When JV and some others claim the peloton is much cleaner, to the point where doping is not essential to winning, one of their main lines of evidence is that power stats on climbing stages are lower than they used to be. Just today, JV claimed that his rider Talansky averaged about 5.9 watts/kg on the final climb of the most recent Vuelta stage, and that JRod/Contador averaged about 6.1 watts/kg. The latter value, he said, was about what the 15th placed rider would do on a typical TDF stage in 2001.
I haven’t made the calculations to confirm this, but will take him at his word. But one of the problems in making comparisons like these, of course, is that there are so many variables that can’t be controlled in comparing different performances. A different climb; different terrain in the stage leading up to the climb; unevenness in the gradient of the climb; different race tactics; different weather conditions; and maybe some others. All of these factors can easily add 10% or more uncertainty to power calculations, which obscures a lot of the real differences in power that we are trying to determine.
The simplest way to avoid all or most of these issues, of course, is to put the riders on a machine, but riders and teams tend to be secretive about this information. Suppose, though, that power output on a machine were part of a GT? Suppose instead of a prologue, riders had to go all out on a simulated climb at some defined and perfectly even grade, say, 8%? The test could be set up so that each rider had to climb a set distance, so time gaps could be recorded. Since this exercise would constitute a stage in the GT—basically, a mountain ITT--and there would be noticeable time gaps, there would be a lot of incentive to go all out. Bingo, we have very clear power data for the entire peloton (or at least the GC contenders), which could be directly compared for different riders, and for the same rider at different times. If organizers agreed, the exact same test could be run at every GT.
I realize this doesn’t sound like real, outdoor competition, and the fans would object. One way to get around this is to have this “stage” be just one part of a two-stage day, perhaps run in the morning of the opening day, with the regular prologue run in the afternoon or evening. Since the latter is usually short, a short or medium-distance climb earlier in the day would not put an undue burden on the riders. It would have to be long enough to result in serious time gaps, and to get meaningful power data, but not so long that it would take too much out of the riders for the later stage. It would also be a wonderful way to set the stage for the GT. The rider with the highest watts/kg would be considered the favorite, at least on the climbing stages.
Another possibility is to have a real stage, on some climb that is fairly even in gradient and fairly well-protected from at least the worst wind conditions. This same stage would be held every year—again, perhaps part of a two-stage day—so that valid comparisons could be made over time. There might still be some weather factors, but by running the same stage year after year, with a couple of hundred riders, it should be possible to estimate how much they were affecting results.
I have long been puzzled that it’s been so difficult to establish if riders are cleaner today, and if they are, to what degree. A test like this should be able to determine power output to a variability of 1-2%, considerably less than what a typical transfusion is thought to supply. Furthermore, by tracking data over time, we might be able to augment the ABP; a suspiciously large increase in power over a fairly short time might be a sign to target some rider. Riders like the Sky bunch can get away with such increases now precisely because the differences in actual riding conditions can sufficiently muddle the situation so that no one is sure. JV will argue that Sky dominates because the peloton is cleaner. Is he right? Data under more controlled conditions that supported this would be much more credible.