Why breakaways work at Roubaix: Strava and rider data tell the story

We trawled through 16 editions of the Hell of the North and the data shows there is only one clear favourite this weekend.

Don’t miss out on the latest CyclingTips updates.

Jump To Comments

What do Yves Lampaert, Nils Politt, Sylvan Dillier, and Maarten Tjallingii, have in common? Stuart O’Grady, Johan Vansummeren, and Matt Hayman have a similar, but even better, thing in common. These riders have all found themselves on the podium at Paris-Roubaix, having survived from the “break of the day”, with the latter group making it all the way to the top step of that podium.  

It seems like 99 times out of 100 the early break at most races has almost zero chance of any success. A small group of riders, doomed from the outset, with only TV time and some KOM points as a scant reward in return for hours of suffering. Not so at Roubaix.

WorldTour performance consultant Dave Bailey and I were discussing what Strava segments might tell us about the physical effort required to get through Roubaix. At some point, we switched our focus to the difference between the intensities for riders in the peloton versus riders in the break, and eventually the success rate of riders in the breakaway. Dave initially identified the high number of top-20 finishers from Roubaix breakaways since 2015. Further analysis revealed a staggering success rate for breakaway riders going as far back as 2005.

Seemingly the early break at Roubaix is the closest thing to a tactical certainty one could find in the chaos that is the Hell of the North. What is it about Roubaix that makes the early break such a low risk, high reward strategy for all bar the heavily marked favourites? 

Even this photo does Arenberg zero justice.

The course

Yes, Roubaix has a lot of cobbles which make for a particularly unique race. No surprises there. But how does this differ from Flanders, which also has a lot of cobbles, especially uphill cobbles, which on paper seem like an added difficulty? 

The answer is two-fold. Firstly the obvious answer. The cobbles in Roubaix are ridiculously rough, astronomically more challenging to ride, and absolutely treacherous. While the cobbles in Flanders are laid in some form of pattern, the cobblestones at Roubaix look more like they have fallen from the sky and landed at odd angles with huge gaps between them. All this means that the fight to be at the front of the peloton – by no means easy in Flanders – is at another level on the approach to each of the Roubaix sectors.

Riders in the peloton need to expend considerable energy just to get onto the pavé in a good position. Once on the bone-jarring stones, the intensity is ramped up with riders riding well over threshold on sectors that typically are much longer than the hellingen in Flanders. 

The frequency and distances for the Hellingen in the finale of Tour of Flanders with power data provided by a certain Mathieu van der Poel.

Secondly is frequency. Analysing Strava data, we can see the final 123 km of Flanders features 18 sectors of hellingen averaging 2 minutes 30 seconds for each with an average of 7 km between them. At Roubaix, the sectors come thick and fast. In the final 116 km, the riders face 23 pave sectors, with an average duration of 3 minutes 6 seconds and with just 5 km between them. 

With such an intense race to get onto the cobbles, the challenge of riding the pave, and the sheer effort and duration required on each sector, it’s no surprise we see a continual stream of riders distanced from the Roubaix peloton. Upfront in the breakaway, the fight for position is almost non-existent (depending on the size of the break) and the intensity is much more controlled with fewer surges. As such, the drafting benefit of riding in the peloton versus the penalty of riding all day in the break is greatly reduced. 

The much more frequent cobble sectors in Roubaix. Power data for Florian Vermeersch’s Strava file from his second place ride in 2021.

While the favourites still most often catch the escapees, quite often it is a significantly reduced and select group that makes it to that point. The selection has already happened, and the riders in the break have escaped the war of attrition in the peloton. With fewer riders remaining and a smaller “fight” for the fewer sectors remaining, although hugely fatigued, the riders from the early break have a chance of surviving to the velodrome. 

The riders

To say the breakaway’s hit rate in Roubaix is solely due to the course would be an injustice to the riders themselves. The Roubaix breakaway is rarely a small group – the past two escapes numbered 29 and 20 riders, but not all riders survive to the finish. It takes a unique type of rider for this unique race and breakaway. So what kind of rider can sustain the intensity, battering, and focus of all 30 sectors?

We analysed every breakaway in Paris-Roubaix since 2005, looking at the number of riders in those breakaways, their number of Roubaix participations, finishes, and breakaway appearances in the Hell of the North for every rider who achieved a top 20 result. We also included each rider’s age, height, weight, number of pro wins, and ProCyclingStats ranking for the year prior to their breakaway appearance to determine the average Roubaix breakaway rider. 

Every breakaway in numbers, top 20s, and top 10s of the past 16 editions.

Who gets on base in breaks? 

A total of 207 riders have “made it up the road” since the riders rolled out of Compiègne for the start of the 2005 edition. In all bar five of those 16 editions at least one rider from the break has secured a top 20. In fact, a total of 36 riders have gone on to achieve a top-20 result while 16 crossed the line within the top 10. That’s a 25% return rate. We can’t think of another race with a similar rate. 

Somewhat surprisingly, those 36 riders who have gone on to finish in the top 20 of arguably the toughest race of the year have an average ProCyclingStats ranking of 397th. Riders need to be obscure enough to be let free of the peloton, but strong enough to survive the 50 km+ of pave.


While it is not impossible for a debutant to achieve success in Roubaix – as proved most recently by Florian Vermeersch finishing second from the 2021 breakaway – experience does help. Our research shows five appearances and four finishes seem to be the sweet spot for the breakaway survivor.

As for career success, in true Vansummeren and Hayman underdog style, the data shows the riders most successful from the Roubaix breakaway have, on average, just six pro wins to their name. Again, good enough to be strong enough, but not so good to be a marked rider. 

While Roubaix is a race that attracts specialists, surprisingly few riders enjoy multiple appearances in the Roubaix break. Marco Haller, Jelle Wallays, Mat Hayman, Alexis Saramotins, Mitch Docker, and Andreas Klier are the only breakaway top-20 finishers with more than one breakaway appearance. Still, there is a Roubaix rider type. Our data shows the average Roubaix breakaway rider is 30 years old, 1.85 m tall, and weighs 75.8 kg. 

In short, if you are 30 years old, 1.85 m tall, 75 kg, have started five Roubaix and finished four, have six pro wins to your name, were ranked ~400th in the PCS rankings in December past, and have previously ridden in the Roubaix breakaway, please make your way down to Compiègne before Sunday morning. In true Moneyball fashion, it’s not the big names who can succeed from the breakaway; rather team managers should look towards the journeymen and domestiques of the peloton for a surprise result on Sunday. 

The 2022 breakaway?

Who then might Billy Beane and Peter Brand pick out from the 2022 startlist as potential surprise packages to get on base the podium? Well, let’s look at our average top-20 Roubaix breakaway rider who fits the bill from the 2022 start list. We decided to extend our sample size slightly by expanding the age, weight, and height ranges, allowing any number of starts and finishes above the standard five and four, and widening the ranking position from 300th-500th.

There are many riders who fit the bill – including some big names – who tick some of our average boxes. Victor Campenaerts, Tim Merlier,  and Matteo Trentin all tick a single box on our list. Kenneth Vanbilsen, Marco Haller, Jasha Sütterlin, Christophe Laporte, Alexis Gougeard, Luke Rowe, Tom Scully, Guillaume Boivin, Jens Keukeleire, and Edvald Boasson Hagen all tick a few more. Big names like Silvan Dillier, Oliver Naesen, Adrian Petit, Hugo Houle, Luke Durbridge, Yves Lampaert, Edward Theuns, Nikias Arndt, Jasper Stuyven, Dylan Van Baarle, and Mike Teunisse all score highly. 

However, two riders stand above the rest. With 14 starts, 12 finishes, 22 pro wins, ranked 249th on PCS at 1.81 m tall and 74 kg, Heinrich Haussler (Bahrain Victorious) is a clear favourite for this year’s race. Only one rider ticks every box, though. Step forward Maciej Bodnar, the perfectly average Roubaix rider to write his name into folklore with a favourite-defying breakaway to victory. With DNFs in almost every race he has started this season, it seems a long shot, but Bodnar gets on base. 

One interesting side note to finish. We also analysed every edition of the Paris-Roubaix Femmes and found that the solo breakaway has a 100% success rate …