Analysis: Ledecky vs. Sjostrom, Which is the Greatest Women’s World Record?

Courtesy: Daniel Takata

A few days ago, I presented a statistical analysis of Adam Peaty‘s 100 breaststroke world record.

I used Extreme Value Theory, which is a branch of statistics which deals with extreme values, or rare values.

Using the times of the 150 best performers in each event, I modeled the data and concluded that Peaty’s world record is the more extreme among all men’s world records.

In other words, Adam Peaty’s 56.88 in 100 breast is the greatest current world record in men’s events.

And, to match this time in other events, it would take 45.71 in the 100 freestyle, 3:35.59 in the 400 freestyle, 50.49 in the 100 backstroke, 2:02.54 in the 200 breaststroke, and 3:59.90 in the 400 individual medley, for example.

But what about women’s events?

Some say that Katie Ledecky‘s 800 and 1500 world records are the greatest among women’s records.

In the 800, Ledecky’s world record is 8:04.79, and no one else has ever swam under 8:14.10 in the event.

In the 1500, the world record is 15:20.48, and the second performer ever is at 15:38.88.

Others say that no record can match Sarah Sjostrom‘s 50 butterfly.

Her world record is 24.43, and no other woman has ever swam under 25.07.

Using Extreme Value Theory, I can compare these records not only with the second best performer of all time, but with a population of world class swimmers, and determine how extreme these records are.

I took the times of the 150 best performers in each event and modeled the data in order to determine the probability of these world records be exceeded.

We can determine the theoretic probability distribution of such data – this distribution is called Generalized Pareto distribution.

We now can evaluate the probability of Ledecky’s 800 world record being exceeded in the universe of the 150 best performers of all time. This probability equals to 0.465%.

Ledecky’s 1500 world record probability is 0.301%.

And Sjostrom’s 50 fly world record probability is 0.593%.

For illustration, the probability of Katinka Hosszu‘s world record in the 400 individual medley being exceeded equals to 1.151%.

And Regan Smith‘s 200 backstroke is 0.891%.

In these terms, the 1500 free world record is the greatest among the current women’s world record in swimming.

And what would be the time in the 100 freestyle, or in any other event, to match Ledecky’s probability in the 1500?

The times are listed below (women’s events only).

It would take 50.72 in the 100 freestyle.

3:54.69 in the 400 freestyle.

2:01.80 in the 200 backstroke.

1:59.52 in the 200 butterfly.

And 4:22.16 in the 400 individual medley.

Unfortunately Katie Ledecky couldn’t swim the 1500 in the last World Championships.

But her world record stands as the greatest of women’s events.

And certainly one of the greatest of all time.

One could ask, how can we compare Ledecky’s record with Adam Peaty’s 100 breast record?

In fact, the probability of Peaty’s world record being exceeded in the universe of the 150 best performers of all time equals to 0.291%.

Remember the probability associated with Ledecky’s record is 0.301%), which means that Peaty’s record is a little bit more extreme than Ledecky’s.

Hence Peaty’s world record is the greatest current world record, among all men and women’s swimming events. But just by an inch.

To match Peaty’s record in women’s 1500 freestyle, it would take 15:20.09 – very close to the actual world record of 15:20.48.

In the following table we can see the times that would match Peaty’s record in every swimming event. Note that the times in women’s events are almost as fast as the times that would match Ledecky’s record, which means that Peaty’s 100 breast world record is almost as extreme as Ledecky’s 1500 world record. These two records are truly in the same difficulty level.

Women’s events

Men’s events

About Daniel Takata Gomes

Daniel Takata is editor of Swim Channel Magazine from Brazil, has a PhD in Statistics and is a college professor. He also works as a swimming TV commentator on SporTV and holds a website dedicated to sports analytics, Esportístico (www.esportistico.com.br).