Sign up for our newsletter

How the pro peloton explains the secrets of life

In-depth
22 Feb 2021
Advertisement

Did you know that the secret of life itself may lie hidden in a professional peloton of riders? We just need to understand the science…

Words: James Witts Photography: Offside

Can cycling explain evolutionary biology? Are the patterns found in pelotons reflected in the flocking of birds and the schooling of fish? And can we truly understand the factors that determine the shape and flow of a peloton as it moves along the course of a bicycle race?

These are the questions thrown up by an academic paper that appeared in our inbox recently, entitled Information Flow and the Distinction Between Self-Organised and Top-Down Dynamics in Bicycle Pelotons.

Now, we’re not scientists here at Cyclist, so we can’t claim to have fully understood every nuance of the concepts presented in the paper. But we are cycling nerds and are keen to learn more about the scientific theories behind our sport – we just need someone to explain them to us. So we decided to call up the paper’s author.

Self-organising or top-down?

Hugh Trenchard is a Canadian living in British Columbia, where he works as a branch advocate for the liquor and cannabis industries. He also loves cycling – really loves cycling. Trenchard raced competitively for 20 years from the mid-1980s and he has written several academic papers about the sport, including the one that piqued our interest.

‘When racing I developed an intimate understanding of the peloton, and that’s where I’ve grown my model of self-organised behaviour of pelotons,’ Trenchard tells us. ‘This paper is written to shed light on how information about the race is transferred amongst the riders, and then how much of the result is “natural” and how much is “human-driven”.’

According to Trenchard, the way a peloton is organised – who’s on the front, who’s at the back, who’s moving up, who’s breaking away, and so on – is partly determined by factors that are unchangeable on the day, such as individual riders’ maximum power output, and partly by things that can be manipulated during the race, such as team tactics or a rider’s own competitive motivations.

Trenchard describes a peloton shape that arises from factors beyond the riders’ control (their own fitness, the weather, etc) as ‘self-organising’. A peloton’s shape that arises from rider volition (decisions made on the day by riders or managers), Trenchard calls ‘top-down’. What interests him most is which of these is actually responsible for how a peloton operates.

To begin with, he posits that information flows through the peloton in two ways: between cyclists (‘intra-peloton’), such as whose bidon is nearly empty, who looks poised to attack, and from sources outside the peloton (extra-peloton), such as orders over race radio or calls from spectators.

Much of the intra-peloton information is observational and up for grabs to anyone – eyeing other riders and their behaviour, assessing the road conditions – while extra-peloton information is solely for a specific rider or team. Regarding the former type, there’s perhaps no better description than in Bernard Hinault’s 1988 book Road Racing Technique And Training.

‘The course and the weather aren’t the only things to watch,’ says Hinault in the book. ‘There are also the other riders. First I note which are the “easy” ones by watching the way they ride and react. Then I look at those who are really racing or who have a teammate who is a contender.

‘If you see five or six riders from the same team beginning to move forward in the peloton, it means they’re cooking up something. You have to move up with them to be sure to be in on whatever happens. If you notice a rider who slips into all the breakaways, it means he’s racing to win… It’s rare that I don’t know exactly who has gone with a breakaway.

‘Since I know the riders well, I know how much of a lead I can let them get away without jeopardising my own chances… If you notice that a dangerous opponent is looking tired, and if the terrain is favourable, it might be worthwhile to attack.’

Trenchard expands on the point: ‘A peloton is a human system and owes its existence to riders who deliberately apply competitive motivations and who act on rational assessments of the information available. So cyclists’ actions that result from these motives are “top-down” in nature.

‘The question naturally arises whether “self-organised” peloton dynamics do in fact emerge or whether peloton dynamics are primarily driven by these top-down influences and are fundamentally not self-organising.’

Coupling up

The starting point for a peloton’s creation is a concept called ‘coupling’ – the interaction between two bodies that are mutually influencing each other.

Through a cycling lens, coupling is commonly seen when two riders take pulls for the benefit of themselves, which has the effect of benefiting the other. Take Stage 11 of the 1985 Tour de France, for example. Bernard Hinault broke free from the peloton alongside Colombian climber Luis Herrera.

While Hinault protected the diminutive Herrera from the wind on the flat, Herrera shielded Hinault in the mountains. The result? Hinault cemented his place on GC and Herrera clung onto his King of the Mountains jersey. This kind of coupling is tactical – that is, born of a rider’s volition – but ultimately dictated by natural behaviour in that it could only work if Hinault and Herrera were both physically up to the task.

It’s this coupling that multiplies to form a peloton and hence leads Trenchard to the idea that the fundamental framework of a peloton is self-organised, primarily arising from natural factors, albeit on the proviso that volitional ‘top-down’ organisation has a role and the weighting given to each is still be to understood.

‘Basically, we can strip away all the human aspects of cycling – the psychological input, strategies, coaching from the sidelines – and distil it down to three basic principles that explain, and to some extent predict, the peloton’s behaviour: the energy-saving benefit from drafting; cyclists’ inherent maximum physical capacities; and their current power output or speed.

‘For self-organised behaviour, it’s critical that we think of the peloton in terms of its macro-structures or collective behaviour,’ Trenchard says. ‘The macro-structure is defined by two major phases: compact, high-density pelotons that move at slower speeds; and stretched, single-file pelotons that move at higher speeds.

‘Each of these phases is determined by the three principles of pace, energy-saving and maximum power output, so if all the cyclists are travelling at between 30-65% of their maximal outputs, we can predict that the peloton will be compact, like during a sprint stage at the Tour. If all the cyclists are travelling between 65-95% of their max, we can predict a stretched-out peloton like in February’s UAE races.

‘The peloton also moves in a convective motion, which is essentially riding to the side, dropping back and mixing it back in with the peloton,’ he adds.

‘If I can extract some of the top-down strategic behaviours – such as the managers instructing riders where to place themselves in a peloton – and analyse the basic biological and physiological principles, can I observe these macro-structures to enhance our understanding of evolutionary biology?’

It might seem some leap, but Trenchard believes he – or at least someone – could. And here’s why…

Energy evolution

The key factor in cycling – and evolutionary biology, as it happens – is that energy-saving element. Broadly, the figures 30-40% are bandied about regarding how much energy a rider can conserve by drafting.

Professor Bert Blocken of Eindhoven University in the Netherlands questioned that figure, especially in a peloton in which the number of cyclists sheltering each other could run well past 100.

To investigate, Blocken had 121 quarter-scale riders made and positioned in generic peloton formation before blasting wind at them. His results were staggering, with many of the riders in the central core and reasonably far back seeing a drag reduction of up to 95% at 54kmh.

Trenchard tempers this by saying Blocken’s presumption of the consistent closeness of riders at 54kmh is unrealistic, but believes that even so energy savings of well over 50% are highly likely.

‘One of the key predictions off my theoretical framework is what I call the “variation range hypothesis”,’ says Trenchard. ‘Say we have a 30% difference between the strongest and weakest riders in the peloton on any one day.

‘Once you have a rider who’s, say, 60% weaker, the drafting benefit is not enough and they’re popped out the back. They’re no longer part of the peloton system.

‘In a natural biological system, you can imagine similar processes are occurring, especially in migrations among birds, fish or trilobites [the now-extinct group of marine anthropods]. Repeated group accelerations and decelerations result in a sorting – a kind of survival of the fittest.

click to subscribe

This is the essential evolutionary process that’s down to those three primary physical parameters: pacesetting speed, individual maximum sustainable output and drafting quantity. It’s why some trilobites measured just a few millimetres, while others reached two-feet long. You get an arms race of larger and smaller groups who separate and form their own sub-species.’

In Yuval Noah Harari’s best-selling book Sapiens: A Brief History Of Humankind, the Israeli historian refers to the work of Harvard biologist Richard Wrangham, who believed that fire drove human evolution.

The theory goes that fire was mastered over a million years ago by homo erectus. Before then, a raw diet dominated, meaning huge swathes of time and calories spent foraging and masticating. Once fire and cooking were conquered, both hunting and eating time dropped, meaning more calories could be redirected to the brain to evolve. Energy conservation had driven evolution and dictated a species’ form. Much like in a peloton, then.

Au revoir, Monsieur Lefevere

The conclusion of all this would seem to be that the pro peloton is a self-organised entity that provides a snapshot of evolution, and sadly the likes of Patrick Lefevere and Dave Brailsford must accept they have little impact on their teams’ results and even less on the evolutionary tendencies of species.

Wrestling with this concept, we proposed several cycling-specific scenarios to Trenchard in an attempt to further understand his model and, beyond that, highlight that this top-down information flow he talks of surely varies depending on the stage or race.

We asked him how a team like Deuceuninck-QuickStep can dominate the Spring Classics, often with a number of joint leaders working together, if not down to significant extra-peloton information such as tactics?

‘You must remember the macro-structure of the peloton,’ Trenchard responds. ‘In my model, does it matter if the first six riders in line are from QuickStep or if they are mixed between the teams? It only matters because someone from one of those teams made a tactical decision to drive up the pace. That’s the top-down part of the race.

‘But the decision to drive up the pace makes no difference to the fact that the peloton is forced into a phase-change, which is determined by those three factors: the pace of the front rider; the maximal abilities of the rider; and energy saved by drafting.

‘Of course there are secondary factors such as collision avoidance and peripheral vision adjustments, but those factors are present regardless of the peloton phase.’

In essence, the importance of top-down influence boils down to the timing of the changes in speed. The rest – power output, drafting, speed maintenance – is down to self-organisation. And while speed change is clearly important in professional cycling (cue Matthieu van der Poel timing his leap from no-man’s land to win the 2019 Amstel Gold) it’s a bit-part player when it comes to our wider understanding of life on this planet.

‘Just think about it: what happens when you have a collection of organic molecules, such as the so-called prebiotic soup that may have seeded life on Earth, and generate the energy to sort it in a certain way?’ asks Trenchard. ‘Could such a sorting of prebiotic molecules give us some insight into the origin of life?

‘It’s pretty “out there”, but it’s the kind of thing that peloton dynamics allows us to be imaginative about and question.’

The correct vision

Does Trenchard’s peloton model stand up to scrutiny? In July 2019, The Royal Society published a paper entitled How Vision Governs the Collective Behaviour of Dense Cycling Pelotons.

The three-year investigation conducted by a team out of Utah University analysed hours of footage from the 2016 Tour de France, leading them to conclude that the changes of peloton structure weren’t down to natural self-organised behaviour, but vision.

Tour de France riders are often only a few centimetres apart from neighbours on all sides,’ explained lead researcher Tadd Truscott. ‘Our image analysis revealed that cyclists align in patterns, and within a plus or minus of 30° are corresponding to the human near-peripheral visual field. This helps them safely react to changes or disturbances from neighbouring riders.’

The researchers also suggested a rider’s field of vision narrows through exertion at the end of the race, which can lead to greater splits among riders, and that their visual model explains the collective behaviour in a variety of animals, including shoals of fish.

Trenchard is not convinced. ‘They don’t propose an actual evolutionary mechanism, like we do,’ he says. ‘It also ignores the effects of a cyclist’s fitness and the drafting effect, implying that formations are dictated by a cyclist’s positional adjustment and collision avoidance through visual cues.’

Trenchard is planning further work to extend his model, most notably to see if it applies to economics where competitive and rational/irrational decisions mix with physical principles.

All we can echo is that analysing self-organised behaviour requires an even playing field, and this is why the pro peloton is the perfect exemplar: it comprises riders with similar physiological make-ups and similar lives, resulting in profiles that means similar behaviour when it comes to pacesetting, drafting potential and maximal power output. And this may well be useful to further study evolutionary biology.

So while it might not help you or the pros ride any faster, by riding in the pack you’ve naturally created a new category of research. Chapeau indeed.