite. He doesn't win with his start. He wins with allocation. In the 400m individual medley, where anyone who has ever analyzed the race knows its four legs are four ticking bombs, Marchand swam the butterfly leg faster than the world-record split, then deliberately reduced speed on the backstroke leg — something I believe was intentional, not a weakness. He knew he could swim the breaststroke leg faster than anyone in the world, and he knew his freestyle finish would be strong enough. So he turned the backstroke leg into a savings account. He didn't race every leg. He raced two decisive legs, and only those two. A spreadsheet has no jersey color, but I still hear the race through every column of numbers. And in Marchand, the columns say one thing clearly: he's an energy-allocation machine, not a speed machine. Those are two different athletes, needing two different training models, dissected by two different analyses. But this is where I have to stop and argue with myself. There's a powerful temptation in my job: once you see a pattern anywhere, you see it everywhere. When you analyze two hundred races and find that most winners had longer underwater phases, it's easy to write "to win, swim farther underwater." That's a mistake about causation, and it's common enough that I remind myself every time I sit at my desk: correlation is not causation. Consider the most famous trap in swimming analytics: stroke rate and distance per stroke. There's an old idea in recreational pools that "fewer strokes but longer glide means stronger than many strokes but shorter glide." People tell each other this in swim classes, in coaching videos, in training sheets. And it's true. For a beginner. For someone trying to swim far. For someone swimming below lactate threshold. But when you reach the elite level, when you measure athletes in an Olympic final, that relationship stops being so simple. I spent many weeks filtering data from major men's 100m freestyle finals over the past seven years, tagging each athlete by final result and by split structure. What I found clearly is this: at peak speed, stroke rate goes up, not down. Champion sprinters stroke faster than their rivals. But they're also longer — meaning they exploit both at once, thanks to something data doesn't measure: the pull on entry and wrist stiffness at the catch. This is what the media doesn't say, and what serious coaches all know. "Fewer strokes, longer glide" isn't a law of elite swimming. It's a developmental phase — a phase every swimmer must pass through before reaching their true phase. Because at high speed, what decides isn't whether you're lazy with your strokes. It's how much momentum each stroke contributes to the water. And here the story gets genuinely fascinating, because underwater momentum is a quantity we can estimate. We can estimate the momentum of a stroke by approximating propulsive force times contact time. In elite athletes, average propulsive force per stroke tends to rise as wrist stiffness rises, as the hand enters at an optimal angle, and as the forearm is positioned as a proper lever. That's why a swimmer with a shorter arm span can still swim faster — because each of their strokes pushes more water per meter, and because they can coordinate more strokes in the same span of time. But this story isn't only about physicality. It's about psychology. And swimming psychology is what data analysts discuss least, even though it's present in every race. I'm writing these lines on a morning when I've just returned from the pool. I swam 1,500 meters, split into thirty fifties. I'm not an athlete; I'm just a man keeping discipline. But when I swim to the 1,200th meter, I understand something data can't teach: fear. That moment when you know you have two hundred meters left, and your body says stop, and your reason says continue. No spreadsheet captures that moment. But that moment decides a great many races at the Olympic level. I think about the young athletes I follow. The ones every prediction says will break out, but who then don't, or who break out later than forecast. And every time that happens, I have to ask myself: where was I wrong? What did my model miss? The answer usually isn't in the water. It's in the locker room. It's in the capacity to endure pressure, in the coach who places faith in a seventeen-year-old, in whether parents can keep paying for another four-year training stint with no guarantee. My data model can't calculate that, and that's why I always have to remind myself that a good model is one that knows its limits. Strategy is a hypothesis. Every hypothesis needs a Korean night to test its fire. Here I must be clear about the so-called "Korean night." In 2026, on a summer evening while I was watching football, I was shocked to see one of my statistical models diverge completely from actual results. Not because the model calculated wrong. Because the model didn't know what only humans know: that a national team can compete with a level of ferocity and focus no prior statistic could predict. That was my lesson, and it applies to swimming identically. You can have every metric, and you can still be beaten by someone who swims faster than you thought for a reason the schedule doesn't state. And here is what I want everyone following swimming this year to remember, especially those reading prediction tables: when you read a number, ask about the third variable. That's the single most important question in my job, and it's simple enough that I have to repeat it like a mantra. What's the third variable? It's anything that could explain a correlation without needing causation. If a swimmer swims faster, is it because they trained better, or because their rivals were weaker? If a national team suddenly erupts in results, is it because their training system changed, or because they just benefited from a rare generation of talent? If a new swimsuit appears, is the performance gain from the suit, or simply because the athletes wearing it also happen to be at their career peak? All those questions can't be answered by looking at one number. They need a whole data series, an understanding of context, and a humility that sports analysts often lack. I've seen too many articles dissecting a race while only looking at the final time. If you want to understand an athlete, you must look at the splits, the whole sequence of form across meets, the schedule, and what happened in the three weeks before that affected their state. You must know how many times they swam that week. You must know where they came from psychologically. You must know that what the media calls "form" is really a function of many variables, most of which aren't in the results table. But I also don't want you to leave this article feeling everything is vague. Because the opposite is true: swimming is one of the best-documented sports in world athletics. Every race at a major meet generates hundreds of data points. Every touch can be split into dozens of segments. The problem isn't a lack of data. The problem is that we read it lazily. And here is what I want to leave for the final part of this article, the part I call signals for the next cycle. I'm tracking three things going forward. The first is the shift toward underwater. Coaches are increasingly attentive to the start and underwater phase, and that will show up in split records. Watch for young athletes with good breathing indices and unusually short arm spans — those are the ones who will surprise, because underwater technique is steadily mattering more than arm span. The second is the return of middle- and long-distance events. As medley and distance swimming become more compelling, coaches will face the energy-allocation problem differently, and athletes who allocate well will rise. That's why I think that over the next four years we'll see a few athletes start their careers in sprints and end in distance, rather than the reverse as before. The third is the friction between national training systems. I think we're entering a phase where three systems — the US, Australia, and Asia — will increasingly compete not only on the water's surface but in the laboratory. Because the real swimming race, the one the media never broadcasts, is the race between data centers, sports-science groups, and split-analysis programs. And I don't know for certain what will happen. No one knows for certain. But I know one thing: while the media argues over one number, behind the curtain, in offices full of spreadsheets, someone is already calculating the next cycle. When football stood still in 2026, I found speed within myself. Four years later, I'm still searching for that speed in people younger, faster, and — I hope — wiser than the generation I watched when they first entered the pool. Because swimming, after all, isn't about touching the wall fastest. Swimming is about understanding how much water is left in you, and knowing when to spend it on the final leg.