You know how your phone buzzes with 8 notifications when some world-changing event happens? My phone is like that too, but when a new training summary gets published.
And on August 29, my phone did an all-day impersonation of a vibrator. I gotta say it was a good day. Those good vibrations were from athletes sending Kilian Jornet’s UTMB training and race summary published by watch-maker Coros. Read the full summary here. Buy a Coros watch too. Their marketing department earned it.
Kilian is the Greatest Of All Time, and this summary of GOAT training is too exciting to pass up. But before analyzing the data, we need to lay down some hot disclaimer beats.
The first disclaimer is the biggest one: I am unsure if the Coros data reflects all of Kilian’s training sessions. The public version of his Strava is only a small snapshot. The screenshot of his training intensities seems to indicate lower volume than he does at other times of the year, but I am not a Coros expert so could be misreading that. In this article, I’m operating from the assumption that the data Coros published is at least a semi-representative sample of intensity distribution, and I will not be talking about training volume in case it’s not the full training load. Either way, I will try to broaden out the takeaways with research studies so they are helpful even if the data is not study-ready.
For context, Kilian’s website has a summary of his training throughout the year. That includes a whopping 1200 hours annually, with 88% of that time at a lower intensity, spread across running and cross-training like skiing and cycling. He wrote a wonderful summary on the website Mtnath in 2019, including a full training diary. Throughout the article, I’ll refer to workout sessions from that 2019 diary as well. Finally, he co-authored Training For The Uphill Athlete with Steve House and Scott Johnston, which you can buy here.
Second, the Coros snapshot is just focused on a 4-week block. Those weeks included Hardrock 100 recovery (where he set another legendary record) and his stellar performance at Sierre-Zinal, one of the world’s most competitive sub-ultra mountain races. Plus he had an asymptomatic COVID infection. (I am so sorry to COVID and its family of viruses, all we can do is pray for them now.) In his post-UTMB iRunFar interview, he talked about keeping his training easy around the COVID infection. So we know he was operating within constraints that won’t exist all the time.
Third, we are assuming the heart rate data is accurate from the training and racing, ideally from a chest-strap monitor. It definitely seems right and overlaps with expectations.
Put all of those disclaimer beats together into this bop: this article is a fun intellectual exercise rather than a definitive statement on what Kilian does or does not do. My co-coach/wife Megan and I went into greater detail on our podcast episode this week. Throughout the article and in the podcast, we tried to be clear about where our biases came in, and how the data may differ from other publicly-available sources. And most importantly: follow Kilian, buy his book, buy his NNormal shoes when they’re available, listen to his interviews, construct a small doll that vaguely resembles him, form a religion around the powers of that doll, etc.
This is glancing through the keyhole and drawing a sketch of the entire house. But it’s rare to get even that glimpse for the house of the GOAT.
Takeaway One: In the snapshot, Kilian likely used a Pyramidal Training structure, with elements of Threshold Training
The Coros site broke down Kilian’s training in a traditional 5-zone model:
- Zone 1: 56.9%
- Zone 2: 20.2%
- Zone 3: 14.5%
- Zone 4: 4.6%
- Zone 5: 3.8%
A fantastic overview of intensity zones is in this video by Stephen Seiler.
57% in zone 1 is a lot of low heart rate training! Even if this data is just a snapshot of a broader training program, it provides another reason to keep most training easy. That aligns with data from elite runners (2022 study in Sports Medicine–Open), cyclists (2022 study in the Scandinavian Journal of Medicine & Science in Sports), and cross-country skiers (summary of the literature).
However, it’s a bit different for Kilian due to the race recovery and COVID infection. Plus, he can move very efficiently at very low heart rates. That ability was honed over decades of high-volume training using a scientific approach. Be cautious applying the launch angles of a bazooka to the launch angles of an intercontinental ballistic missile. One of Kilian’s quads should be tattooed with “Mutually Assured” and the other should say “DESTRUCTION.”
Big takeaway: there is no such thing as too easy for around 40-50% of training for most athletes. Yes, if you did all of your easy training extremely slowly, it might lead to worse biomechanical and aerobic outcomes. But this 40-50% is just one part of the base-training pie, and it enhances recovery rates while improving mitochondrial function and lactate clearance.
Kilian had a tweet earlier this year that said it best (translated from Spanish): “It’s the aerobic base. It is essential, and it is what allows you to work on other skills. Then it is the periodization.”
Zone 2 moves up to steady running, but would still be considered easy to easy/moderate in most training plans for most levels of athletes. These efforts are also great for base building, but could break an elite athlete down over time since they are relatively high output and not purely aerobic. A pro marathoner might be running 5:30-45 minutes per mile pace in Zone 2, and even if that feels easy aerobically, it will wear them down biomechanically.
For athletes that are volume-limited or have less running background, it might be beneficial to have higher than 20% in this range, and a bit less in pure Zone 1. Uphills can venture into Zone 2 more often as well, since the impact on the body is lower. However, I think early-career advanced athletes sometimes make the mistake of spending way too much of their time in Zone 2, chasing those slightly sexier easy run paces. Meanwhile, they produce more cortisol and more nervous/endocrine system stressors, plus more musculoskeletal breakdown, but without much aerobic fitness benefit that can’t be achieved in Zone 1.
Zones 3 and 4
Zone 3 and Zone 4 involve more moderate running and traditional tempo training. Threshold is a sweet-spot for aerobic power adaptations. (Note: different sources/calculations use different delineations for where Zones 3 and 4 transition to lactate threshold, so the exact intensity distribution relies on those methods, plus location within the zone, which is not info we have here.)
At the lower end of moderate in Zone 3, it might be what is traditionally called the “gray area,” depending on the exact effort level. And Kilian spent 14.5% of his time there. That higher percentage could be owed to steady uphill running in training, where the heart rate will naturally be higher, or moderate workouts. That proportion demonstrates that the gray area is a fine spot to hang out for adaptations at the meeting point of aerobic and musculoskeletal stress, just don’t take up residence there.
The higher-end Zone 4 threshold+ work is just 4.6% of his training time in this block. That may have to do with the recovery from Hardrock, since he has completed many difficult high-end threshold sessions at other times of the year.
For example, in his 2019 training diary, he had some of the following threshold sessions: 3×5000 (14:58-15:20-15:15) / 4×2000 (5:32-5:55-6:10-6:00) / 5k (15′) 4k (12’20) 3k (9′) 2k (‘5’50”) 1k (2’50”). Monstrous! There’s a reason he is the GOAT.
Zone 5 involves more VO2-max+ style efforts, and he is still doing a pretty large quantity of training in this intense zone for an ultra runner. Imagine a 12-hour training week–3.8% of that time is 27 minutes. A 5 x 3 minute hill workout would only be 15 minutes (I AM A MATH GOD), so even though the number is low, that may be more indicative of focused hard workouts without hammering unreasonably than a lack of high intensity.
I wonder if we may just be getting the impact of Sierre-Zinal on the numbers in that percentage–it’s possible that he wasn’t doing this time as specific weekly workouts. In his 2019 diary, he includes some fast track-style workouts, but I’m unsure if that was included in the UTMB sharpening window. Whatever the driver, it shows that small amounts of high-intensity work can still lead to ultra excellence.
Interestingly, a 2019 study in the Journal of Strength and Conditioning Research on the training of 85 elite male athletes showed that long intervals above 1 kilometer had the lowest correlation with long-term growth, at just 0.22 after 7 years. That may be due to athletes running these intervals too hard and accumulating too much Zone 5 time, undercutting aerobic growth. Contrast that with Norwegian training approaches, which involve plenty of long intervals but with strict intensity control based on lactate production that keeps the effort out of Zone 5.
Aligning With The Research
Most physiology studies now use a 3-Zone model for ease of quantification (see this fantastic 2022 study in the Scandinavian Journal of Medicine & Science in Sports for an overview), plus the barriers align with physiological principles:
- Zone 1 is easy to steady (approximately Zone 1 and most of Zone 2 in 5-Zone model),
- Zone 2 is a wide band for threshold (~Zones 3 and some/all of Zone 4 depending on the model), and
- Zone 3 is above threshold/critical velocity (Zone 5, with some of Zone 4 depending on the model).
In that model, Kilian’s training in this snapshot is strongly Pyramidal (with some Threshold elements), meaning he does more work in Zone 2 than Zone 3. That aligns with what he has said about his training in interviews, with an aerobic emphasis and plenty of threshold work.
To put it all together, even if the snapshot is just a brief moment in time for Kilian, I think it provides good guidance for intensity distribution more generally for all athletes.
- 40-55% very easy. Stack up some fun, chill hours doing what you love. It can be as easy as you want! Kilian spent even more time here during the 4-week snapshot.
- 20-35% mostly easy and up to steady. While purely chill time is good, you don’t want to become a human snail. The biomechanics of moving more efficiently matter. The first two bullet-points should usually add up to at least 80% of your training.
- 10-20% moderate to moderately-hard. There is wide latitude for what exactly constitutes threshold training, and that’s where the magic of specific training approaches really come in.
- 5% hard or fast. The biomechanical, neuromuscular, and musculoskeletal adaptations from actually putting out lots of power are essential to support efficient performances at lower intensities. However, excessive intense work may counteract aerobic growth and contribute to injuries.
Athletes racing shorter distances under 1 hour should likely polarize their training more leading up to events, with a higher percentage of hard or fast (up to 10-20% for track racers in competition season). Even ultra runners can pulse in short periods with more Polarized training, making sure they always return to the long-term aerobic growth emphasis of a Pyramidal structure.
Two: Kilian’s vert profiles were specific to race day in his final 4 weeks of training
According to the Coros data, he averaged 75-150 meters of climbing per kilometer in the runs included in the snapshot. That equals 396-750 feet per mile. So an average 10 mile run might have anywhere from 4000 feet of climbing to 7500 feet of climbing!
What makes Kilian’s training so fascinating to me is thinking about how this snapshot stacks on top of previous training. Over the winter, he focuses on skiing and builds his aerobic system. In the spring, he often does some traditional speed training–he is so damn fast and I have no doubts that he could run a legendary marathon if he dedicated a couple training blocks to it. And as the seasons change, his training gets more specific to his events.
These high vert ratios align with the demands of the UTMB course. While every approach varies, I think that for an event as difficult as UTMB, at least 6-8 weeks before the race should be focused on the specific demands of vert. During specific training, try to do steep trails like the event 2-3 times per week at least, particularly on long runs (adding treadhills or downhill repeats if needed to achieve that specificity where you live).
Three: Kilian’s internal effort fluctuated in fascinating ways during UTMB
Interestingly, Kilian’s heart rate was BY FAR the highest near the start and in the first half of the race, before drifting downward even as he gapped his competitors. Ultra physiology is bizzaro world stuff–in a road marathon or any shorter race, you’d expect the reverse, with the internal metrics getting more strained later on. That flip is likely due to the body burning through glycogen stores, along with some weakly understood nervous system impacts. That unexplored valley of physiology is where fatigue resistance lives, so the more we can learn about it, the more clues we might have about how to improve it.
During the first 30k, Kilian ran some portions above 170 heart rate! That shows how the race dynamics involved athletes hammering off the front this year, and likely shows that to have a shot at winning in future years, you have to be ready for a barn burner out of the gate. His average heart rate for the 2:38 stretch was 151, much higher than you might expect to kick off a 20-hour race. He’s so efficient that he is sparing most of his glycogen at this effort, but it’s right on that razor’s edge and it required incredible aerobic development.
Over the next 30k, his heart rate still averaged 148, with uphills approaching 155. My guess is that Kilian has it dialed in enough to know that he can manage those uphill heart rates without breakdown as long as he is fueling a ton along the way. And sure enough, in the aid station videos, he looked like the world’s most chill competitive eater. As he tweeted: “Eat, eat and eat more. An ultra is an eating challenge while running and hiking in nice places.”
Around halfway of the race, Kilian’s heart rate average dropped dramatically, all the way to 132 for the section to 103 km. This is also when Jim Walmsley got away from him, a moment he talked about in the iRunFar interview. Then on the long downhill to 127 km, he let his heart rate go all the way down to 119 average! Jim got 14 minutes ahead. But Kilian likely was making a calculated decision while listening to his body–push the pace at the start to winnow the field and compete with Jim, back off to allow for glycogen recovery.
Kilian was caught from behind by Mathieu Blanchard, and he used that as an impetus to start a steady push. While his heart rate stayed low (125-130 average), he surged multiple times. My guess is that he understood exactly how many matches he had to burn, likely via his scientific training methods and decades of self-analysis.
Kilian should have an honorary Ph.D. in training theory and physiology, with a dual degree from medical school too. There’s a reason you can call him the GOAT.
David Roche partners with runners of all abilities through his coaching service, Some Work, All Play. With Megan Roche, M.D., he hosts the Some Work, All Play podcast on running (and other things), and they answer training questions in a bonus podcast and newsletter on their Patreon page starting at $5 a month.