The climb to Lagos de Covadonga is a particular climb: it's not very long (12.6km), but the first 8.6 km at 9.1% of average gradient are always done at a very high speed, both because the climb starts at an altitude of only 190m and the few curves do not slow the pace of the riders.

In fact on these 8.6 km we always get some of the most impressive and highest VAM values in professional cycling: Sastre in 2005 Vuelta climbed up in 24'22" (VAM=1907m/h), with Heras, Simoni and Menchov just a little behind (1891m/h).

This year's group of favorites, with Piepoli, Sastre and Menchov took 25'30", with VAM= 1823m/h.
Wind and drafting can always affect the times of the ascent and the ensuing valuations, but assuming the body weights of Sastre, Menchov and Piepoli to be 58, 64, e 52 kg respectively, then the average watts developed would be 364, 402, and 326.

Therefore Menchov produced 10% more power when climbing compared to Sastre.
A difference confirmed by the time trial in Saragoza, 52.2 flat km with the russian's speed of 53.6 km/h versus the 51.2 km/h of the spaniard's: 2877 versus 2624 are the respective square speeds, with a difference of 9.6%.

The following 2 mountain stages confirmed a substantial equality in the uphill performances of the 2 riders; Menchov with 3' of advantage theorically is not much in jeopardy, but the road is long and he doesn't seem to have that great a team to support him...