How To Win Paris-Roubaix

There are some crucial differences between the Tour of Flanders and Paris-Roubaix (enough to make for a second post, later), but the most obvious and important distinction is the lack of hills. While this observation is off-the-charts banal, it leads to what makes Paris-Roubaix so interesting: the inability to forecast any particular strategy. And so begins my non-preview Preview...

Face it, as wonderful as De Ronde is, the presence of certain climbs results in some pretty conventional wisdom about how to strategize a winning move. And this conventionality is the only thing I don't like about the race. Everyone gets twitchy around the Valkenberg, the Muur, or whatever is in between, leading to thrilling attacks... or stalemate.

Paris-Roubaix has its cobbles, of course, and they tend to compartmentalize the race, but in the last 50km there's no one place where the course dictates strategy, where an attack can be expected or ruled out. There are no crushing inclines on the horizon to wait for before testing your rivals. Instead, all you've got is a long, narrow, flat, bumpy course into a head- or side-wind... in other words, it all sucks, so you might as well make your move whenever you see tongues wagging around you.

As a result, Paris-Roubaix strategy is somewhat more heavily reliant on pure instinct. The one thing you can count on is that the cobbled sections will slowly reduce the field. The Wallers-Arenberg stretches will detonate the back half of the peloton, at least, with 90km to go. And the Mons-en-Pévèle section at T-minus-50km will likely shed all but the strongest contenders, assuming their steerer tubes are still in one piece.

Who makes the selection at that point, and in what shape, is impossible to predict. And from that moment on, how to win the race is entirely dependent on who's in the front groups and how they're feeling. It defies forecasting and replaces it with the riders' own split-second judgments. How cool is that?

Looking back a few years (prior to which the, um, purity of the product becomes a tad murky) gives a sense of where to "expect" a move:

• Last year, Stuart O'Grady sauntered away from the field near Bourghelles, km 234 (of 260), and just after sector 6. [Remember, the cobble sectors are numbered backwards from 28 to 1.]

• In 2006, Fabian Cancellara bridged a gap to the leading attacker, Vlad Gusev, at Camphin-en Pévèle (km 239/sector 5). He dropped Gusev shortly after, never to be seen again -- train or no.

• In 2005 Boonen, Hincapie and Flecha reached the 'Drome together. They were part of a dozen leaders to survive Mons-en-Pévèle break, and the group slowly winnowed itself down to final 3 at Carrefour de l'Arbre (km 242/sector 4).

• 2004 was pretty much the same story as '05, with Backstedt beating a 4-man group in the Velodrome.

• In 2003 the key split happened before Cysoing (232), between sectors 7 and 6, while Van Petegem bridged at Carrefour (sector 4) and won the sprint.

Naturally, the closer you are to the Roubaix Velodrome (one of my five favorite buildings on the planet, along with the Ice Hotel, the Madonna del Ghisallo church, a certain villa in Sicily, my house, and, what the hell, something by Frank Gehry), the more likely you are to see a decisive move. But even when the winning move goes it's hard to spot. Just when you see someone jump, the next section of stones arrives and a countermove overtakes the earlier one. No more careful Ronde-style riding; this is a slugfest.

For more info on the cobbles... I prefer Le Tour's official cobble sector listing but it doesn't include the sector star-ratings. CN's does.