Lessons from Watching the World’s Best Runners

An in-depth biomechanical analysis of nearly everyone at the world track championships yields unexpected results

At last year’s World Track and Field Championships in London, researchers had 49 high-speed cameras rigged up in the stadium and out on the marathon course for a massive biomechanics study led by a team at Leeds Beckett University. If you’ve ever wondered about the nuances of hammer throwing—should you spin three, four, or five times before releasing your projectile?—then the recent release of detailed reports on 38 individual events will be a treasure trove for you.

Among those reports are some interesting nuggets on the best distance runners in the world. During the marathons, which were run on a four-loop road course, the researchers used a four-camera setup to watch for subtle changes in each runner’s stride, foot strike, joint angles, and so on as they fatigued. During track races like the 10,000 meters, they analyzed the pacing of each runner in 100-meter intervals throughout the race in addition to collecting biomechanical data. Here are a few highlights from their findings:

The heel strikes back: One of the main tenets of the barefoot/minimalist running movement is that landing on your heel is bad for you. And it’s true that studies of people who grow up without shoes in places like Kenya show that they tend to land on their forefoot or midfoot when they run barefoot. But what about the elite runners who emerge from East Africa? Do they continue to run that way once they have access to shoes?

Of the 70 runners analyzed during the fourth lap of the men’s marathon, 47 of them (67 percent) landed on their heels, 21 (30 percent) landed on their midfoot, and two (3 percent) landed on their forefoot. The percentages were similar in the 78 competitors analyzed in the women’s marathon: 57 (73 percent) heel, 19 (24 percent) midfoot, and 2 (3 percent) forefoot. And this pattern wasn’t confined to particular countries or finishing places. In the men’s marathon, the top four finishers, hailing from Kenya, Ethiopia, Tanzania, and Great Britain, were all heel strikers during all four laps of the race.

This is particularly interesting because the idea that elite runners don’t land on their heels is so pervasive. When Nike was developing shoe concepts for its sub-two-hour marathon project, they at one point produced a prototype with the heel mostly stripped off to save weight. The problem was that virtually all the elite runners who tried it hated it, so they ended up pivoting to the chunky, cushioned heel in the new Vaporfly 4% shoe.

Of course, there’s a bit of a chicken-and-egg problem here: do the runners prefer shoes with cushioned heels because they land on their heels, or do they land on their heels because it’s all but unavoidable once you’re wearing a bulky cushioned shoe? This study can’t answer that question, and by no means should we now conclude that heel striking is “better.” But the claim that the best runners in the world land on their toes—and the corollary that you need to learn to do the same if you want to maximize your performance—should be abandoned in the face of the new data.

Fatigue changes your form: One of the complicating factors in discussing foot strike is that it depends what part of the race you’re talking about. Rose Chelimo, the Kenyan-born runner representing Bahrain who won the marathon gold medal, was a heel striker for the first two laps of the race, but switched to a midfoot strike for the third and fourth laps as the pace accelerated. In the men’s race, eighth-place finisher Daniel Wanjiru did the same, even though he was decelerating rather than accelerating during the second half.

There are some more subtle nuances to fatigue-induced form changes, too. In a commentary accompanying the study, elite running coach Steve Magness points out that several runners start to exhibit asymmetries in the fourth and final lap of the marathon. Fifth-place finisher Gideon Kipketer, for example, had a nearly symmetric stride in the third lap, with only a 1-centimeter difference between his right and left stride lengths. In the fourth lap, the difference ballooned to 9 centimeters. Similarly, seventh-place finisher Yohanes Ghebregergis’s asymmetry jumped from 6 to 11 centimeters. “As athletes fatigue,” Magness writes, “they start to compensate and their ‘weak links’ in the chain start to show.”

Another example is cadence (how many steps per minute you take) and stride length. Most athletes increase both these parameters when they accelerate and decrease both when they slow down. Men’s silver medalist Tamirat Tola, on the other hand, faded after the 35K mark almost entirely due to a 14-percent decrease in stride length, with little change in cadence. That’s unusual.

These sorts of insights don’t lend themselves to big generalizable conclusions—on the contrary, they remind us just how unique each person’s response to fatigue is. But the rise of wearables is making this sort of analysis much more widely accessible. Someone on Twitter recently asked me about data they collected showing an increasingly big left-right asymmetry in their stride that starts after about 20 minutes of running. While it’s hard to know exactly what causes it, this is a pretty clear signature of something going wrong—and it will be easy to see if it’s getting better after trying things like stretching, strengthening, or stride alterations.

The mirror may lie: For runners like Kipketer (or my Twitter correspondent), asymmetry may only emerge with fatigue. But some runners have significant asymmetries right from the start. Almaz Ayana, the 10,000-meter gold medalist and world record-holder from Ethiopia, had the biggest stride asymmetry of anyone in the study—and it was apparent right from the start. After 5 and 10 laps of the 25-lap race, one stride was 17 centimeters longer than the other; by the last lap, the difference was 20 centimeters. That’s massive.

The big question is what you should do with this data. After all, Ayana is already the best 10,000-meter runner in history. Are we really sure that “fixing” her stride would make her better? The finding is reminiscent of the news last year that Usain Bolt apparently also has an unusually asymmetric stride, exerting 13 percent more peak force with his right leg than with his left leg. That may result from a slight discrepancy in leg length, which means he has optimized his form for his unique body. “In other words,” Southern Methodist University researcher Peter Weyand told the New York Times, “correcting his asymmetry would not speed him up and might even slow him down.”

That caveat is something to bear in mind for all the data collected at the World Championships last year—and for all the wearable data that runners around the world are gathering on themselves. If you want to tweak your own biomechanics, the first step is to collect a bunch of baseline data to make sure you understand what’s normal for you. Then make changes slowly and cautiously, allowing plenty of time to see if it’s having the effect you’d hoped and watching for any undesired side effects. And remember that, like Bolt and Ayana, a seemingly unusual stride trait might be optimal for you.

Oh, and if you’re still wondering: the theoretical optimal release angle for throwing a hammer is 45 degrees. In the 2017 World Championship final, the average release angle for the best throws of the 12 men’s finalists was 41.3 degrees. Only one man, Poland’s Pawel Fajdek, exceeded 45 degrees with 46.2-degree release angle. He won the gold.

---

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.