When conspiracy theorists examine data that tend to support their theories, they use them copiously to prove their point. That has been the case for decades with those cynics who try to connect the increased speeds in professional cycling with increased (or more sophisticated) use of doping products. Those theorists say that the faster average speeds seen at grand tours through the past couple of decades were all the proof they needed—and they would debunk the counter-arguments that racing has become faster because of smoother road surfaces, lighter bikes, fitter athletes and the more intense competition each day for stage wins in those grand tours.
Conveniently, the theorists tend to ignore the more reliable results obtained from studying the speeds in one-day classics, each of which take place on near-identical or similar courses every year. Take the case of Paris-Roubaix.
If you watched the NBC Sports broadcast of the French über-classic last Sunday, you probably heard Paul Sherwen speculate at one point whether the event’s record speed would be beaten this year. The British commentator made his remark because the first two hours or Sunday’s race were ridden at more than 48 kilometers per hour (kph) and the course record stands at 45.129 kph, set by the 1964 winner, Dutch legend Peter Post. In the end, Tom Boonen’s average speed for the 2012 Paris-Roubaix was 43.476 kph. Hmmm….
On first look, it would appear that racing is getting slower, not faster, but the big difference between Boonen’s ride this year and Post’s 48 years ago was the course. For the first seven decades of its existence, Paris-Roubaix actually started in Paris (in the northern suburb of St. Denis) and used mostly national roads (which were still cobbled through the first part of the 20th century) on a fairly direct south-to-north route. That changed in the late-1960s when the start was first moved 30 kilometers out of Paris to Chantilly, and then 75 kilometers northeast of the city to Compiègne.
The course was changed because the cobbled main roads of the early route were all getting resurfaced with asphalt, leaving just short stretches of pavé in the latter parts of the course. By moving the start (Compiègne is only 160 kilometers from Roubaix on the direct route), the organizers had an extra 100 kilometers to play with, seeking out the oldest sections of cobblestones on narrow farm roads, including the infamous Forest of Arenberg, much farther to the east than the old course.
And so, if we are to compare apples to apples, we have to study the race speeds on the two versions of Paris-Roubaix separately. On the old, more direct course, the prevailing southwest winds often provided a tailwind all the way from start to finish. And so, through the postwar years, there was a spike in the average speed when the wind was most favorable. Rik Van Steenbergen set an all-time record in 1948 at 43.612 kph—which, interestingly, is faster than Boonen’s speed this year!
In ensuing editions, both Germain Derijke in 1953 (43.523 kph) and Pino Cerami in 1960 (43.538 kph) came close to the record before Post broke Van Steenbergen’s mark in 1964 at 45.129 kph. Factors in that speed were a tailwind, a strong, harmonious four-man breakaway, and the lack of cobblestone roads.
So, you may ask, why are we comparing speeds set on a course that no longer exists with one that contains 27 sectors of true cobblestones totaling 51.5 kilometers? The answer is: we’re not. Those faster speeds from a half-century ago provide the background to what has been done at Paris-Roubaix in the four decades since the route was changed.
If you believed the conspiracy theorists, there should be a continual increase in race speeds from the 1970s onwards, perhaps with a spike in the 1990s when the use of the blood-booster EPO became widespread. That’s not what we see. Let’s compare two similar editions of the Hell of the North classic. Last year, the Roubaix winner was a Belgian journeyman team worker named Johan Vansummeren, who, on a course of 258 kilometers, averaged 42.126 kph. Going back 40 years to 1971, the race was won by a Belgian journeyman team worker named Roger Rosiers, who, on a 266-kilometer course, averaged 42.108 kph. Virtually identical performances and identical speeds, 40 years apart.
The first time the 43-kph barrier was broken on the current route was in 1980, when Italian great Francesco Moser averaged 43.105 kph in his winning solo victory. That course record stood for 10 years until it was improved incrementally by Johan Museeuw to 43.305 kph in 1990. And that speed remained at the top of the charts for another 18 years until Boonen’s winning 43.406 kph in 2008—again, just a small improvement. And last week, in a race that was fast from start to finishing, and ended with his phenomenal 56-kilometer lone breakaway, Boonen went slightly faster at 43.476 kph, which is now the record for the current course.
But this is only part of the story. Another reason why speeds at Paris-Roubaix have been consistent on the new course over the past 40 years is because the weight of the bike is not much of a factor, as it is in hilly races. In fact, last Sunday, the French runner-up Sébastien Turgot rode a cyclocross bike, a Colnago Cross Prestige, which weighed 7.8 kilograms (17.2 pounds) compared with a road bike’s usual 6.8 kilograms (15 pounds). And that two pounds heavier bike didn’t stop him from out-sprinting Alessandro Ballan for second place.
A few years ago, a French statistician presented a paper to the French Olympic Committee that mapped the average winning speed of 11 major events (the three grand tours, four weeklong stage races and four classics) since the start of bike racing in the late 19th century. The major conclusion of that study was that, averaged out over the 11 races, there was a 6.2-percent increase in speeds between the late-1980s and the early-2000s. And EPO was cited as the most likely cause of that increase.
But none of other changing factors—lighter bikes, et cetera—was fully explored, maybe because they didn’t fit the theory. For instance, the French scientist’s chart showed that the average speed in the Giro d’Italia has remained consistent for the past four decades, while the Tour de France speed has gradually increased.
Those results can easily be explained by the nature of the two grand tours. The Giro has more mountain stages and is less competitive in terms of team strength, whereas the modern Tour has stronger teams and more serious contenders, and every stage is raced like a classic. Even so, as long ago as 1981, the Tour average was 38.960 kph, and in 2007 it was 38.980 kph. Sixteen years apart, same speed. Last year, the speed was 39.788 kph, and in 1992 it was 39.504 kph. Nineteen years apart and, again, not much difference.
Which goes to show that statistics can be used to prove virtually any theory. But tying race speeds to drug use is just as unreliable as any conspiracy theory. Times change, bikes change, roads change, but the best riders always come out on top. Chapeau to Tom Boonen for his great victories over the past few weeks. Now it’s time for the hilly classics and a new set of exciting winners. That’s what really counts in this beautiful sport!
You can follow John at twitter.com/johnwilcockson