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.

Sunday, July 19, 2009

Tour 2009: Contador takes yellow

Alberto Contador soars into yellow in Verbier: Report and analysis

Alberto Contador rode himself into the yellow jersey by dominating the mountain-top finish in Verbier today. An attack by the Spaniard with 5.5 km to go saw him open up a gap on the Tour's elite, and then continue to build it. The damages by the finish line were 43 seconds to Andy Schleck (the only rider to respond to the initial acceleration), 1:03 to Vincenzo Nibali, 1:06 to a group containing Frank Schleck, Brad Wiggins and Carlos Sastre, 1:26 to Cadel Evans, and 1:35 to Lance Armstrong.

The result leaves the overall standings of the Tour as follows:
1. Alberto Contador
2. Lance Armstrong @ 1:37
3. Bradley Wiggins @ 1:46
4. Andreas Kloden @ 2:17
5. Andy Schleck @2:26
6. Rinaldo Nocentini @ 2:30
7. Vincenzo Nibali @ 2:51
8. Tony Martin @ 3:07
9. Christophe Le Mevel @ 3:09
10. Frank Schleck @ 3:25

A level above the Pyrenees: How it unfolded

The Tour's first mountain-top finish in Arcalis just over a week ago failed to produce the fireworks. The same can't be said of this one. At the bottom, Saxo Bank charged into the lead and the tempo was pretty close to maximal from the outset. The peloton was systematically shredded under Saxo-Bank's pressure, and with about 6.5km to go (within 2km of the start of the climb), the peloton had been cut to about 13 riders. Compare this to the 40 who made it 7km up Arcalis and you realise that today, the Tour shifted into top gear.

At 6km to go, Frank Schleck was first to attack, and his move cut the elite down to only 5 - Schleck, Contador, Armstrong, Wiggins and Andy Schleck. Noticably absent were Cadel Evans, Carlos Sastre and Kloden, though all would gradually return to the chase group. By the time they did return, however, Contador was long gone - a fierce attack with 5.5km to ride saw him open a gap of 17 seconds on Andy Schleck and 30 seconds on the group of Armstrong within one kilometer. It was an extra-ordinary acceleration, and I wish had the data to analyse it some more detail - unfortunately, I've been unable to find a kilometer-by-kilometer breakdown of the climb to Verbier - if one knew the gradient of each kilometer, and the time, it would be possible to do some interesting analyses, so if you have this, please let me know!

Contador continued to build his lead, and by the time he hit the summit it had been increased to 43 seconds over Schleck, but perhaps more importantly, over a minute to those with aspirations in the time-trial. Contador seems peerless on the climbs, and the Tour now seems his to lose, barring a dramatic change in form or an accident.

For his part, Armstrong didn't drop back onto the wheel of the chasing riders, he rode at the front, along with Kloden. Perhaps they were simply riding tempo and allowing Contador away, but it seems odd to me that they would not just sit in the group and find a wheel (that of Evans, perhaps) to follow, given the supposed "obligation" to the team. That is the usual approach, to the best of my knowledge of what happens when a team-mate is out ahead.

Attacks come from behind

In any event, it was noticeable how the fast pace was neutralising the attacks from behind. A group of eight had reformed with 3.5km to go, but no attacks came until about 3km from the finish. Then it was Brad Wiggins, perhaps the revelation of the Tour in the mountains who attacked and stretched the group. The next attack came from Frank Schleck, in an attempt to see whether he could join brother Andy to combine forces to catch Contador. He couldn't, but his move did cause the split in the group that would eventually see them come over the line in pairs or alone.

Each attack put Armstrong more and more into difficulty, though many will have expected this, given his age. What will be interesting (apart from the race situation, of course) is to see how he manages himself through the next few days - as described below, he should face attacks in this final week.

Climbing rate of Contador - a record climb

As I mentioned, I don't have a detailed breakdown of the climb - the gradient each kilometer, for example. So doing a detailed analysis must wait until hopefully someone can provide that (any takers? Alex...?)! But until then, some
basic analysis reveals how fast this climb was:

The climb to Verbier is 8.7km long, at an reported gradient of 7.5% (according to the official Tour website. CyclingNews has it as 7.1%). The climbing times of some of the top riders were:

1. Alberto Contador - 20:36 for the climb at an average speed of 25.34 km/h
2. Andy Schleck - 21:19 (24.49 km/h)
3. Carlos Sastre - 21:42 (24.06 km/h)
4. Lance Armstrong - 22:11 (23.53 km/h)

One of the measures of climbing ability is Vertical climbing rate, which is the distance STRAIGHT UP, expressed as a rate in meters per hour.

For the climb to the Verbier, the vertical climb is 652.5 m (8.7km x 7.5%). Therefore, Contador's climbing rate is an extra-ordinary 1,900 m per hour. I say "extra-ordinary" because this is the fastest climb in the history of the Tour de France, in terms of vertical climb rate. I have data, courtesy one of our readers (thank you Alexander) that tracks all the climbs in the last twenty years, and I can tell you that the previous record for vertical ascension rate was Bjarne Riis at 1843m on Hautacam in 1996 (and we all know what powered Riis to that summit).

I am going to post on these figures during the next week (don't worry, you don't have to blindly take my word for it!) - I don't want to overload the race report with too many numbers. But the Contador climb today was incredible.

It was a short climb, taking just over 20 minutes, and of course that allows a higher intensity than the 37 minutes on Alpe d'Huez or the 49 minutes on the Tourmalet. So a higher VAM is partly expected. I would point out that there have been other 8 to 9km finishing climbs in the Tour, and many climbs this length, but none have ever come close to a rate of 1900m/hour. Race tactics no doubt played a part as well - the Tour has been dormant for about 8 days leading into this, and suddenly it exploded into 20 minutes of action. Then there is the wind, which is reported to have been a following wind. A lot of factors may have gone into today's record, but I guess it's only natural to wonder...? To those who are no doubt wondering whether that means anything in the big picture of doping control, you are not alone...

By way of comparison, Andy Schleck's climbing rate was 1,837m/hour, and Armstrong's was 1,764m/hour. Looking even more broadly, last week up Arcalis, Contador climbed at a vertical ascension rate of 1,687m/hour, and when the Tour dawdled its way over the Col du Tourmalet on Sunday, the rate was 1401 m/hour. The Tour really did step up a notch or two in Switzerland today.

Stage implications: Team leadership, podium implications and days' disappointments

The stage was always going to lay the foundations for what everyone expected would be a very brutal last week in the Tour. To a certain extent, that happened. At the very least, the debate around team leadership of Astana was laid to rest today - Contador is clearly the number 1 man, Kloden and Armstrong seemingly close for number 2. In fact, I'd give number 2 to Kloden going into week 3. For all the inevitable talk of "team obligations", Lance Armstrong would not have had the firepower to match Contador or the Schlecks in the mountains.

He may yet have enough for a podium finish in the time-trial on Thursday - lot depends on whether Astana instruct Kloden to ride the TT conservatively. If Kloden races hard, he becomes the number 2 for Astana, in my opinion - he did a great deal of work pulling today and seems to have better form now. Time will tell...

What will be interesting for the rest of the Alps and the climb to Mont Ventoux is that attention will shift to the question of whether Armstrong can respond to what should be some telling attacks from those currently ranked 4th to 8th in the general classification. Among those who will have eyes for the podium (if not for Contador, who seems imperious just at the moment) are both Schlecks, Evans, Sastre and maybe even Wiggins. All will recognize the need to attack, given the time-trial to come.

Whether they will have the strength to do this is another question - today did not exacly fill me with hope that anyone can match Contador. More worryingly for the race as a spectacle, even attacks from the group were not damaging. I have my reservations over whether anyone will be aggressive enough to attack on the next two Alpine days - Contador's aggression threw today wide open, I'm not sure it will happen again. So, with all the hype around a super-competitive final week, it might be that the riders neutralize one another all the way to Thursday's time-trial. Contador, for his part, may try to steal a few more seconds in the final kilometers, but need not attack like he did today. And with no mountain finish until Mont Ventoux, even that is unlikely.

So, looking ahead, it's a rest day tomorrow, then two more days in the Alps. Wednesday in particular is a monster stage, the so-called "queen stage", but that stage may be neutralised by the race situation and the fact that we still have a time-trial and a Mont Ventoux finish to come. The onus now lies on the likes of Schleck, Schleck, Evans and Sastre to attack - will they have the ammunition? Only time will tell...

Join us for more race analysis as it happens!


Donate and support The Science of Sport

Thank you for taking time to read and follow our coverage of the Tour de France here at The Science of Sport. We hope you've enjoyed the insights we've tried to provide! Your time and energies are greatly appreciated!

We run this site as a "labour of love" and will continue to do so, but of course, any support is greatly appreciated. So, if you would like to donate and support our continued efforts to bring you the insights and analysis, please consider donating to our site. You can do so in one of two ways:

  1. If you are reading this on our site, then simply scroll up to the top of the page, where you will find the DONATE button at the top right of the page.
  2. If you are reading this in our email newsletter, click here to be taken to our site, where you will see the DONATE button at the top right of our homepage.
Thank you once again, and keep on reading!

Ross & Jonathan


Anonymous said...

Excellent report, good analysis. 1900 VAM is beyond any credibility. And I believe he good have gone faster if he had been challenged "man to man"...Its sad for the sport!

Ross Tucker and Jonathan Dugas said...

Hi anonymous

Thank you - you're right, though. I didn't factor in that it was a solo effort and could have been faster.

I guess the optimist/naive side of me says that the climb was short and came at the end of what was a pretty sedate Tour so far, where it had effectively been "dormant" until the 20 minute unleashing. I think this has something to do with it - certainly, a great performance doesn't necessarily PROVE doping, I don't mean to imply that!

But, given cycling's history, it's difficult to see that Pantani, Riis and Ullrich, pumped full of drugs, couldn't hit the same level, even on a longer climb...

Anyway, we'll see what comes of the retrospective testing and the 'detective work'. I believe (or should I say hope) that one day, the truth (whatever it might be) will emerge!

Thanks again!

Unknown said...

O_o Wiggins!
Surprised you didn't mention him enough. This is an amazing improvement. In recent years such breakthroughs were followed by doping charges so I am still a bit afraid of things to come.

About Contador I would say indeed a great climb but his vertical climb rate is less than 5% better than Riis on Hautacam and the altitude of this climb is quite low so 2009 juice is not as good as 1996 juice.

Also interesting how Sastre and Evans got back up after dropping to the initial attacks. It is likely that they made better pacing decisions, avoiding running anaerobicly at any time. So indeed a consistent pace gets you there(Wanjiru rivals nodge-nodge-wink-wink)

Other notes would be that I don't think they will try to slow Kloden in the time trial. He may be put on to work in front of the pack and might tire a bit but Wiggins is a dark horse and nobody knows what will happen on the tt. By the way Lance said in an interview that Contador is better than him now and he would be proud to work for Alberto(a little dishonest maybe), anyway Astana will try do get the maximum possible from this tour and that might mean Kloden. I think if Leipheimer had not crashed, he would also have been ahead of Armstrong.

And it is Vincezo so correct the spelling. This guy may take the white jersey even. He is better than Schleck at time-trialing, and we do have to consider that these guys are so young that there will be high variance in their performance, everything is possible on Ventoux


Chad Black said...

Hi Ross--

Found a profile of the climb here: http://www.climbbybike.com/profile.asp?Climbprofile=Verbier&MountainID=8971

The question is, where did the Tour officials start and finish the climb, because they obviously weren't counting from the bottom, and didn't go all the way to the top.

There is another profile from the Tour de Suisse here: http://www.dailypelotonforums.com/main/index.php?s=&showtopic=8654&view=findpost&p=154260

In either case, Contador's initial attack was amazing. You could see on Andy Schleck's face after trying to bridge up after the attack just how hard the pace was. I thought I heard at one point in the stage there was a prevailing tailwind on the day. Have to check the map, but if that was the case on the climb it would be a factor.

By the way-- who was it again that popularized the concept of VAM? Wasn't it a certain doctor with real talent at getting extra-ordinary performance out of his athletes?

Ross Tucker and Jonathan Dugas said...

Hi Mircea

Thanks for the comments. You're spot on about everything

Regarding Wiggins - too little time to discuss him in more detail! His story is interesting though, and we'll see what comes. He has always been very outspoken about doping, and I'm of the belief that people give away their beliefs by how they talk about doping. The likes of Armstrong et al, you'll have noticed, never ever condemn doping in the Tour. Hamilton, Millar, all of them. Wiggins has, so I'd be surprised. But we'll see.

On the pacing, absolutely right. Sastre rode his own race and it paid off. So too did Evans, though Evans is usually quite attentive to attacks. So he either changed his approach, or he didn't have the legs initially. Regardless, his pacing benefited him.

On Kloden, fair point. Astana will probably try, I'm sure they'd love for Armstrong to be that "best result", especially given that Armstrong is so influential there. But we'll see how the attacks affect the race before that TT. Interesting one...

And finally, thanks for pointing out the spelling - I'll blame it on my haste in writing these posts.


Ross Tucker and Jonathan Dugas said...

Hi Chad

Thanks, that's awesome. IN your second link, it says the Tour finished at 1468m, which means it corresponds to the first profile finish point, so I'll be able to work backwards to match the profile up with the measured times on the climb.

Should make for interesting analysis, hopefully I can get that up tomorrow!

You're dead right about the popularity of VAM - Touche! Dr Ferrari lives on!

Thanks for those profiles!


Anonymous said...

Ross: Great work! Is it possible to quantify how much of an effect the tailwind would have on a climb like this (even to ballpark it)? Are we talking about a couple percentage point gains in speed or something much greater? This seems to be the big question for me to decide whether Contador's climb is believable.

To Mircea about Wiggins: you're right that performances like this usually raise red flags, but Wiggins is entirely different. He has a history as one of the most powerful riders on the track, but since the Olympics he's dropped 8kg in order to compete on the road. He's a totally different rider, and you'd expect him to be able to translate his past power into climbing now.

Anonymous said...

I have just a question about the finish. The last K seems to be at the Church of Verbier, meaning 7200m from the Begin of the Climb. Therefore the Climb could be considerabely shorter than you calculated it. How sure are you about the right location of the finish, and consequently the exact course?

I think this is a very important point considering the severe conclusions you draw from your calculations.

Alex Simmons said...

Given the following assumptions*:
Speed: 25.34km/h
Gradient: 7.50%
Mass bike (8kg) + rider (60kg): 68.0kg
Crr: 0.0040
CdA: 0.280m^2
Air density: 1.053kg/m^3

Average Power required:

With no wind:
422W @ 7.03W/kg (body mass only)

With a 1m/s (3.6km/h) tailwind:
408W @ 6.80 W/kg

With a 2m/s (7.2km/h) tailwind:
397W @ 6.61W/kg

With a 3m/s (10.8km/h) tailwind:
387W @ 6.45W/kg

So you can see that even on a relatively steep climb, wind has a sizeable impact on the estimated power required and has a big influence on how VAM should be interpreted.

Certainly a tailwind brings 20-min power estimate for a relatively "un-fatigued" world class climber well into the realms of the plausible.

So, knowing wind vector is pretty important.

Of course if mass is wrong, it changes the absolute power requirement but the W/kg doesn't vary by nearly that much (e.g. in the 2m/s tailwind example above, adding 1kg to body mass (+1.47%) requires 402W (+1.37%) @ 6.59W/kg (-0.29%))

And if the speed or gradient assumptions are wrong, then all bets are off.

* Note: Of the above assumptions, the most important are speed, gradient, mass and wind speed. The estimate of power is less sensitive to changes in the other assumptions.

So if someone can provide validation of the rider's speed, gradient, mass and wind speed (at rider level, weather station wind speed is not usually at rider level), then I can re-reun the numbers.

djconnel said...

Whoops... I accidently put my comment on the wrong post (the previous one).

The full route profile is available from mapmyride.com:

It shows 651 meters gained over the final 8.31 km. This, however, includes a relatively gradual beginning to the climb, which then steadies at 9%. I'm not sure where the timer begins on the climb, but 53x12.com shows that VAM increases by around 40 for every 1% increase in grade, due to the reduced influence of rolling and wind resistance. So a steep, relatively short finishing climb in a tour which has left the favorites with fresh legs is likely to to produce a VAM number significantly higher than you might get in many Alpine stages in Tour history. Still, it's extremely impressive, obviously: Sastre was 1782 climbing Vesuvius in his Giro stage win. If he did the same here, he'd have finished 80 seconds behind Contador. Instead he finished 66 seconds behind Contador, but with the benefit of more drafting.

Note the official route profile shows only 638 meters gained in 9.0 final km. It seems like the mapmyride data should be more reliable.

djconnel said...

I looked more carefully at the mapmyride map. I had estimated the climb from the start of the upward gradient, at 198.366 km. If instead I start it at 199.1 km, the sharp left turn which marks the beginning of the steep climb, total altitude gained = 630 meters over 7.68 km ( 8.2% ). Errors in the routing may change the distance but not likely the altitude gained. So this reduces Contador VAM to 1836 m/hour. This is more human, still extraordinary, but not unprecedented. Google 53x12.com for VAM numbers from history.

Alex Simmons said...

On the basis of these revised values for gradient and distance:

Speed is revised to 22.36km/h on an 8.21% gradient.

Maintaining other assumptions per my previous post:

Tailwind (m/s) ___ Watts ___ W/kg
0 ___ 391 ___ 6.52
1 ___ 381 ___ 6.34
2 ___ 372 ___ 6.20
3 ___ 365 ___ 6.09

More than plausible 20-min power from a world class climber.

IME, Map My Ride data can be unreliable. e.g. the MMR profile for the opening ITT was simply wrong. My experience with developing TT pacing strategies has shown this is a not an uncommon "feature" of MMR data.

Anonymous said...

Would the effect of a tailwind really be that large given how the climbing route is very sinuous? It seems like it would act more like a crosswind the majority of the time.

djconnel said...

True, but drafting is an effective tailwind. Contador was going around 6.6 meters/second, the others more like 6.3 meters/second, so a 10% draft benefit would be an effective 0.6 meter/second tailwind, a 0.11 W/kg advantage interpolated from Alex's calculation.

Chris said...

One factor that I don't believe has been mentioned is the quality of Swiss roads. If you watch the replay of stage 15 on Versus.com, there's a different commentator than the live stream, and he mentions the road quality several times (and of course it's obvious from watching the footage).

djconnel said...

Alex: I checked mapmyride on heavily switchbacked Old La Honda Road, my favorite local climb, and it did quite well on altitude, but was off substantially on distance, claiming a grade of 8.4% instead of the established 7.3%. I suspect something similar applies here: the altitude close but the distance short. A 1% difference in gradient is worth around 3.2% in VAM, by Ferrari's estimate, so that's a non-trivial error.

robert merkel said...

I'd be very surprised if the road quality made much difference.

I don't know the formula you guys prefer to use, but using analyticcycling.com's calculator and changing the coefficient of rolling resistance from between 0.002 and 0.005 only changes the watts numbers by about 14 watts.

And, frankly, after having ridden the routes of the Pyrenees stages of this year's tour (in June). Even on the Col de Serra-Seca (which is on an extremely isolated, lonely road by western European standards) the road surfaces were either very good or being resurfaced, and I assume that this is normal practice when the Tour comes through.

Therefore, it's hard to see more than a couple of watts difference being made by the roads.

Marcos Apene do Amaral said...

Maybe this video series can help on some more analysis about that!
May I translate it into portuguese for a personal publication?
Cheers, Marcos

Marcos Apene do Amaral said...

The first part of the video!

Will said...

Very interesting article. And some very useful comments.

I would suggest that focusing future analysis on power / weight data might be more revealing than vertical speed.

For example, it would show that Armstrong was in fact exerting more power than the much lighter Contador. (which might make one less likely to hint that Contador is doing something suspicious).

Unknown said...

Average gradients for each km of the final 8.8km climb from Villette-le-Chable:
K1: 5.5%
K2 8.5%
K3: 7.5%
K4: 8.5%
K5: 8%
K6: 8%
K7: 8.5%
K8/finish (0.8km): 3.2%
Total climb is 638m.
Hope this helps.

Anonymous said...

Wrt. tailwind:

From what I saw, ther was tailwind at the 10km mark. It looked more like the usual afternoon uphill wind developing due to thermical effects.
The final climb had several hairpin curves, therefore you have rather some wind from the side, but noct too much, as the side of the road ar coverd by fans an the road is cut in the hill. All in all, I don't think there was a really relevant tailwind factor.

Anonymous said...

And data from an SRM-system used by a tour rider yesterday:

8,74 km distance
640 hm

djconnel said...

Wow -- thanks for the gradient info! So indeed the mapmyride numbers are short on distance and 8 meters short on height. Add around 1.1% to Alex's numbers as a first guess.

WRT "only 2 watts": Out of around 470 sustained for the climb 2 W is a VAM difference of around 8 meters/hour. So quite significant. Definitely worth considering when comparing different performances.

Marius-H said...

You can use Chris Anker Sorensens Garmin GPS data to get the exact figures.


The climb is 2090 ft and 7.3 grade.

I calculated an average wattage on Contador of about 450 watts and 1960 VAM.

Marius-H said...

Sorry 1855m/h VAM.

Ross Tucker and Jonathan Dugas said...

Hi everyone

Thanks so much for contributing to an awesome discussion on this thread - a lot of you have taken quite a bit of time to calculate, to read up, to find profiles and so forth, and it's really, really appreciated! Alex, DJ, Robert, Marcos, Marius, Jonathan - everyone, thanks so much!

I was shocked at the number of emails in my inbox when I woke up this morning!

I can't possibly do justice to all of them, and so instead of trying to respond here, i decided to do a second post, trying to analyse the climb in more detail. You can read that post as our latest article on the website:


I'm sure it will generate some good discussion as well!

ANyway, just so you know, I read all these comments, they were great - thanks again for the profiles, the calculations and the input!


Alex Simmons said...

I would still remain sceptical about:

- MMR data or other similar sites (MMR has a tendency to cut corners when measuring distances)

- GPX data from an on board cycle computer (settings chosen on these unit influence their accuracy)

- any cyclocomputer data relying on air pressure changes to determine altitude changes

- Tour data. think about it - what are they using to measure these things? We don't know, so if they are using similar flawed technology, who knows?

There have been plenty of cases of tour signage being placed where it is convenient, rather than where it is right.

When analysing info like this, it is far better to rely on power meter data from a known calibrated and zeroed power meter.

Movly said...

Well I for one don't belive your numbers. If I read all this comments right what you're saying is that those fat people with flags runing in front of riders are going at the speed of 22km/h uphill? I can hardly run 12km/h on a flat road and you want me to belive that those fat f**** are runing 20-24 km/h uphil, even though for just 50-100m.

Anonymous said...

1900??!! it may be too high or may be not. But if 1900 is too high to be true (keep in mind that Contador is better climber than Schleck) than Schleck was doped too as well as the others. Just simple calculation gives you Schleck's climbing number is 1836.

Supertramp said...

I used Ssörensens SRM-file together with analyticcycling.com to estimate Contadors power for the climb and got the following result:

time 25:32
weight: 64+8=72kg
Calculated power: 341W
Real power: 358W (+5,0%)

time: 20:36
weight: 62+8=70kg
Calculated power: 430W

Assuming the same correction (+5%) for Contador gives estimated power of 452W.