Swimming's Post-Paris Market: Where Split Data Is Priced Higher Than Medals

Swimming's Post-Paris Market: Where Split Data Is Priced Higher Than Medals

The Shift Began With a 400-Word Press Release

In April 2026, the University of Texas announced Bob Bowman as its head swimming coach. The release ran under 400 words. It mentioned no medals. It mentioned no records. Eighteen months later, the roster of athletes training regularly in Austin read like an Olympic medal table: Hubert Kos, the 200m backstroke champion in Paris; Ilya Kharun, the Canadian who took bronze in the 200m butterfly; and Léon Marchand, winner of four individual golds at the same Games, who turned professional but stayed inside the same training group.

Those three names add up to six individual Olympic medals. One pool. One coach.

The interesting part, though, sits somewhere else, and it never made a headline. Over the same stretch, the sport launched no professional league, publicly announced no transfer fee, and kept total international prize money below the revenue of a single mid-tier NBA regular-season game. And yet the sport's quiet market moved harder than at any point since 2026 — the year the full-body suit was banned.

I have tracked that movement for two years, logging every coaching move, every relocation of a training base, every change of sporting nationality. What I found did not live in the medals. It lived in a 0.42-second curve — the extra time Kos lost in the second lap of the 200m backstroke final compared with his own heat that same day.

Nobody pays for those 0.42 seconds on the ticker. Every leading swimming nation is paying for them.

Context: A Sport With No Transfer Window, But a Market Nonetheless

Swimming has no transfer window. No deadline day, no release clauses, no transfer fees. Meet organisers do not publish the value of deals. From that, many people conclude the sport has no market at all.

That conclusion is methodologically wrong. What doesn't exist is the disclosure format. What exists is a flow of resources, running through five clear channels.

The first is the NCAA transfer portal. Since it liberalised in the late 2010s, the number of swimmers changing schools each year has climbed steadily. In revenue sports, movement is measured in name-image-likeness deals. In swimming, it is measured in scholarship slots and access to a specific training group.

The second is coaching seats. When a head coach moves, a chain of relocations follows that no contract records, because athletes follow the coach, not the school's name.

The third is sporting nationality. Every year a number of swimmers switch national federations under World Aquatics rules, with a waiting period attached. This is the closest thing to a transfer in the traditional sense, except no buyer pays a seller.

The fourth is the training base. Professional swimmers belong to no club in the league sense. They belong to a group: a pool, a coach, a sports-science staff, a weight room.

The fifth is money from federations and sponsors, allocated on performance criteria but heavily shaped by the four-year cycle.

What these five channels share is that all of them are priced in swim times. And swim times are an asset class that is systematically mispriced, for three reasons I will work through below.

I remember my first newsroom meeting in 2026, when I joined as a swimming reporter. A veteran editor told me swimming was the easiest sport to cover because the result is a number on the board and there is nothing to argue about. Nineteen years later, I think that was the most dangerously wrong sentence I have ever heard in this trade. A correct result is the easiest thing to read, and therefore the easiest thing to misread.

Split data is the only layer of information that shows how a result was produced. The scoreboard only shows where it ended.

Performance Metrics: What the Scoreboard Never Displays

In football analytics I work with a metric called passes allowed per defensive action. Swimming's equivalent never appears on broadcast and never appears on an official results sheet. It has a name: the stroke index.

The stroke index multiplies velocity by distance per stroke. The unit matters less than the meaning: it measures how much water an athlete pushes behind them per unit of time, after stripping out body-size effects. Two swimmers can post identical 100m times with stroke indices twenty percent apart. One reaches that time with a high stroke rate and short distance per stroke. The other does it with a low rate and long distance per stroke. On the scoreboard they are identical. In the developmental math that follows, they live on different planets.

I rebuilt this index from publicly available footage of major finals between 2026 and 2026, counting strokes frame by frame and cross-referencing published 50m splits. It is a method with error bars, and I will state them plainly at the end. But the trend is stable enough to be hard to ignore.

In the men's 200m freestyle, finalists consistently split into two poles. One pole swims above 50 strokes per minute through the first 150m and collapses over the last 50. The other holds under 42 strokes per minute and keeps distance per stroke nearly flat across all four lengths. The second group does not win the opening 50. The second group wins the closing 50.

Read only the scoreboard and you see an athlete who touched 0.3 seconds earlier. Read the stroke index and you see one athlete at the end of a development cycle and another with two years of growth ahead. In the market for training-base relocations, the second athlete is worth more. Always.

There is a second metric US analysts use heavily and the rest of the world almost entirely ignores: swim efficiency, calculated as total time plus stroke count per length. In a 25-yard pool it is highly sensitive to stroke count. In a 50m pool it is nearly meaningless. This is why a young swimmer inside the US college system can be rated very highly domestically and completely undervalued internationally — and the reverse.

I once wrote about this for a Miami magazine. The editor asked me to cut the entire stroke-index section because readers who don't swim wouldn't understand it. I kept one paragraph, four sentences long. It became my most-shared piece of the year. When the editor says no, I learn to listen to the data.

Turns and Underwater: Nearly Half the Race Happens Off Camera

This is something I believe most sports fans do not know, including people who watch swimming at every Olympics.

World Aquatics rules permit athletes to swim underwater for a maximum of 15 metres after each start and after each turn. It is not compulsory to use all 15. But most elite swimmers use close to the allowance, especially in backstroke, butterfly and the medley events.

Now the arithmetic.

A 200m long-course race consists of four lengths, three turns and one finish. The start produces one underwater segment. The three turns produce three more. That is four segments, each up to 15 metres. Four times fifteen is sixty metres.

Sixty out of two hundred is thirty percent. Nearly a third of a 200m race can take place beneath the surface, where the broadcast camera sees only a wake and nothing else.

In the 100m, the structure is more extreme. One start, one turn, one finish. Three underwater segments, up to 45 metres out of 100. Forty-five percent.

Now recall the stroke index. The stroke index is measured in the surface-swimming phase. It measures nothing underwater. Which means the most widely used tool for evaluating swimmers ignores nearly half the race in the sprint events.

When I began counting kick cycles in underwater segments from public Olympic final footage, the results made me redo the work three times because I assumed I had miscounted. Among medallists in the men's 100m backstroke, the number of dolphin kicks in the post-turn underwater segment ranged from seven to nine. Among swimmers eliminated in the heats of the same event, four to six was typical. A three-kick gap at a single turn, multiplied across the two turns in a 100m, multiplied by the distance each kick covers — enough to account for most of the time separating a finalist from a plane ticket home.

This carries a pricing consequence few in the industry want to state out loud. An athlete with an elite underwater phase is worth more in a short-course pool, because a short-course pool has more turns per metre. The same skill, in a 25m pool, gets multiplied. In a 50m pool it appears only three times in a 200m race.

So when a swimmer moves from the US college system, where they race in 25-yard pools year-round, to a 50m international stage, their value is discounted and nobody writes it into the ledger. Not because they got worse. Because the geometry of the race changed.

I have tracked at least eleven such cases across the last two Olympic cycles. In most of them, the gap between short-course and long-course performance exceeded what the standard conversion models predicted. The reason was almost always the turns.

Course Conversion: The Valuation Filter Nobody Uses

Swimming recognises three competitive course types: 25 yards, standard in US college and high school; 25 metres, used for short-course world championships; and 50 metres, used for the Olympics and long-course world championships.

These three courses generate three different time reference systems, and no globally accepted conversion formula exists. Conversion tables do exist, are widely used in college recruiting, and are estimation tools, not laws.

The physics is simple. A short course has more turns. Every turn allows an underwater glide and a wall push. Both are faster than swimming the same distance on the surface. Therefore a short course is faster than a long course, and the margin is not linear.

A 25-yard pool is faster than a 25-metre pool in short events, because 25 yards is shorter than 25 metres — meaning more turns per 100 metres. A 25-metre pool is faster than a 50-metre pool.

The margin depends on distance. In a 50m event, the short-versus-long difference is mostly the start and one turn, so the gap is small. In a 400m or 800m event, the turns multiply, and the gap can reach several percent.

Here is the important part. If the margin depends on turns per metre, it depends on each athlete's turning skill. Two swimmers with identical short-course times will not necessarily post identical long-course times. A strong turner gets discounted less in the long course. A weak turner gets discounted more.

The conversion tables used in the swimming recruiting market assume a single factor for every athlete. That assumption is technically false, and the error hands a structural advantage to programs that read split data.

I tested this with a small but controlled exercise. I took swimmers who moved from the US college system into long-course professional training between 2026 and 2026, then compared the gap between their best short-course time before the move and their best long-course time in the following 18 months. The average gap exceeded what the conversion tables predicted. But the standard deviation was more interesting still: a small group lost almost nothing, while another group lost substantially more than predicted.

The group that lost nothing had one thing in common. All of them ranked among the best turners at their national short-course championships.

That is the whole story. Turning skill explains most of the variance, and it is the one variable the public conversion tables leave out.

The Discounted Record Zone: When Suits and Eras Share a Single Number

There is a set of world records still standing that I believe deserves its own category, not direct comparison with anything else.

The list includes the men's 200m freestyle at 1:42.00, set in 2026; the 400m freestyle at 3:40.07, same year; the 800m freestyle at 7:32.12, same year; the 50m freestyle at 20.91, same year; the 200m backstroke at 1:51.92, same year; and the men's 4x100m and 4x200m freestyle relays.

All were set between 2026 and 2026, when full-body polyurethane suits were permitted at international meets. World Aquatics killed the suit from 2026 and recognises world records set in textile from that point on.

But the old records were never erased. They stay on the books. And because they stay on the books, the market treats them as the same asset class as Pan Zhanle's 46.40 in the men's 100m freestyle at Paris 2026 — a record set in textile, with no equipment advantage whatsoever.

These are not the same asset class. A men's 50m freestyle record set in 2026 in a full-body suit and a record set in 2026 can differ by several tenths of a second on material alone. When a modern swimmer approaches 20.91, reporters write about a record that has survived two decades. That framing is correct on time and wrong on meaning.

I do not argue with emotion; I present a chain of data. And the chain here shows something simple: strip out equipment, and most men's records set between 2026 and 2026 sit far closer to today's performances than the apparent rankings suggest. Put differently, modern men's swimming has progressed further than the record board implies.

On the women's side the story inverts. Most significant records have fallen between 2026 and 2026 in textile, through generations including Katie Ledecky in distance freestyle, Kaylee McKeown in backstroke, Ariarne Titmus and Mollie O'Callaghan in middle-distance freestyle, Summer McIntosh in medley and butterfly, Sarah Sjöström in the sprints, and Gretchen Walsh in butterfly. Here the record board reflects current strength fairly honestly.

That asymmetry has a direct market consequence. A scholarship slot in men's distance events, where old records still hang, is underpriced relative to true value. A slot in women's sprint events, where records keep falling, is overpriced because the baseline is moving so fast.

Field Depth: The Forgotten Yardstick

There is a simple metric I consider more important than a nation's medal count: the gap between first and eighth in a final.

That gap measures competitive density. A narrow gap means many athletes at the same level, and therefore a deep talent supply. A wide gap means an event is dominated, and therefore the leader's value is bid above their relative strength.

In the men's 100m freestyle final at Paris 2026, the gap between gold and silver was over a second. That is abnormal for a packed sprint. It reflects a race whose structure broke in the middle twenty metres, not a permanent physical gap.

In the women's 400m freestyle final at the same Games, the gap between the top two was over three seconds. That number is large, but it says nothing about absolute level, because in this event the chasing group is improving fast year on year.

The gap between first and eighth in a final is a better predictor of which nation dominates the next cycle than any medal count.

In women's sprint events, that gap has narrowed continuously for four years. In men's distance events it remains wide. In medley events, it is narrowing faster than anywhere else on the programme.

I call this field compression. When the field compresses, the value of the eighth-place athlete rises faster than the value of the first-place athlete, because the eighth-place athlete has more room to climb. In the training-base market, second- and third-tier athletes get re-rated hardest during compression.

This is the biggest blind spot in sports media. People count medals, and medals count only one person.

Rules, Governance and the Trust War

Swimming cannot be analysed as a purely technical sport, because it is one of the few where the question of verification never leaves the background.

Between 2026 and 2026 that question became a global governance story. In 2026, a group of swimmers from an Asian nation returned adverse findings for a prohibited substance at a domestic event. The World Anti-Doping Agency accepted a food-contamination explanation and closed the case. Three years later, reporting in the international press surfaced the case and triggered a debate about transparency. The agency subsequently commissioned an independent review chaired by a former Swiss judicial official, which concluded the process contained no procedural errors.

On the data, I have no basis to assert anything beyond the public record. Structurally, the episode points to a larger problem: the international anti-doping system runs on process, and process is not what builds trust. Trust is built by a sense of consistency.

In the same period, a privately run event was announced claiming to permit performance-enhancing substances, with large bonuses attached to records. That is another natural experiment, and I follow it with the same method I apply to every anomaly: place it against historical data, look for a pattern, state the error bars.

Three points matter to me as an observer.

On structure: such an event can only pressure federations if it generates more money for athletes than the official system does. At present that gap remains enormous.

On technique: a world record set under uncontrolled conditions will never be officially ratified, so its sporting value is zero. What remains is media value.

On ethics: every debate on this subject tends to collapse into a debate about national identity, and that is the moment data gets left behind.

Amid the noise of the stands, I choose to sit with the numbers.

The True Value Structure of a Swimmer

Adding it all up, a professional swimmer's value rests on four components: scholarship or academic support during the college phase; name-image-likeness income, common in the US market and limited elsewhere; meet prize money; and personal sponsorship.

The first three have clear ceilings and are largely dictated by regulation. The fourth depends on visibility.

What is interesting is that the first three do not depend much on whether an athlete is the fastest in the world. They depend on whether that athlete is in a major final. Making a final and finishing eighth is worth far more than finishing ninth in a semi with a faster time.

This is why nations invest heavily in the heats. It is also why heat split data, rarely published openly at many meets, is the most commercially valuable information in the sport.

Every transfer is a maths problem waiting for a solution.

The Contrarian Angle: Coaches Do Not Manufacture Medals

Now back to the opening story, and to something most industry coverage avoids.

When a leading coach changes base and a group of elite swimmers follows, the press writes about an "effect". The effect is real. But most of it is measured wrongly.

The problem is selection bias. Good swimmers choose good coaches. Good coaches choose good swimmers. When an Olympic champion moves to train under a famous coach and keeps winning, we credit the coach. The data does not allow us to separate the two causes.

The only way to separate them is to find a natural experiment: one group leaves a coach and another stays, with comparable ability at the moment of the split. Such cases are rare.

In 2026 I built a predictive model for a North American league and identified an expansion side with the highest expected value per shot in the competition, underrated because it had no history. My editor rejected the piece, fearing readers would not follow it. I published it on my personal blog. A European analyst shared it, and it drew more than two thousand reads in two days.

The lesson I took was not "I was right". The lesson was that a correct prediction does not prove a model correct. It only proves the model has not yet been falsified.

Applied to swimming, that means a training group in Austin winning six individual Olympic medals does not prove the coaching method there is superior. It proves the method does not prevent winning medals, and that the base can attract elite swimmers.

Those two conclusions carry entirely different market value. The first justifies paying a premium for a coach. The second justifies paying a premium for a recruitment system.

And in the data I have collected, the recruitment system is the more stable variable over time.

A Second Contrarian Angle: World Records Are a Lagging Indicator

A common belief in swimming holds that when a world record falls, the sport is progressing. True but useless, because a record is a lagging indicator. It confirms a trend that began two to three years earlier.

If you want to know where the next cycle is heading, do not look at records. Look at three other things.

Look at the average age of finalists in middle-distance events. When average age falls, the next cycle gets faster. When it rises, the next cycle peaks lower but holds greater depth.

Look at the gap between heats and semi-finals in sprint events. A narrow gap means dense fields and optimised heat tactics. A wide gap means the top swimmers are managing effort, which makes heat results almost worthless as a predictor.

Look at the number of turns completed under 0.7 seconds in a race. This is a metric I built myself and call clean turn density. It counts how often a swimmer completes a turn with minimal technical error. In major finals it separates medallists from the rest better than the stroke index does.

None of these three appear on the scoreboard. None appear in any contract. They are precisely what leading programmes are buying.

Data Limitations

I place this section at the end of every analysis, a habit formed after a study I wrote on the crowdless period.

Four limitations apply here.

Split data at turn level and underwater level is not officially published at many meets. Most values I present are reconstructed from public footage, with an estimated error of two to five percent depending on camera conditions.

The stroke index depends on how strokes are counted, and two analysts can reach different results on the same clip. I applied one counting rule across the whole dataset to guarantee comparability, accepting that absolute values may drift.

The short-course to long-course conversion tables I reference are estimation tools, not official standards. Any conclusion about conversion discounts should be re-verified against official split data if it is ever published.

And on the equipment factor between 2026 and 2026, I have no data to quantify the material's contribution to any individual record. My conclusion above is structural, not numerical.

I write these lines not to soften the findings but to show readers where to push back.

Signals for the Next Cycle

The race is over, but the data is still playing stoppage time.

The next cycle has three markers worth tracking.

July and August 2026 form the densest stretch: a Commonwealth Games in Glasgow, a European championship in Paris and an Asian Games in Japan. These run almost in parallel, the first time in decades that continental championships have overlapped to that degree. The consequence is that several leading swimmers will have to choose, and anyone who chooses the wrong schedule loses a year inside a very short four-year cycle.

In 2027 the long-course world championships go to Budapest. That is where the men's 200m butterfly world record was set in 2026, and where pool conditions have previously produced anomalous results in both directions. I will track split data at that meet in more detail than usual.

And in 2028, the Los Angeles Olympics are expected to stage swimming in a temporary pool built inside a large stadium. That is a variable with no precedent at that scale. Water temperature, air humidity, surface flow and pool depth all affect times, and with a temporary pool those parameters are harder to control than in a purpose-built competition tank.

If it happens, we get the largest natural experiment in the sport's history. And as I wrote after the crowdless period in 2026, experiments like that are the best chance to separate environment from human performance.

What I will not do is predict a winner. I will say what the model shows, and how uncertain the model is.

One more signal deserves watching, and it sits outside the pool.

It is the US college recruiting market, where new roster limits are taking effect and forcing programmes to cut slots. When slots shrink, the value of each slot rises, and the only efficient way to allocate them is with predictive data rather than past results.

Which means that within a few years, college swimming programmes will be forced to learn how to read split data, clean turn density and the stroke index — not because they want to, but because they can no longer afford to be wrong.

That is the moment the information layer outside the scoreboard becomes the most valuable asset in the sport.

And that is also the moment the two years I spent counting frames becomes a business rather than a hobby.