Welcome to the Science of Sport, where we bring you the second, third, and fourth level of analysis you will not find anywhere else.

Be it doping in sport, hot topics like Caster Semenya or Oscar Pistorius, or the dehydration myth, we try to translate the science behind sports and sports performance.

Consider a donation if you like what you see here!

Did you know?
We published The Runner's Body in May 2009. With an average 4.4/5 stars on Amazon.com, it has been receiving positive reviews from runners and non-runners alike.

Available for the Kindle and also in the traditional paper back. It will make a great gift for the runners you know, and helps support our work here on The Science of Sport.

Friday, July 15, 2011

Tour de France: Luz Ardiden analysis and thoughts

Analysis of Luz Ardiden: ± 5.9W/kg and thoughts

It's a day late, for which I apologize (grant proposals due at work!), but hopefully better late than never, some analysis of the first big mountains of the 2011 Tour de France.

The race's first foray into the Pyrenees went some of the way towards 'cleaning up' the top page of the overall GC, but also left many questions unanswered, and introduced more than a few new ones!

Obviously, with three more mountain-top finishes and a time-trial to go in this year's race, this was never going to be the race's decisive day, but given the profile and the stakes after a tough first 11 stages, much was expected.  At the very least, yellow was expected to change shoulders, but even that did not happen as Voeckler produced a brave, inspired ride on Bastille day.

In the end, the stage is tricky to analyze and interpret, because at times, it was done quite conservatively.  The fireworks that were expected never really came, at least until about 4 km to go, when the first real testing moves came from Andy and Frank Schleck.

The pace of the lead group - the time gaps tell the story

The rhythm and intensity of the climb is best judged by looking at the time gaps between Sammy Sanchez, the eventual stage winner, and the peloton.  Jeremy Roy and Geraint Thomas began the climb with a lead of just over three minutes, while Sammy Sanchez and Vandendert had left the peloton and began the climb with a 20 second advantage.

The 3 minute lead was cut to 1:43 with 10km to go, and it seemed that the race would come together before long.  Sanchez and Vandendert were however edging ahead of the peloton, and when they caught Thomas and Roy at 8km to go, the gap was 44 seconds, meaning that Sanchez and Vandendert had grown their advantage on the yellow jersey group by a further 24 seconds over the first 5km of the climb.  At that stage, there were around 20 riders in the yellow jersey group (there had been 30 a full 3 km into the climb), which itself tells a story.

Meanwhile, back in the peloton, Basso and Sylvester Szmyd of Liquigas had gone to the front with about 10km to the summit.   Their presence may have nudged the pace up very slightly, but it did little to erode Sanchez's lead.  It dropped to 40 seconds with 7.5km to go.  The Liquigas drive was also the impetus behind thinning the lead group slightly  - it had been cut to 18 men with about 5 km to go, an extremely large group so late in the climb.  One of those to find the pace slightly too hard was Chris Anke Sorensen.  Immediately after the stage,  I was very eager to see his power output data because that would confirm the initial impression that the pace at the front of the race was solid without being exceptional.  More on that below...

But over this period, from 8km to 4km to go, the peloton actually slowed down.  Sanchez's lead grew from 40 seconds to one minute, then even more, and with 4km to go, the gap had grown to 1:20.  That means that Sanchez and Vandendert covered a 3.5km stretch 36 seconds faster than the big group chasing from behind.  It was quite clear that the tempo at the front had dropped.  And as well as Sanchez and Vandendert climbed, and as pedigreed a performer as Sanchez is, their climbing performances relative to those of the GC riders reveal that for all the talk and hype about the "brutal effort on the front", the top riders were not throwing every weapon at the climb.  When that finally happened, it was dramatic, and a 1:15 lead would be cut to only 30 seconds within about 2km, around 25 seconds per kilometer (and even more for Frank Schleck).  But on average, the climb today was a shade below the level we may see in future climbs.

Schleck attacks and the race hots up

The attacks, as mentioned, came in the final 4km, when the Schlecks did the expected and alternated going off the front.  The elite group was now cut to only eight men.  The fast-slow tempo with the attacks allowed others to rejoin, and by 2.6km to go, the gap to Sanchez had grown back up to 1:14.  Then Schleck made the decisive attack, and was able to go clear.  This was the catalyst to suddenly reduce the lead - it fell rapidly and with 500m to go, and Schleck had basically bridged the gap to Sanchez.  Sanchez responded, sprinting away to take the stage win, but Schleck had effectively knocked away a 1:20 second lead over 2.5 km.  The rest of the field were only 20 seconds back.

That field did NOT include, most notably,  Alberto Contador.  He lost 13 seconds to the main peloton, apparently all in the final two kilometers.  It doesn't augur well for Contador, because this was not an all-out effort by his biggest rivals.  Their efforts were hard, certainly, and it would be naive to think they had too much more, but the intensity could have been lifted slightly, and that suggests that unless Contador can recover remarkably well, he's in for a very tough few days, first on Saturday and then in the Alps.

A look at the power output values from yesterday will confirm this...

The power output analysis - under 6 W/kg, even at the front

First, take a look at the direct measurement (always the best) of power in Chris Anke Sorensen.  As mentioned, Sorensen stayed with the front of the race for the first 5km on the climb to Luz Ardiden.  He then dropped off soon after Liquigas went to the front.  His power output values, especially for the first 5km, would tell us exactly what kind of power output the lead riders were producing.

Very important point that I must make here - after Sorensen fell off the lead group, the pace of the Big 5 (and Voeckler and co) did not actually increase.  If anything, it slowed, as judged by Sanchez (who himself may have been slowing, remember).  So when you look at Sorensen, you're seeing a measure of the higher range of power outputs recorded on the climb, with the exception of the attacks that came later.

So here is Sorensen on Luz Ardiden (click to enlarge):

The key points:
  • The first 5km, when Sorensen rides with Contador, Schleck and co, the power output in the front is ± 5.8 W/kg (376 W)
  • Sorensen drifts off the back soon after, and completes the stage (final 8km) at 5W/kg
  • His heart rate reaches 172 bpm.  You may recall that on the climb of Super-Besse last Saturday, he attacked on the final climb, having ridden the first 6 min at 6.4 W/kg, and his heart rate there reached 176 bpm.  So this was pretty close to a maximum effort.
  • Once he dropped off the lead group, and settled down to the 5W/kg, his heart rate dropped to around 160 bpm, which is approximately 90% of maximum (I think it's safe to say his max is 176 to 180 bpm)
Frank Schleck's power output

Unlike Sorensen, none of the big GC contenders provide their power output data, but it's still possible to estimate power output knowing the time taken, the gradient and the distance.  This has been done (courtesy halamala over on the cyclingnewsforum), and the calculation is below:

Change in elevation: 979 m
Distance: 13.3 km
Time: 37:21
Weight rider: 67 kg
Total weight: 75.0 kg
Weight bicycle, clothes etc: 8 kg
Grade: 7.4 %
Average speed: 21.3 Km/h

Calculated power output = 394.6 W, or 5.9 Watt / kg

The same calculation over the final 10km, incidentally (where it's steeper, and where more of the climb includes Schleck's attack), produced an estimation of 396 W or 5.9 W/kg.  That is, basically the same.

Implications and testing the estimation

Of course, there is always error in these calculations, and situations where the actual power output will differ from what is estimated.  Wind and drafting factors are particularly important.  However, the data from 2011, courtesy Sorensen, give some real support to the accuracy of the calculations for Schleck.  Remember, Sorensen was riding at 5.8 W/kg to stay in the group for 5km.  Once he drifted backwards, the overall pace was probably slightly slower, until the attacks (distance of about 4km), when it picked up again.  So to estimate that Frank Schleck rode it at 5.9 W/kg over the entire climb seems reasonable to me.  If anything, it's too high.  It would be similar to Sanchez and Vandendert, and their subsequent performances will be interesting to see.

And again, part of the discrepancy might be explained by people saying that the race tactics were in play, that the situation was different, and that Frank Schleck and the leaders were not riding as hard as possible.  And sure, this may be true.  But if you think that can bridge the 10% gap between the likes of the 2003 performances and what happened in 2011, well, I find this optimistic. 

How much faster can the top men ride?  An increase of even 0.1 to 0.2 W/kg, which would project to a sustained power output of 6 - 6.1 W/kg might be expected to give a rider an additional 3 to 4 seconds per km, and they would cover the final 10km of a climb around 40 seconds faster than we saw today.   

For comparison's sake, when Lance Armstrong rode up to Luz Ardiden on the way to winning the 2003 Tour (the famous stage with the fall), his time for the final 10km was 27:08, a full 2:02 FASTER than Frank Schleck yesterday (again, thanks to halamala for the numbers).  The estimated power output for Armstrong then was 458 W, or 6.4 W/kg.

Where all this leaves us is with a couple of realizations:
  1. The big shake-up is yet to happen.  To repeat, I don't think we're going to see massive increases in the power outputs from Luz Ardiden, but I do expect that the climb to Plateau de Beille and Alp d'Huez, for example, which are of similar length (37 to 42 minutes) will be climbed slightly faster.  Look for 6-6.1 W/kg on those slopes, and if that seems a trivial increase, it's enough to create gaps of five to ten seconds per kilometer compared to a guy riding 5.8-5.9 W/kg.  So the big effort still awaits.
  2. All of this further confirms, to me anyway, that there is a "limit" to how much power a rider can produce going up a hill.  I've given the theoretical argument why this is the case before, and I estimate that around 6 to 6.2W/kg on these long climbs is the ceiling for normal performance.  But now, as we get more and more data, we're seeing the same thing emerge over and over - guys just do not ride these HC climbs to finish stages at 6.3, 6.4 W/kg or higher. 

    This is obviously a topic worthy of much more discussion, but that is for the gap between the mountains, when I'll look at it in more detail.
Right now, attention turns to the big stage on Saturday, which finishes on the Plateau de Beille.  As I said right at the outset of this post, there are some unanswered questions and some new questions from Luz Ardiden.  Contador's ability on the climbs is probably the big one - was his bad day a carry-over from the first week, in which case one extra day might have allowed him more recovery and he'll return to better form come Saturday?  Or was his performance the carry-over from the Giro, in which case things will only get worse?

One thing is for sure - if his form is off by even 5% thanks to the Giro, then the Plateau de Beille will expose it more than Luz Ardiden will, because the tempo at the front is likely to be harder (only slightly, but it counts!), and the attacks likely to come slightly earlier.  But don't expect big, decisive moves at 13km to go - those days are gone, along with the days of 6.5 W/kg climbing performances.

The other big question is how do Evans and Basso compete against the alternating attacks by the Schlecks?  Their tactic seems reasonably certain, and it's to go one and then the other, until eventually one move sticks.  Basso and Evans have never really been known for their ability to accelerate on the climbs, as good as they are in the mountains.  And so Contador's apparent lull in form does threaten to make the race a Schleck affair, because Contador was the one rider who could counter-attack.  It'll be interesting to see if the Schlecks just continue to chip out 30 seconds advantages on each stage, up to the time-trial when the race will be decided.

And Voeckler?  Great effort to keep yellow, but will it cost him on the second mountain-top finish?  Only time will tell!

Enjoy the racing!


P.S.  One last word on power - Chris Anke Sorensen's power output data is available for all three climbs in yesterday's stage, and you can see it here.   

Just to summarize:

La Horquette d'Ancizan, 9.9km at a power output of 338 W or 5.2 W/kg

Col du Tourmalet, 17.1 km at a power output of 349 W, or 5.4 W/kg

Luz Arideden:  5km at 376 W (5.8 W/kg) and 8 km at 325 W (5 W/kg)

And finally, here is the climb of the Plateau de Beille, to set the scene for tomorrow's finish: