Pelotonomics: How the power of the pack shapes the Tour de France

Peter Gray, Senior Director of the Sports Practice – Technology at Dimension Data

The ‘peloton’ – the pack of competing cyclists from rival teams working together to hunt down the leaders in a cycling race – is one of the most iconic and puzzling aspects of the Tour de France.

The peloton is a fast-moving contradiction, simultaneously representing co-operation and fierce competition, as rivals are forced to work together in order to have the best chance of individual and team glory. Over the course of a race, it can turn what looks like an unassailable lead from breakaway riders into an apparently foolish gamble, as it catches and consumes them.

And that’s the key to the peloton’s success. For all the risk and forced co-operation with the competition, hunting in a pack brings great reward. Riders in the middle of the peloton can enjoy a reduction in wind resistance of as much as 40 per cent, meaning teams will often hide their prized sprinter away in the centre of the pack, letting them conserve energy until near the end of the race, before delivering them to frantically expend those reserves in a final burst of pace. Those riders who broke away and built a lead have had no such protection from the elements and, as a result, have none of the same energy reserves stored up in the peloton.

Our name for this complex balancing act is ‘pelotonomics’. And each year at the Tour de France, millions of data points are collected from the competing cyclists, from the second their wheels start turning to the final stage along the Champs-Élysées in Paris.

During this year’s Tour, we’ll be crunching over 150 million data points, meaning fans will have access to an incredible amount of information about their favourite riders, teams, and stages. If the breakaway group during Stage 5 of the Tour is set to outpace the peloton, we’ll know about it, and so will Tour de France supporters across digital and broadcast, as data visuals flow seamlessly between all channels and platforms.

And most exciting of all, with the power of machine learning and predictive analytics, we’ll be able to understand even more about how those spectacular pelotons really work.

Now picture this

Hundreds of cyclists power through the sleepy French village of Mouilleron-Saint Germain on a 185-kilometre epic journey during the second stage of the Tour de France 2018. Taking in the gentle rolling hills of Western France, this route is tailor made for a fiery sprint finish.

Live GPS co-ordinates reveal that a small group of riders has broken away from the peloton early in the race and is desperately trying to open up a gap ahead of the main group. The peloton is grinding through the first half of the race, subtle changes in speed and rider position highlighting how various teams are plotting how to steal the advantage at the vital moment. The GPS sensors pinpoint an almost imperceptible increase in acceleration, as the team near the front of the peloton shift gear.

Thousands of data points have just been collected in the past few minutes, and if the predictive analytics are correct, the peloton will catch the breakaway group just before the explosive final sprint. With one kilometre to go, the peloton flies past the burnt out breakaway at the flame rouge. Teams jostle in order to place their chosen sprinter in the best position, while from the relative comfort of the race centres, data specialists frantically relay information to their cyclists via earpieces, detailing their exact speed, the pace increase required in order to reach the front of the peloton, and the crucial moment to make their move in order to claim victory.

The tech behind the stories: How do we track the peloton?

These kind of stories are commonplace at the Tour de France, but few realise how much technology and science are required to weave them into a coherent narrative. Prior to the start of the Tour, state of the art sensors are placed on the back of every rider’s bike, which provides access to basic information such as cyclist speed and location. This is then combined with more unpredictable variables, such as wind strength and direction, along with the potential effects of gradient and altitude on the speed of the peloton.

Anything can happen in this high-octane race; in 2015, over 20 cyclists were involved in a terrifying peloton crash during Stage 3 of the Tour, with data revealing theaverage rider speed during the fall was a blistering 42 km/h.

This year, we’re translating these variables into a thrilling ‘rate of catch’ prediction, where we’re able to calculate when the peloton is about to catch the breakaway group – and exactly how they’ve managed to accomplish it.

We’re also sharing detailed rider profiles on every cyclist. For example: it’s widely known that Mark Cavendish is one of the sport’s strongest sprinters, but we’ll now be able to delve in greater depth into his performance throughout every stage to understand more about why specific sprinters finish where they finish and how they perform throughout the stage.

Our increased insights will do nothing to detract from the wonder and spectacle of the peloton, but they will help us all understand more how it shapes and decides the outcome of the most unique sporting event in the world.