Sweat Science

What does the evidence show?

Are Marathoners Getting Faster or Is It the Shoes?

Plotting top-100 marathon times for the past decade shows that Eliud Kipchoge’s world record is a huge outlier. The question is why.

Was it Kipchoge, or was it the shoes? (Markus Schreiber/AP)
Photo: Markus Schreiber/AP marathon

When you buy something using the retail links in our stories, we may earn a small commission. Outside does not accept money for editorial gear reviews. Read more about our policy.

Plotting top-100 marathon times for the past decade shows that Eliud Kipchoge’s world record is a huge outlier. The question is why.

For better or worse, one of the biggest questions of 2018 for running fans is: was it Eliud, or was it the shoes? Eliud Kipchoge’s insane marathon world record of 2:01:39 at the Berlin Marathon in September made him the fastest man in history by over a minute. But he was wearing Nike’s controversial Vaporfly shoes, introduced last year, which feature a thick layer of foam and a curved carbon fiber plate that reportedly make runners four percent more efficient on average. So is Kipchoge really any faster than previous record holders?

There’s no simple answer to this question. But one way to think about it is to ask what the broader trends are in marathon running. Are times getting faster overall? Even if only a fraction of elite marathoners—say a quarter of them—are wearing new supershoes, then those runners will leapfrog past non-Nike runners and push them further down the rankings. That should produce a ripple effect that makes the 10th or 50th or 100th fastest marathoner of the year faster than in previous years, regardless of what shoe he or she is wearing.

So, with that in mind, I’ve plotted the top 100 marathon times for men and women each year for the past ten years, drawing on data from the IAAF and including only the fastest time for each runner. Bear in mind that 2018 isn’t over yet, so it’s still theoretically possible that a few new performances might squeeze into the top 100 (the data is as of Dec. 10).

Here’s the men’s data:

marathon
(Alex Hutchinson)

And here’s the women’s data:

marathon
(Alex Hutchinson)

Does the appearance of the first Vaporfly prototypes in early 2016, or their general release in July 2017, bend the curve? Admittedly, these plots are basically glorified Rorschach tests: you (and I) can see whatever you want in them. But there are a few things worth noting.

Looking at the broad patterns (as opposed to the big jumps in the top few individuals), there’s no doubt that 2018 is faster than 2017, which in turn was faster than 2016. But the improvements aren’t noticeably bigger than the fluctuations in annual performance in previous years.

In fact, the most obvious feature that’s present in both men’s and women’s data is a steady improvement leading up to a peak year in 2012, followed by a few years of declining performances. Then things start getting better again sometime between 2014 and 2016 (depending on where you look), leading up to the current streak of improvement. What happened in 2012? I’m not sure. I asked a few running-expert friends, and they couldn’t reach any consensus. It’s true that 2012 was an Olympic year, which generally brings out the best in people. But the same bump doesn’t show up in 2016. My leading theory, I’m very sorry to say, is that 2012 was when the first doping positives under the IAAF’s new biological passport began showing up. Perhaps that led to a subsequent chill in performances—at least temporarily.

Is the general improvement in performance in 2018 enough to raise a red flag? For comparison’s sake, I’ve plotted similar top-100 data for a swimming event (I picked the men’s 100 freestyle). Swimming’s results were famously skewed by the introduction of ultra-fast textile suits like Speedo’s LZR in early 2008; those suits were banned from 2010 onward. Here’s what that data looks like, drawing from the FINA’s official rankings:

marathon
(Alex Hutchinson)

The thing I should highlight is that I’m not confident in the completeness of the rankings from the earliest years. FINA’s website showcases rankings from 2008 onward, but the earlier data is available in archived form. So some of the apparent improvement between 2005 and 2008, particularly outside the very top performances, may simply be a result of more complete data. That said, the LZR swimsuit didn’t emerge out of nowhere: there was steady improvement in swimsuit technology leading up to the big dip in 2008 and 2009.

What’s unambiguous is that the introduction of new swimsuit rules in 2010 led to a slowdown across the board. At the time, many commentators (including me) wrote about the prospect of swim records that would never be broken, thanks to the distorting effects of the now-banned suits. Curiously, that’s not what happened. Overall top-100 performances are faster now than they were in 2009. Of the 20 official men’s world records, eight have been set since 2010; 16 of the 20 women’s records are post-2010. There are many possible explanations, like faster turns and hyper-optimized swimming pools. Regulating swimsuits didn’t stop the tech-aided march of swimming progress; nor did it destroy the sport.

There are some key differences between swimming and running, most notably in the greater role of drag underwater, and the consequent importance of subtle changes in technique. Swimming’s records have always fallen more frequently than running’s records. But there’s one point that stands out to me in the graph of swim rankings. In 2008 and 2009, everyone seemed to get faster. The top swimmers were no further ahead of their competitors than they were before or after.

In contrast, take another look at where Kipchoge’s new world record sits in the marathon graph. He’s way out of sight relative to his competitors. Lots of top runners are wearing the Vaporfly. In 2017, these runners took 19 of 36 podium spots at the six World Marathon Majors. But none of them are anywhere near Kipchoge. So it can’t all be the shoe.

The marathon, of course, is uniquely challenging. Kipchoge’s record run was his fifth official marathon in the Vaporfly, but the first when fitness, course, and weather conditions combined to give him the opportunity to shatter the record. Maybe there are other runners who will follow Kipchoge into 2:01 or at least 2:02 territory when the stars line up for them, and in doing so demonstrate that his run in Berlin wasn’t as special as it appeared. After all, by the end of 2016, prior to the Vaporfly’s release, 31 men had run 2:04:59 or better. If the Vaporfly really alters performance by several percent, we should see someone threaten Kipchoge’s record soon.

The tough question is whether we can afford to wait. If, in five years, a half-dozen men all wearing the Vaporfly have run 2:01, we’ll have our answer. But by then it may be too late to convince authorities to regulate shoe technology—if it’s not already.


My new book, Endure: Mind, Body, and the Curiously Elastic Limits of Human Performance, with a foreword by Malcolm Gladwell, is now available. For more, join me on Twitter and Facebook, and sign up for the Sweat Science email newsletter.

Filed To: Marathon / Running / Science / Running Shoes / Swimming