On Sunday, from the moment 30-year-old Dennis Kimetto breasted the tape at the Berlin Marathon at 2:02:57 to set a new world record by 26 seconds, conversation in certain circles turned to what this meant for the assault on a sub-two-hour marathon.
There are actually two separate issues here. The first is a matter of human evolution: Can the species produce some physiologically absurd outlier, a la Usain Bolt, someone so many standard deviations off the norm that he upends distance running's traditional calculus for speed, intensity, and efficiency? And the answer is sure, maybe, why not. The second is a matter of the sport's evolution: How long it will take for the pack of elites, or at least the select few who have the potential to break a world record, to inch gradually to the point where two hours is within range? They have 177 seconds to go.
The first thing to cover is that 177 seconds is an enormous amount of time. To a casual runner, it doesn't sound like a whole lot, just under seven seconds per mile over the course of a marathon. You can drop or gain whole minutes from mile to mile in a 10k. But at elite marathon pace, already run at sub-five-minute speed, asking for even a few seconds per mile can break a race, or a runner.
Here's a chart that shows the progression of the marathon over the last 100-odd years:
The chart comes from a New York Times article by Justin Wolfers, which points out that the world record has been broken six times in the past 12 years by an average of 27 seconds. Indeed, if the next dozen years keeps pace with the past dozen, the world record will be lowered by around half a minute another half-dozen times—which will be enough to break the two-hour barrier.
But the past 12 years of marathon records don't exist in a vacuum. It's true that records have fallen at an increased clip, but that progress was in large part due to the boom of 10,000m runners in the '90s and the subsequent conversion of those runners into long-distance runners. Ross Tucker of the Sports Science Institute of South Africa laid this out in a post from 2010.
So now, why is the relative stagnation of 10,000m times important to this debate about the marathon?
The reason is that the golden-era of 10,000m running in the 1990s, where the records fell so regularly from 1993 to 1998, is one of the main factors responsible for driving the drop in marathon performances in the 1990s and 2000s. Regardless of the factors that drove that spate of records in the 1990s, it was the conversion of these runners to the longer distance that helped to lower the marathon world record and raise the standard to a point where a 2:05 now longer even guarantees a win.
The reality is that there is a strong correlation between the speed an athlete possesses over 10,000m and their ability to run a fast marathon. It's not perfect, otherwise the best 10,000m runners would win marathons all the time (Zersenay Tadese, for example, has not converted his amazing 10,000m and half-marathon achievements yet).
There's a basic conversion of speed from one event, like a 10K, to another, like the marathon. The training will change, but you can adjust and extrapolate how fast a runner can complete a longer race by how fast he can complete one of its component pieces. And right now, progress in the half-marathon is not close to a projected 57-minute benchmark that would line up with a two-hour marathon (the world record for the half has been 58:23 since 2010). The 10K meanwhile has flatlined since the heyday of the '90s:
But this is the actuarial case against the imminence of a two-hour marathon. And the conditions that set these tends into motion—the shift of elite track stars into marathoning in their later years—are changing.
"The stigma for marathoners," says David Epstein, author of The Sports Gene, "was that it's what you do after you lose your track speed. But the money moved away, and guys skipped track careers, all of a sudden guys were in marathon in their prime. ... Kimetto is a little unusual—the last few years the age of top finishers in top marathoners was coming down every year."
Not that the influx has made the prospect of breaking two hours much more likely. "If someone shatters the 10k record in 26-flat," Epstein says, "I'd say, OK, maybe this is a guy who can run a two-hour marathon. But the record is 26:17, and hasn't moved since 2005." Meanwhile, the 10K equivalent to a two-hour marathon is 25:35. Not very encouraging.
But let's say a glitch in the system produced a runner so physiologically aberrant that he overcomes the trend lines. What would he look like?
"You'd start with really long legs," says Epstein, "a giant chest cavity with huge lungs with surface area to put a lot of oxygen into the blood, a stiff and long Achilles tendon, a massively strong heart, tremendous blood volume." And then you come to finer points, like distal elongation (where the lower legs are even longer than the upper legs), or tiny calves and ankles that don't weigh down the pendulum-like motion of running.
A lot of this is deterministic. If you have a lot of blood—or even just a lot of plasma—your body can deliver more oxygen; this influences aerobic capacity, after all. Or perhaps you have blood that is abnormally efficient at delivering oxygen. An athlete's body will wind up with any number of useful peculiarities, and the ideal runner would need just about all of them to break his way. Thing is, we don't have the best grasp on which way that is.
A Japanese study of Kenyan runners, for example, noted that that they have exceptionally long and elastic ("stiff") Achilles tendons. The Achilles basically functions as a spring in the ankle, and the longer it is, the more energy it can store. The study found that the Kenyans' Achilles were much more elastic than non-athletic whites', which they measured through a double leg-hopping exercise. The Kenyans performed far better, but of course they did—they were elite athletes matched against guys off the street, basically. Ideally, you'd match them with elite European runners, but we don't have anyone doing that study yet.
The information we do have is suggestive, though. Gary Hunter at the University of Alabama at Birmingham has conducted multiple studies showing that people with longer tendons were more economical using energy when walking on both flat terrain and on an incline. It doesn't translate directly to running an elite marathon, but it's close enough to figure it matters, and it's one more thing that has to break exactly right for our ideal runner, whichever way "right" ends up being.
There's room for spontaneous physiological divergence in a lot of these cases, but in some instances, we're bumping right up against the limits of the human body. How thin can a lower leg get before it's structurally unsound? "I don't think it's known," Epstein says, "but if you look at the lower legs of the best marathoners, you have to have a bone and an Achilles, and there's not room for much else. We aren't too far out."
Within endurance sports, there are two measurements that mark off the capability of an athlete—VO2max and economy. You've probably heard of VO2, in relation to Lance Armstrong's hot-air balloon lungs or the limits of human breath-holding . Essentially, it's the maximum capacity of oxygen your body can store. Running economy, in turn, is how much oxygen your body requires to maintain a certain speed while carrying a certain amount of mass (your body weight) over a certain distance. Economy is measured in milliliters per kilogram per kilometer (for example, 200 ml/kg/km), while VO2max is measured in milliliters per kilogram per minute (e.g. 70 ml/kg/min).
To break two hours in a marathon, you'd need an exceptional combination of the two. Ross Tucker has crunched the numbers, extrapolating the average measured intensity that runners can keep up when running given distances:
An elite athlete is usually able to sustain running speeds that require about 85% – 90% of VO2max for about one hour. Marathons are usually run at an intensity corresponding to approximately 80% of VO2max, while 10,000m is run at around 95% of VO2max. So for a two-hour marathon, we can make the assumption that the intensity will be equal to between 80% and 85% of maximum.
From there, you can suss out how big of a VO2max you'd need to combine with how tight an efficiency and get a very rough sketch of the bare minimum to reach two hours. Here's the chart that Tucker came up with:
The two white lines on the chart are the observed running efficiencies of African runners (example A) and European runners (example B). The blue line shows how much VO2 would actually be used at a two-hour pace for a given running economy, while the other two lines show the requirements for VO2max. A runner with an elite efficiency would still need a VO2max north of 80 to drill down to two-hour pace. A more common efficiency pushes that up north of 90. In case you're wondering, yes, those are gigantic numbers.
"Think of a car with big engine that's also fuel efficient," Epstein says. "It doesn't happen in cars, basically, and it looks like it's similar with runners. Guys with a really high max tend to not have the best economy. The reasons aren't 100 percent known, but it has to do with the fact that when you get bigger you have bigger lungs, for example, and more surface area to diffuse oxygen, but big isn't good for economy. People who have the highest in one or the other don't tend to come in the same body."
In many ways, Dennis Kimetto is an aberration. He's a 30-year-old in a sport that's getting younger and younger at the elite level. He began running competitively just a few years ago, having more or less been dragged out of a shopping mall on a lark to begin training. His world-record marathon was a shock in just about every way except for one: Like each of the past six world records, it was run in Berlin.
Course conditions, layout, and a field packed with peaking contenders are just as much a prerequisite as exceptional athletic ability. And it's just as constraining, as well. Even before the recent ubiquity of the Berlin marathon, there were maybe a half-dozen courses in the world, max, that could feasibly host a world-record attempt. And even then, many are finicky, like London, which features a home stretch given to slight headwinds—more than enough to tack on a few seconds per mile. Further complicating things are prize pools that are perhaps too big to get competitors to focus on their times, since they'll focus on beating their competition, at any time.
Stack all this on top of itself and the handful of elite runners who are even in the conversation will get just a handful of clean shots at breaking the record. This is a big deal, since breaking any record is largely about exposing it to enough legitimate attempts. This is true of marathons just the same as it's true of, say, NFL field goals .
There are a few other factors that could push times down a little further.
One is the nebulous mix of science, pseudo-science, and infomercial pitch known as "sports science." Here we're talking less about the gimmicky stuff related to fitness tracking—though Epstein has a funny story about his time in East Africa, when the only heart-rate monitor (and he is the first to say that heart rate monitors are largely silly) was strung up between two trees, holding up the laundry with its straps.
Instead, we're talking medical science that we largely take for granted here. Elite runners will win a race and then drop off the face of the planet, over something as minor as tendinitis. Extending the careers of great runners, or maybe broadening out their primes, or at the very least making sure they don't retire over some shit like plantar fasciitis, would go a long way toward adding to the number of attempts on the record.
And then there's pacing. It's weird, since your high school cross-country coach likely pounded the virtues of a properly paced race into your brain in between every vomit-drenched interval, but as records have fallen, an "ideal" pace to a record-setting marathon has never arisen. This isn't the case for other events.
In 2010, Tim Noakes analyzed 32 world-record runs in the mile, and found that a clear pattern had emerged. Almost universally, the third lap was the slowest, followed by the second, and the first and last were the fastest. Here are a few charts from the paper:
The "end spurt" being present and successful in this many world records doesn't necessarily mean it's the best approach, but it's the one that has worked consistently. And this just doesn't exist for marathoners.
"Given Kimetto's 5K splits," Epstein wrote in an email, "I have a hard time believing that's perfect pacing, dropping a 14:10 from 30-35k, so I bet there's some improvement to be had there somewhere: 14:42, 14:42, 14:46, 14:25, 14:33, 14:29, 14:10, 14:42." Those are each of the successive 5Ks making up Kimetto's 2:07:57 time. Blasting out to 14:10 in the second-to-last 5K is what really stands apart, and it runs counter to the pattern for the mile records. Epstein's best guess is that's a move you make to break your opponents, not to break a record.
A world record is a ridiculous thing to expect under any circumstances. If a breeze rolls in off the water at just the wrong angle, or if the wrong low-pressure system pushes the wrong clouds in the wrong direction, or if your nose is running, or if your shoe comes untied, your shot at history blinks away, just like that.
Now take all of the physiological coincidences that go into a world record, all the stars aligning on the course and in training and in the pace of the field, and squeeze them. That little gust of headwind on the London home stretch matters that much more at 2:00:01 than it did two minutes slower. At the very edges of human capacity, fewer and fewer things can turn out less than perfect and keep a record attempt intact. So maybe there's something to the 15-seconds-per-year projections, or 10 seconds, or 20. But the preponderance of escalating complications on the way will make each step harder than the last.
"After the four-minute mile," Epstein says, "people were like, 'The limit is nothing!' The record was broken eight times a decade, then hasn't moved at all. ... Kimmeto would have been a 10k record holder, but there's a long way to go."