Five Thoughts (Hot Takes?) On The Olympics & The Sport

What a week! What a sport! And what a community we have to follow along with.

Now that the pool swimming has finished for yet another Olympiad, I wanted to share some thoughts about the meet and the sport.

So as not to repeat what others have said better, I figured it would be more interesting to share a few ideas that might be at least a little bit out there, perhaps even controversial. After all, what is sport, and what is the internet, without arguments over trivialities?

Despite the length, these aren’t intended to be airtight arguments (although I really do believe them), but more like suggestive musings, and possibly conversation-starters.

Let’s start with:

Hot-take #1: McIntosh and Marchand sure are super swimmers!

Hmm, not controversial enough? OK, how about:

Hot-take #1: McIntosh and Marchand are the SUPER DUPER-EST of swimmers!!!

Still not that interesting? I guess we need to take it up a level…

1. Regan Smith is very nearly as good as Kaylee McKeown

Surely no one wants to hear this argument right now, after Smith was hyped-up once again, and McKeown swept the backstroke events once again. But I actually think now is exactly the time for such an argument, that time of year when it sounds most false.

******

I can imagine that to a subset of McKeown fans the backstroke conversation always sounds like this:

Irrational American:

It’s a new year and Regan’s swimming great
I was wrong last year, but let me restate:
Greatness does not abate, nor ablate, but merely gestates,
Until the golds flow forth in a triumphant spate.
Kaylee had her time, but now she’s out of date!

Reasonable Australian:

If you would recollect, last I checked
These two great athletes have mutual respect.
Despite the stakes, there’s not the least hostility.
So pump the brakes, and have the same civility,
And, dare I ask, a smidgen of humility?
As to the picture you drew: that’s not a reason,
Since Kaylee herself is lightning-fast in-season.
You’re grasping at straws, you argue accretion–
But take a pause; let’s talk Laws: Kaylee delivers each year like fucking Santa Claus.
Every year, you forget her. But she’s gold every time; perhaps she’s just better?

Irrational American:

Aha! You’ve fallen into my trap, you poor sap.
Let us recap: in 2019, Regan was a sensation,
The new-crowned queen of a swimming generation;
Records to Smith-ereens was her occupation;
Kaylee’s got nothing on that demonstration,
So polish the throne for the Regan Restoration!

Reasonable Australian:

I’m sure you know—but 2019 was several years ago;
Compared to records now, those old times were slow.
Your story is cool, and not to be cruel,
But the last five years, Kaylee’s been taking Regan to school.
McKeown, her name ends with “win”; the rest of the field fight for scraps at her shin.
OK, fine…I’ve got a proposal,
So that maybe next time you’ll be less mucosal.
Predict what you want about the medals’ hue,
But if Kaylee again is golden times two,
Then give in: take her dominance for granted,
Until the day she’s actively supplanted.

Irrational American, Lying Even To Themself:

Of course! You’ve got a deal.
But keeping it real: that’s not going to happen since this is Regan’s year!
Our National Anthem will be a worm in your ear,
Our undead fandom a phantom of cheer.
Here’s a memorandum: the US has no peer.
You’re merely our geopolitical pawn; I’ll go on:
We sort the medal table by bronze, two rights make four wrongs, and we’ve swapped out our cowbells for gongs!

All of SwimSwam Comment Section:

USA! USA! USA! USA! USA!

******

So what does it mean that McKeown always wins?

Here’s what I think it means: you can have two evenly matched swimmers, and it just happens that one of them wins every time. That’s not a very satisfying explanation. We like to think of our sport as dispositive, the Olympic final as settling arguments once and for all.

McKeown and Smith have raced 6 times at global championships. McKeown has won 6 golds; Smith, 5 silvers and a bronze. If you flip a coin 6 times, there’s a 1/64 chance they will all land heads.

Wait…is that what I’m saying? That McKeown’s victories, nay, her unadulterated dominance over a half-decade, is the result not of sweat and tears and talent and will, but merely of crass and implausible probability? No, of course not! Well, mostly not…

Even as close as they are in ability, it’s not actually that surprising that McKeown has won every time she’s faced Smith. For one thing, they weren’t actually that close in 2021, when McKeown beat Smith once and won two golds versus one bronze. So it’s down to 5 races. And in the 200 back, despite it being the setting of Smith’s original big breakout, McKeown does seem to be just a little bit better.

Which leaves 3 races. The 50 and 100 back from 2023, and the 100 back last week. McKeown won all three races. Here are her margins of victory:

• 2023 50 back: 0.03
• 2023 100 back: 0.25
• 2024 100 back: 0.33

The 50 back was super close; the 100s a bit less so. But these weren’t the only times they raced the 100: at both meets, Smith’s relay splits were better than McKeown’s. This year, in two separate races on as big a stage as any save the Olympic final, Smith has set both the Olympic Record and World Record.

So, yeah, I’d feel comfortable saying that McKeown and Smith have been very even in the sprint backstrokes for the past two years. For a couple of years before that, McKeown had the edge, and she still does in the 200 back. (They both have excellent secondary events that I won’t compare here). And McKeown, through some combination of consistency, happenstance, and rising to big moments has won all three of those races, as she won the three before.

Two of the top several swimmers in the world swim the women’s backstroke events. One of them is slightly better (very slightly better, is my thesis), and she has won every time they’ve met in a global final. If I had to bet on their future races, yeah, I suppose I’d pick McKeown to keep winning in the absence of other evidence. But if you choose Smith to come out on top, especially in the 100, that’s hardly a stretch. More than anything, bet on the women’s backstrokes to continue as exhilarating, top-class events for years to come.

2. Clear water is overrated; drafting is underrated

For the amount of talk it gets, people around the sport think clear water matters a lot. And I understand why: when you’re caught up in the wash, it sure feels like it’s slowing you down! And watching a sprint race, particularly one like an Olympic final, it’s hard to see the swimmers for all the churn.

And so a lot of sprint strategy is premised on avoiding this hazard. Put your fastest swimmer on leadoff, get in front of the field going out. And if you’re a skinny guy swimming next to big guys, you should expect to pay a penalty in your finish time.

But I’ve got to say: I kind of don’t buy it.

To be precise, this is not to say that clear water doesn’t matter at all. All else equal, choppy water does slow you down relative to clear water. Some of those effects come from the pool; others from the race. I just don’t think the latter has quite the effect that it feels or looks like.

Why does this matter? Well, fighting for clear water means being in front. Catching a draft means being behind. You can’t really do both! And while it’s a hard question that merits real study, I believe that drafting is far more important.

That’s a difficult assertion to analyze, but we’ll try to shed a little bit of light on it, starting with drafting. Here’s a list of the fastest-ever men’s 100 freestyle relay splits, helpfully listed in a recent article by Anya Pelshaw.

ALL-TIME TOP RELAY SPLITS MEN’S 100 FREESTYLE

1. Pan Zhanle (China), 45.92 (2024)
2. Jason Lezak (USA), 46.06 (2008)
3. Duncan Scott (Great Britain), 46.14 (2019)
4. Cesar Cielo (Brazil), 46.22 (2009)
5. Alain Bernard (France), 46.26 (2009)

What do these swims have in common, besides being dramatic and memorable? Actually, that gives a clue: everyone began their leg behind a swimmer in an adjacent lane.

In fact, four out of five swims (all but Cielo) followed exactly the same pattern. Our hero dives in slightly behind a swimmer in an adjacent lane, whom they catch–in each case on the second length–and overtake. (Cielo both started and finished his split roughly two seconds behind the Americans in the next lane, so he likely got some draft, but less.)

In other words, the way most of these legs played out was almost designed to maximize the drafting benefit. And it’s borne out in the times. Cielo and Bernard owned the two fastest flat-start times in history when they swam in the same 2009 medley relay. Their splits in that relay were of at least World Record quality, roughly 0.7 seconds faster than their flat-start times. Pan’s differential was slightly smaller, 0.48 seconds, but still close to the quality of his immense World Record*. Meanwhile, Scott and Lezak each outswam their best flat-start swims by a stratospheric** 1.5 seconds, far more than what the differential “should be” under any accounting.

*See below

**Also see below

I don’t want to belabor this point because anyone who’s ever sandbagged at the back of the lane during practice knows that drafting makes you faster. But it deserves emphasis because none of these swimmers had anything close to clear water. If that’s what mattered most, we would expect the list of fastest splits to be made up of swimmers who started with a lead. Instead, we see a bunch of swimmers who fought through the chop and caught a wave.

******

Let’s examine the benefits of clear water more directly. It’s not easy to investigate since there’s no obvious counterfactual. We can see that drafting helps turn Lezak and Scott superhumans by comparing to their flat-start times. But there’s no similarly good comparison here without the ability to make everyone swim a solitary time-trial.

In addition, what do we actually mean when we say that choppy water slows you down? Which swimmers? And in which races? The specific way in which you expect this to matter makes a big difference for what you look at to suss out the effect. Therefore, nothing I can say here will rule out a big effect that I just didn’t look for correctly. Let’s say this: if you think clear water has a big effect on race times, it’s on you to explain exactly how you expect it to manifest, and then find it.

So with that in mind, here’s the idea as I most often hear it: choppy water is caused by big guys sprinting fast. It’s worse on turns, and worse for stragglers than leaders. As a result, the race most affected is the men’s 100 freestyle, since it’s the fastest race with a turn. In the Olympic final, you have the biggest, fastest, splashiest sprinters, and they all drown each other in churn. If you’re a smaller guy in that race, or if you fall behind on the first length, you’ll lose a lot of time.

Given that, let’s look at the Olympic final of different men’s freestyle events. If the 100 freestyle was surprisingly slow, or if there’s another pattern that seems to correlate with the amount of wash, that will be evidence for the clear water hypothesis.

Here are the average Olympic final times for the men’s 50-400m freestyle events relative to each swimmer’s entry time:

From seed times to Olympic final:

• 50m: 0.15% slower
• 100m: 0.65% slower
• 200m: 0.71% slower
• 400m: 0.40% slower

And here they are relative to each swimmer’s semifinal (or prelims for the 400m) time:

From previous round to Olympic final:

• 50m: 0.05% faster
• 100m: 0.19% faster
• 200m: 0.06% faster
• 400m: 0.54% faster

The final times were slower on average than the entry times, but faster than the semifinal times. Notice that no particular event jumps out. The 100 was one of the slowest events relative to seed, but the 200 was ever so slightly slower. And how do you explain why the 50 had by far the least slowdown? Maybe you really need turns to see the effect?

And a semis-to-final slowdown fails to materialize entirely. In fact, not only was the 100 final slightly faster than the semifinals, it slightly outpaced both the 50 and 200. If the finalists are out front during their semifinals but caught in the wash in the final itself, we’re not seeing any evidence of that here.

Again, this isn’t a knock-down disproof of any clear-water effect–for one thing, it’s only one meet! Maybe choppy water affects all events equally? Or maybe the effect is subtle, and “drowned out” by other variance in the data?

But while the drafting effect screamed out from the first data we looked at, the clear water effect, if it exists, is harder to find.

3. Like every meet for a decade, Katie Ledecky has an argument as the best

Let’s define a new term: a “WRINSS” (World Record If Not for Self or Supersuits) is a swim that either

• Is itself a World Record, or
• Is only beaten on the all-time list by supersuit swims and other races by the same swimmer.

(One could choose not to exclude supersuits, which I guess would be a WRINS.)

This may seem pretty arbitrary, but it’s actually one of my go-to ways to compare swims. We can call this the “Coconut Tree mentality”*: ask where this performance would rank in a world where the swimmer in question has just fallen out of a tree, unburdened by past performances, and swims this time in their first race? A WRINSS means you are at the pinnacle: no other swimmer has ever been better. While any metric is only partially-illuminating, it’s a good way not to penalize athletes for their prior excellence.

*To paraphrase the Vice President

So who WRINSS-ed the competition (get it?) in individual events in Paris?

• Katie Ledecky: 2 events – 1500 free (0.94%) and 800 free (0.07%)
• Leon Marchand: 1 event – 400 IM (1.09%)
• Pan Zhanle: 1 event – 100 free (0.98%)
• Summer McIntosh: 1 event – 200 fly (0.65%)
• Sarah Sjostrom: 1 event – 50 free (0.34%)
• Gretchen Walsh: 1 event – 100 fly (0.18%)
• Regan Smith: 1 event – 100 back (0.09%)
• Bobby Finke: 1 event – 1500 free (0.04%)

The percentages represent the gap back to the next-fastest (non-self, non-supersuit) time in history.

It’s a pretty good list: Marchand, Pan, and McIntosh are right near the top, as they should be. It’s interesting to see Walsh and Smith on there when they didn’t win those events. And there’s a distinct lack of Australians: McKeown, Titmus, and O’Callaghan have many past WRINSSes between them, but for various reasons didn’t quite get there this time.

And Ledecky is the only swimmer with two WRINSS-es, including the third best single-event WRINSS. Lest you think this stat was engineered to be maximally in Ledecky’s favor: if we kept supersuit races, Ledecky’s numbers would be unchanged, but Marchand would move well down the list and McIntosh would fall off entirely. I believe in this version, though.

Now, there are certainly cases against such an approach, and one can easily argue, probably correctly, that other swimmers had a better meet. I’m less interested in litigating that war and more interested in asking: how does a swimmer whose times were well off her best, who few people would rank in the top 5 of the meet, somehow still end up at the top of a list like this one?

One reason we might underrate Ledecky is that we’re not weather balloons: we can’t always tell the stratosphere from the troposphere. Maybe the Ledecky of 2016 was alone in the exosphere and has since descended to the ionosphere*, but is still flying higher than merely-stratospheric swimmers in her events and others.

*Why yes, I did look up atmospheric layers for this essay. Why do you ask?

While there’s a fair bit of nuance, I do think this is the answer. There’s some, vaguely-defined, absolute level of performance. And while Ledecky’s “exospheric” swims over the past decade plus have completely rewritten the record book in her events, the absolute levels haven’t changed. The challenge of beating Ledecky last week was precisely to swim faster than 15:30 and 8:11, a standard no other swimmer has ever met. And so, Ledecky remains unbeatable at long-distance to this day, with margins that compare well with superstars throughout the sport.

(We’ll leave this for now, but I think a few things that seem outlier-ish about Ledecky–her range, her pacing, her aging curve relative to other distance swimmers–appear much more “in line” if you consider 2016 Ledecky to have been a level above even the best of the best).

4. Bobby Finke is a distance swimmer

Yes, the header is a bit of a joke: Finke just set the 1500m World Record, we do know he’s a distance swimmer!

But it seems Finke’s race strategy has partly obscured this. Until this most recent race, his usual style on the international stage has been to stalk the leaders, biding his time and pouncing on the final length to take the win (or come close). I mean, with that finishing kick, isn’t he at least sprint-y relative to other distance swimmers?

I think no. For one thing, Finke is by no means the only distance swimmer to have fast finishes, even in PB-type races. A final split of 26-point and even 25-point was done before Finke debuted his strategy, and by several more since. Certainly, some are better finishers than others, and Paltrinieri in particular doesn’t seem well-suited to this method.

But more to the point, why look at strategy when we can look at results? Finke’s results have always been far better in the distance events; even one step down, his 400m PB is only 3:46.27 (interestingly, Paltrinieri has done nearly the exact same time: 3:46.29). And last year, Finke swam a near-world-record of 14:31 in the 1500m for silver, while his 800m was “only” 7:38 and bronze.

Actually, let’s go back to Finke’s double-triumph in Tokyo 2021. His 800m shocked the field, closing in 26.39 to win a surprising gold. His final split in the 1500m later that week? An astonishing 25.78, and quite a bit larger winning margin. Kind of seems like he had more left in the latter race, doesn’t it?

To state the obvious: any distance swimmer in an Olympic final can close in 25-26 seconds if the pace is slow enough. The hard part is the first 1450 meters.

(One last thing: you might think Finke broke through and got the World Record because he finally decided to push the pace. That’s not entirely false, but not entirely true either. Finke’s opening splits on Sunday were just slightly faster than his opening splits in last year’s final, about a second and a half total over the first 400, a gap which mostly held all the way to the finish. Finke’s closing 50 in his World Record swim? 26.27.)

5. Pan Zhanle’s 100 free wasn’t that much of an outlier

I say this not to diminish the clear swim of these Games, one of the truly great performances in our sport. But I’ve heard some people try to cast suspicion by framing Pan’s swim as an impossible outlier. No, it’s merely an impossible non-outlier, on par with other famous impossible swims done by beloved athletes around the world.

The best comparison for Pan’s race from a dramatic perspective might be a 100m race from another sport, perhaps the most famous Olympic race of them all. I’m talking about Usain Bolt’s 100m dash in Beijing. You had the same set-up: a young stud, already the world record holder, but can he perform in an Olympic final? And then he starts out well, he pulls ahead, he pulls away, he pulls awwaaayyyy. And then…whoa! The clock. The future is present, that present is past, just like that. The only thing Pan didn’t do was celebrate the last 10 meters, but he certainly would have had time for it.

Meanwhile, the rest of the field was just swimming a different race, as if two videos were superimposed, but slightly off. When Ledecky wins big–or Titmus, McIntosh, Marchand–the field is strung out behind them like cars on a desert highway. But when Pan does it (or Sjostrom), it’s a single headliner followed at a respectful distance by a synchronized line of backup dancers.

And in Pan’s case this was especially striking because of the blanket finish behind him. The gap from Chalmers to Popovici? 0.01 seconds. From Popovici to Nemeth? Also 0.01 seconds. The gap all the way from Chalmers in second to Giuliano in 8th? Exactly half a second. What do you expect–it’s an Olympic 100m final! But the gap from Pan to Chalmers? A full 1.08 seconds, plus some relativistic time-dilation.

Back to the thesis, which is quite simple. Pan’s new World Record is 0.98% faster than anyone else has ever swum (and next on the list is Popovici, if anything more of a 200-meter swimmer).

Meanwhile,

• Sjostrom, 50 fly: 2.48% ahead of 2nd on the all-time list
• Peaty, 100 breast: 1.40% ahead, and much larger when he set it

And there are several other events with a gap as large or larger than Pan’s.

Similarly, Pan won the race by the massive margin of 2.27%. However,

• Marchand, 400 IM: 2.28% ahead of silver
• McIntosh, 400 IM: 2.08% ahead

Pan stacks up very well here! It was the swim of the meet, after all. But he was roughly in line with some of the other very best swims from recent history. It was a swim to remember, a swim to redefine the 100m freestyle, redefine the sport. But not break it.