The Ethiopian running secret
The Ethiopian running secret
One school of training is highly personalised, technical and data-driven. The other is the one that wins marathons
by Michael Crawley & Geoff Burns + BIO
A training group in Sebeta, Ethiopia in 2016. Photo supplied by the author
is an award-winning author and social anthropologist based at Durham University, UK. His latest book, To the Limit: The Meaning of Endurance from Mexico to the Himalayas (2024), was shortlisted for the Sunday Times Sports Book Awards.
is a sports physiologist for the United States Olympic and Paralympic Committee and a faculty member at the University of Michigan’s School of Kinesiology. His research focuses on sports science and human performance, and his work has been featured in publications including The New York Times, the BBC, The Wall Street Journal, and Runner’s World. He is based in Colorado Springs, Colorado.
Edited byRichard Fisher
Gojjam wipes a streak of vomit from the corner of his mouth and turns to his friend Zeleke. ‘I did your turn at the front today,’ he says, ‘and my soul almost came out.’ He squirts water from a bottle into his mouth and spits. ‘Leading is hard. It’s like carrying someone else’s burden.’ The two athletes sit on the side of the Chinese-built road that leads southwest into Oromia from Addis Ababa, Ethiopia’s capital city. Beyond the tumult of cars and buses and the occasional horse-drawn cart, farmland stretches to the horizon in every direction. They have just run 25 km with 14 other athletes at a pace designed to prepare them for an upcoming marathon. Before they started, their coach, Messeret, carefully divided the responsibility of leading sections of the run between them, emphasising the importance of doing their ‘duty’ as pacemakers and invoking them to ‘share their energy’ with their teammates. As the exchange between Gojjam and Zeleke illustrates, monitoring the effort of training is understood as a collective endeavour, requiring a great deal of trust and reliance on others.
In 2025, athletes from Ethiopia and the nearby East African nations of Kenya, Uganda, Eritrea and Tanzania filled 69 and 74 of the top-100 spots in the World Athletics marathon rankings for men and women, respectively. This is an extraordinary level of dominance, with few parallels in global sport. In these countries, distance running expertise is seen as something that is intuitive, learnt from others, honed through experience, and deeply dependent upon a group training dynamic. Increasingly, though, this approach goes against the grain of cutting-edge sports science, which advocates the monitoring of an ever-increasing number of physiological variables and individualised, precisely engineered training.
Sabastian Sawe of Kenya leading Yomif Kejelcha of Ethiopia during the 2026 TCS London Marathon on 26 April 2026 in London, UK. Both runners crossed the line in less than two hours. Photo by Warren Little/Getty Images
So do Ethiopian runners and those with similar methods succeed in spite of, or because of, their training philosophies? Viewed through the dual lens of anthropology and sports science, we believe this question runs deeper than athletics results, speaking to what it takes to stretch the limits of human ability, and how expertise is used, assessed, embraced or rejected. For a long time, African knowledge has been seen by the West as intuitive, superstitious and practical, and some might view the Ethiopian methodology as unscientific. We argue it may be just the opposite. So is it Western sports science that reveals new frontiers of human potential? Or is it merely catching up, while the true innovation occurs on the dirt tracks of Addis Ababa?
Join more than 270,000 newsletter subscribers
Join more than 270,000 newsletter subscribers
Our content is 100 per cent free and you can unsubscribe anytime.
Our content is 100 per cent free and you can unsubscribe anytime.
Our mutual interest in running training methods began more than a decade ago, following a conversation as competitors at the World 50 km championships in 2015. Since then, Michael – a social anthropologist – spent 15 months doing ethnographic work in Ethiopia, learning Amharic, living and running alongside athletes as they sought to change their lives through the sport. Geoff, who is now a sports physiologist for the United States Olympic and Paralympic Committee, has come at the problem of human performance from an engineering background, having previously worked in the automotive and medical fields.
There is no better place to understand the engineering and data-driven approach to running with which Geoff is familiar than a university town in the highlands of Arizona: Flagstaff. With miles and miles of relatively flat, soft dirt paths, it is situated at the perfect altitude – around 2,100 m – to promote beneficial oxygen-handling adaptations in the blood and muscles. It also has all the modern conveniences a detail-oriented runner could want: accessible tracks, stocked gyms, organic grocery stores, and skilled masseuses and physiotherapists. Top runners visit from all over the world and, as such, it has a further X-factor: a community of athletes engaged in the collective pursuit of getting faster. Yet whereas in Ethiopia the athletes work in groups, here the training is often more solitary. Modern trends in individualised, precise, engineered training have resulted in more and more of these athletes pursuing common dreams but in self-tailored, idealised isolation.
British long-distance Olympic runner Mo Farah training above Flagstaff, Arizona. Photo by Michael Steele/Getty Images
Visit the iconic blue track at Northern Arizona University any morning, and you’ll almost certainly find myriad bubbles of Olympians, national champions and dreamers running carefully controlled laps. Look closely, and you’ll probably even find little purple strips with drops of spent blood peppering the side of the track – overlooked modern refuse from the precise monitoring of energy flux. Those strips, similar to the ones a diabetic would use to monitor blood glucose, are from measurements of blood lactate, the current gold-standard biomarker, ex laboratorio, for an endurance athlete’s relative rate of energy expenditure. Lactate, a molecule once erroneously vilified as a source of muscle-inhibiting acid, is now better understood to be a barometer for the internal rate and sustainability of a given physiological effort. Consequently, that energy flux that the athletes in Flagstaff are monitoring is of a distinctly different flavour than what concerns the athletes in Addis Ababa. It’s the same energy that’s more typically on the mind of a physicist or an engineer: the standardised quantification of work, be it chemical, electrical or physical.
The frequent measurements serve as a constant, controlled feedback loop for the athlete and coach
In Western sports science, endurance training is now viewed and prescribed as an energetic, metabolic stress. Metabolism itself is simply energy transfer: the chemical energy in our foods gets converted into electrical energy inside our cells, which then ultimately gets converted to mechanical energy in the contractions of our muscles. The rate and efficiency of these processes is the intensity of our exercise and training. Certain rates are relatively unstressful (ie, you could hold a conversation while maintaing them), higher rates are stressful but sustainable (where one runs a marathon, for example), and the highest rates are quite stressful and very much unsustainable (racing a mile or a 5 km race might require this rate of energy expenditure).
The aforementioned molecule, lactate, is simply the product of breaking apart sugar in the muscle, glucose specifically, for rapid energy in the cells. The rate that it appears and disappears in the blood is a tight proxy for the internal rate of energy combustion in the cells, and its stability – accumulation, or lack thereof – reveals the sustainability of that rate of energy flux, revealing whether the athlete’s body is fundamentally in control or out of control. More and more athletes are relying on this biomarker, along with heart rate – a more established but sometimes less reliable marker of physiological strain – to guide the precise speeds and intensities at which they perform their individual training. It’s not uncommon for elite distance runners to pause every few reps in a session to take a blood sample to calibrate their paces, speeding up or slowing down for the next few reps, even by just a few seconds, depending on what the test reveals.
This approach to training, often referred to as the ‘Norwegian Method’ for its Scandinavian pioneers in distance running and triathlon, can be seen more agnostically and simply as an engineering approach. Each athlete has very precise intensities at which they should train to get the most out of the session without compromising the ability to do the next session, thus maximising the total amount of intense work that they can do (or energy they can expend) in a training cycle. It requires frequent measurement of heart rate and blood lactate. To train above or below those precise intensities is to compromise the potential adaptations for the athlete, and thus the effort and pursuit itself. And the frequent measurements of internal strain against external stress themselves serve as a constant, controlled feedback loop for the athlete and coach. If the athlete is continuously doing their intervals to control their blood lactate at 3.2 mmol/L, and the speeds that elicit this level go from 3:07/km down to 3:05/km down to 3:03/km, the success of the prescription is obvious.
Such control and precision are exactly at odds with the Ethiopian valuation and management of their energy. A tailored, individualised management of physical energy is necessarily non-social, while in Ethiopia, the important properties of energy are that it is understood to be a limited substance that must be carefully monitored and protected. It is understood to be a ‘transbodily’ substance – that is, it can flow between people, as well as between people and their environments.
While monitoring technologies such as GPS watches are now widely available in Ethiopia, they are used selectively, and seen as inappropriate for some kinds of running. On the eucalyptus-forested slopes of the Entoto Mountains, which reach an altitude of 3,200 m at Mount Entoto, the precise quantification of speed and distance is seen as a distraction. Rather, the emphasis is on creating zigzagging routes through the forest, using the uneven ground and slow pace to enable the legs to recover. Some runners are seen as better at creating routes, so are sought out to lead. The sessions are different every time and rarely follow the forest trails.
The rejection of technologies of quantification was especially clear on one particular run, where the athletes were told by the coach to cover between 17-18 km in an hour and 20 minutes. Bogale, who was put in charge of leading, was given a Garmin watch, and the runners set off on a particularly meandering route, often turning 180 degrees around trees, and at one point having to use tree roots to pull themselves up a slope. After running into a deep hollow and disturbing (to some of the runners’ evident glee) some hyenas, one of the other runners shouted to Bogale to speed up, or we wouldn’t ‘have enough kilometres’ on returning. He refused, saying that with this kind of training the important thing was to ‘go up and down’. Here, the intuitive and creative approach to exploring the forest rendered the GPS device – and the coach’s authority – superfluous. When the athletes returned to the clearing at the end of the run, the coach was told that the watch was not working.
A training group in Sendafa, Ethiopia in 2016. Photo supplied by the author
In Ethiopia, it is the ability to run in a way that protects your own and others’ energy that is seen as the primary skill of an endurance athlete. In this instance, doing this well was often seen to require the rejection of quantified data rather than its embrace. Rather than seeing energy as contained within the individual body, and the athlete as a system of inputs and outputs, Ethiopian athletes see the process of expending and monitoring energy as a collective responsibility. Because the total amount of energy is seen as limited, for one athlete to gain within this system necessarily involves another athlete losing something. For this reason, there is a complex ethics of training in Ethiopia, which ensures that energy is shared as evenly as possible.
Where is expertise located – in the laboratory and sports scientist, or on the trails and with the athlete themselves?
Among Ethiopian athletes, running alone is seen as deeply antisocial in the same way that eating alone is. Running together is an important way to ensure that people control themselves and avoid ‘burning up’, as they put it, by training too hard. Similarly, it is important that food is shared equally between athletes, and if people fail to do their ‘duty’ as a pacemaker, they are often required to redress this energetic imbalance by ‘sponsoring’ bread or bananas for the rest of the athletes.
But for all the benefits of sharing, using and exchanging energy in a social context, a cautious engineer might look at those practices and see that some athletes are surely expending energy at rates that are not ideal for their own physiological adaptation and growth. If every athlete has a distinct, individual capacity and efficiency, with distinct, individual demarcations between sustainable and unsustainable rates of energy expenditure, then surely the ideal intensities at which they should train are distinct and individual?
Some might look at the Ethiopian approach and see echoes of the ‘old school’ approach of Western training, where more was good and harder was better, lending itself well to big groups training together. Top athletes could lead the way and set the standard, and developing athletes could push themselves to hang on with the hope of making the jump to the next level. And everyone would hold each other accountable. Training wasn’t tailored to everyone, but dose and adaptation were assumed to be somewhat crude concepts, and the power of the collective was known to supersede any nuance in them.
The Norwegian style, by contrast, seemed to move things on with its new-school, ‘Goldilocks’ engineered approach, where the best intensity is the most fragile of lines between not-hard-enough and a-bit-too-hard. If training with someone else, the idea is not to compromise the ideal adaptive stimulus that is your dose and intensity, and obfuscate your progress and feedback. At one time, you could watch the three Ingebrigtsen brothers – Henrik, Filip and Jakob, who popularised the Norwegian approach – train ‘together’ as simultaneous world-class 1,500 m runners. But while they were often doing the same workout, they were very much in their own worlds. Can you argue with their results? They would be the first to tell you that their cautious, tailored, individual approach took them from obscurity to global titles and medals. But the old-school coach would be quick to give you an expletive-laden description of lactate meters and heart-rate monitors – and the athletes using them – best summarised politely as ‘neutered’.
In some ways, this old-school controversy comes down to ideas about expertise and where it is located – in the laboratory and lactate strip and sports scientist, or on the trails and with the athlete themselves. While many have assumed that East African athletes’ success comes ‘naturally’, or is derived almost automatically from the advantages of genetics or altitude – there is a huge amount of expertise about endurance running in Ethiopia. It is not ‘old school’ at all, but more refined, built upon decades of cumulative knowledge. It just can look a little different to Western sports science: less about lab testing and utilising data, and more about creating a balance in training between different kinds of environmental conditions and learning to share energy with others.
The way in which Ethiopian runners approach their training is in fact highly scientific. Britt Rusert, professor of Afro-American Studies at the University of Massachusetts Amherst, has coined the term ‘fugitive science’, broadening the definition of science to include other kinds of embodied empirical practice. East African runners have often been approached by researchers primarily as a source of physiological data, rather than being taken seriously as people with in-depth knowledge about endurance.
In some ways, the Ethiopian and Norwegian approaches look very different from each other but there is in fact one major similarity – and here there are lessons for elites, amateurs and beyond. While much of the messaging about endurance sport, especially on social media, is about pushing harder or suffering more or descending deeper into the ‘pain cave’, in fact the best athletes in the world are concerned with restraint. How can they improve at a sustainable rate without ‘burning themselves up’ through overtraining? Whereas the Ethiopian approach to this is profoundly social, the Norwegian method privileges the ‘objectivity’ of ever-increasing quantification. Both methods are themselves ‘checks’ on an individual’s maladaptive tendencies to push too hard or not hard enough – one is social, and one is technological. And for each, their use taken to an extreme could itself be maladaptive: the Ethiopian approach leading an athlete too far from their own path of consistent and predictable growth, and the Norwegian approach leaving the athlete too disconnected or too fragile to the turbulent demands of reality.
One thing that does seem clear is that the Norwegian approach has a great deal of currency with both amateur endurance athletes and the general public. It coincides with advances in technology that generate more data than ever before, as well as a wave of companies seeking to commercialise it. Now, anyone with a GPS watch and a smartphone can monitor their pace, heart rate, cadence, elevation, VO2 max, sleep and more – and, if they choose, share detailed performance graphs on social platforms like Strava. Companies such as Whoop market themselves as providing privileged insights into the body (one advert features the line ‘You know the inside of everything. Except you’) encouraging people to make decisions based on their HRV (heart rate variability) used as a proxy for stress. And the latest twist is to add artificial intelligence, with various subscription services offering personalised, algorithmically tailored training plans. All this raises a question about what happens to the skill of monitoring how running actually feels intuitively.
Running was not just about careful calibration, but also about a sense of ‘dangerousness’, adventure and fun
Perhaps the golden mean lies in using these engineering tools to calibrate intuition, while simultaneously creating space to engage the dynamics of a collective. The social structures tied to the pursuit of self-improvement – be it races and competitions, or simply clubs and friends – can stretch and knead rote tendencies and perceived limits. And tools of quantification – watches, meters and heart-rate monitors – can refine those perceptions for a more reliable understanding of stress and response. Humans are undeniably creatures prone to bias, emotion and error, and these modern tools can help runners check those demons and refine behaviours.
There is also a broader question here about how societies and cultures value particular kinds of knowledge. While concerned about maintaining and monitoring their energy levels, Zeleke and Gojjam would occasionally indulge in seemingly excessive and unnecessary expenditures of energy. They would get up at 3 am, for instance (imagine the diligent sleep-tracker’s horror!) to run up and down a particular hill in the darkness, before taking a bracing outdoor shower. They did this because running was not just about careful calibration, but also about cultivating a sense of ‘dangerousness’, adventure and (dare we say it?) fun.
It is possible to imagine a future in which sporting performances are attributed to personalised technologies or biomedical protocols, rather than to characteristics such as grit, determination and sacrifice, or the communities that nurtured an athlete in the first place. As societies increasingly delegate human attributes in wider spheres of life to machines and pharmaceuticals – intuitive feeling to self-tracking devices, thinking to AI, self-discipline to appetite suppressants – it is important not to lose sight of human connection along the way. The self-quantification that is increasingly available in devices and apps can indeed help to hold us accountable to our individual selves and positively change behaviour, but the scope of their measurements and guidance will always eventually be eclipsed by the totality of being human. Whether you are an athlete or not, we are all, at our core, social creatures and, to be at our best, we need others to push and pull us there.
We publish hard-won knowledge from real people who have grappled deeply with their subjects.
We publish hard-won knowledge from real people who have grappled deeply with their subjects.
Your donation, whatever the size, supports our mission to ask the big questions and deliver fresh, original insights from leading thinkers.
If you value what we do, will you support us?
Your donation, whatever the size, supports our mission to ask the big questions and deliver fresh, original insights from leading thinkers.
If you value what we do, will you support us?
I would like to donate:
Select amount (US dollars):
The rules on doping in sport are incoherent – should we change them to allow the right kind of performance enhancement?
essayPhilosophy of mind
Thinking on your feet
Don’t just do it, think about it too: how Gilbert Ryle’s philosophy of mind can help athletes teach themselves to improve
essaySports and games
Performance-enhancing vices
Selfishness channels ambition, envy drives competition, pride aids the win. Does it take a bad person to be a good athlete?
essayComputing and artificial intelligence
There’s an app for that
What to eat, when to meditate and whether to call your parents: can self-monitoring tools make a difference?
Who is a sportswoman?
Elite female athletes are subjected to invasive gender tests, and hormone treatments if they fail. This is deeply unfair
Prosthetics, doping, computer implants: we take every upgrade we can get. But what is waiting for us at the finish line?
Sign up to our newsletter
Updates on everything new at Aeon.
© Aeon Media Group Ltd. 2012-2026. Privacy Policy. Terms of Use.
Aeon is published by registered charity Aeon Media Group Ltd in association with Aeon America, a 501(c)(3) charity.
