The Sprint

There’s a sequence in the great La Course en Tete, the documentary on Eddy Merckx, in which the filmmakers show a sprint unfold in a head-on view as the riders barrel toward the finish line. To a rider, their cadences are north of 120 rpm and they rock their bikes side-to-side hardly at all. In fact, each and every sprint in the film or any film of the era have something in common. The bunch sprints of the day are displays of redline agility, typified by furious pedaling in wound-out gears.

In another segment in the film the Cannibal climbs on a set of rollers in his garage. They consist of large drums, at least 15cm in diameter, and Merckx rolls them up to centrifuge speeds until his bike is bouncing on those drums. For those who watch carefully, Merckx’ frame flexes under his effort and his bottom bracket sways as if there was a wind stiff enough to torque the frame.

To my eye, these efforts feature two limiters: agility and frame stiffness. Cultivating the agility necessary to pedal 150 rpm is an incredible challenge for many riders. To hold that kind of cadence for more than a second or two is superhuman. It’s fair to ask why riders needed to pedal that fast. The fact is simple: the only way for a rider to generate his full wattage was to do so by pedaling a relatively small gear at gyroscopic speeds. Big gears caused riders to flex the frame too much, making the bike harder to handle. Rocking the bike exacerbated the flex issue.

So what has changed in 25 years? Quite a lot, when you add it all up.

First is the frame and fork. The difference between a 28.6mm-wide down tube and a down tube measuring more than 60mm in diameter is enormous, but that change didn’t happen overnight. Increases in stiffness occurred gradually and an increase in stiffness in one part of a bike showed a weakness elsewhere in the bike. It was only after frames became stiff enough that you couldn’t make the chain rub the front derailleur when in the big ring that frame twist became an issue.

But of course, the frame and fork were only the starting point. Rims were beefed up, and spoke tensions rose, decreasing the need to tie and solder them at crossings. Bars and stems got a dose of stiffening as did crank arms. And let’s not forget the changes to both pedals and shoes that increased a rider’s ability to deliver power to the bike.

All these changes can be summed up in a single part: The 11-tooth cog. Greater stiffness meant that bigger gears could be used, and while for the most part sprint speeds did inevitably rise, the bigger change was that more meat in the drivetrain brought cadences down.

It’s inaccurate to say that the riders are more agile now, but the drop in cadence has given riders a greater degree of control. While I haven’t conducted a survey of sprint finishes over the last 40 years, I wouldn’t be surprised if you found that crashes in final sprints had actually gone down.

More than anything, what changes in bicycle technology has given us is the ability for an athlete to show his full potential in the final 200 meters of a race. A sprint, after all, should not be defined by a rider’s ability to refine movement and flex the bike no more than can be controlled, rather, it should be a measure of his full power.

Cavendish Image: John Pierce, Photosport International



  1. Marco Placero

    Lightness, stiffness, and comfort– ahh the benefits of living on a carbon-based planet.
    Reflective thinking on Padraig’s post, one wonders whether changes in training methods have also contributed to decreased sprint cadences. Coaching methods crossing intersport bounds may have increased muscle fiber recruitment ergo leg power.

  2. randomactsofcycling

    Perhaps the technology has contributed to the lower cadences in the sprints but that can’t explain the vast differences in style between the ‘stompers’ like Boonen and Cavendish and the more fluid, high cadences of Petacchi and Zabel.
    Or could it be that the bunch sprint now is almost exclusively the domain of the sprinter whereas in the past there was less specialisation in cycling and the GC riders also got mixed up in the madness? Not too many GC riders would have the outright power to pedal the big gear/low cadence, hence in the past the sprints had higher cadence.
    It’s an interesting topic. Has anyone got any old track footage? I wonder if cadences on the track have changed to the same degree?

    1. Author

      Randomactsofcycling: While I agree that Zabel’s sprinting style harkens back to a previous generation, Petacchi and Cipollini both pedaled at noticeably lower cadences than sprinters from Merckx’ time, though certainly not as low as McEwen. I, too, have been curious about why GC riders just don’t contest bunch sprints at all anymore. It stopped after LeMond. In ’89 he and Fignon were still contesting sprints on flat stages. It may be that Indurain proved it just wasn’t necessary to dominate the way Merckx and Hinault did. As for the track, there are some good reasons for high cadences there: You can’t rock your bike in close quarters and higher cadences allow for more sudden acceleration, which is much more critical for getting away on the boards.

  3. Charles Cushman


    That track question intrigued me since it took a while for me to find my optimum gearing at the track without spinning out or wasting energy when trying to accelerate. According to what I could find on the web, Theo Bos used 104 gear inches, but if you watch his world record on youtube he is still sitting in the saddle doing a high cadence sprint.

    I am guessing that on the track since there is no lead out gearing can’t get too high because you have to build your own momentum, and in races such as points, madison, and scratch the gearing has to be low enough to be able to accelerate efficiently. Maybe in the Kieran the gearing can larger.

  4. Curt Goodrich

    Interesting premise but I think you reached a little too far. Comparing 28.6mm steel down tubes to 60mm carbon fiber down tubes no conclusion can be made. One can’t just isolate the OD for comparing the stiffness because the two down tubes are different materials.

    Rims have been beefed up because spoke count has gone down. Spoke tension has gone up because spoke count went down. Wheels have gotten more aerodynamic because of these changes but lateral stiffness hasn’t changed radically.

    I’m not arguing stiffness doesn’t matter. I just don’t conclude that the changes we see in bikes are good or bad. Sometimes they’re just changes. Also this sort of analysis feels like the arguments about who would win if they fought- Joe Lewis or Ali? There’s no answer to be found because they didn’t fight in the same era. Same for why didn’t Merckx ride an 11 tooth cog? There just weren’t any 11 tooth freewheels. To conclude anything from that seems to be reaching.

    I think the cadence observation is interesting and I dig your writing. Thanks for a great blog.

  5. Touriste-Routier

    There are two other important change in sprinting between today and the Merckx era; changes in the team structures and the advent of the leadout train.

    As chaotic as they are, I think today’s sprints/leadouts are better controlled by teams than in previous generations. While leadouts certainly existed previously, the trains weren’t as big, and the build and leadout phases were not as long or structured. Furthermore the riders today are of similar caliber and ability, whereas previously there was greater disparity between the stars and the domestiques.

    As for the changes in technology, it would be interesting to see a study comparing older bikes and the current ones. We can conjecture all we want, but a controlled study utilizing power meters, accelerometers, etc. would reveal a lot in regards to stiffness and net power delivery.

  6. roomservicetaco

    Interesting article. Not disputing your conclusions regarding sprints, but if frame flex were such a big issue, why is it that riders in the days of Mercyx climbed at very low cadences/tall gears instead of spinning up them?

    1. Author

      Touriste-Routier: Agreed, today’s leadout trains have contributed to greater control, and yes, there was a great deal more disparity between the stars and water carriers.

      Roomservicetaco: The issue with the low gears in years gone by was a technical one. The rear derailleurs could only handle a 13 or 14-tooth differential between the largest and smallest cog. Additionally, the front derailleurs could only handle a maximum of an 11-tooth differential between big and small chainrings. A bike with a 52×13 generally didn’t have a gear lower than 42×26.

  7. MattS

    For me this discussion highlights the importance of matching a bicycle design to the rider’s style. Jan Heine has tested numerous vintage bikes, which you can read about in Bicycle Quarterly. He has become rather enamored with standard diameter tubed frames. In one test, he found the lightest and most ‘flexy’ frames climbed the best (seated). He has mentioned to me that he’s working on learning how to ‘sprint’ in the saddle, hearkening to the bygone era Padraid speaks of here. This post clarifies in my mind why this approach is best suited to such ‘whippy’ frames. Since not everyone sprints their bikes, I think its worthwhile for riders to think about whether their style demands a 60mm downtube. I personally like to hammer sprints standing up on my ‘road’ bike, so more ‘stiffness’ is probably good. But on my ‘allroad’ bike I rarely spring, but need good traction and efficiency on the climbs (and of course everywhere else). Comfort is exteremely important when riding all day on dirt roads. Horses for courses.

Leave a Reply

Your email address will not be published. Required fields are marked *