Power Data Estimates for the climbing stages

[Von Mises]

From Chris Horner´s SRM http://www.srm.de/index.php?option=com_content&view=article&id=563&catid=112&Itemid=260&lang=en it took 52.22 minutes to climb Tourmalet.

"Chris Horner climbed very well. It was very hard in the beginning with the lead outs by Saxo Bank - average 377 watts (5.9 w/kg) for the first 9.3 kilometers, then 347 watts for the second half of the climb with a very consistent power output and speed, just perfect pacing (5.4 w/kg). "

If thats correct, then Contador and Schleck climbed 50.37 minutes.

 

[The Science of Sport]

Hi again

I must apologize - in my previous post, I erred in the time reported for that stretch from 20km to go banner to the 18.6km start of the climb. I said 4:42, it was 3:42 - a typo.

It changes the result very slightly, but I see that you've already been debating that.

In the end, the SRM data posted above resolves the issue anyway, because the data are for 18.6km and an ascent of 1387m (7.45&#37]More analysis here for those interested[/URL]

 

[Cozy Beehive]

Ross : That's fine. Looks like we all had to correct ourselves. I had to go back and change the time taken to climb for Contador-Schleck, this information coming from clues from Horner's power output file. Thanks to Le Breton who did call this out as strange before from reading my post. The Eurosport footage tricked us all a bit today. Yellow Jersey goes to Eurosport?

 

[131313]

directed at no one in particular (though I think Dr. Tucker should really take note), I think it should be illustrative that using charts with single data points taken from stages years ago to show that the peloton is "clean" or "dirty" is fraught with a lot of error.

We have a couple of very methodical, rigorous individuals estimating the power output and power files from the actual race...and yet there's huge disagreement in even the time and length of the climb! That doesn't give me a lot of confidence in estimating VAM or power output from a stage that happened 20 years ago...

Looking at Horner's file, it appears to be 5.7-5.8 w/kg depending on his weight on the day. That would probably put Schleck around 6.1-6.2 and Contador a little under that.

That's certainly a lot of power this late into the race, but then again these are the best two guys in the world, so I wouldn't say it's impossible to do clean. That would probably put their rested power around 6.4 w/kg, doing some more handwaving.

 

[Mr.38%]

Looking at Horner's file, it appears to be 5.7-5.8 w/kg depending on his weight on the day.

It's 5.6 W/kg actually as the other values were given for 64 kg.

 

[Le breton]

directed at no one in particular (though I think Dr. Tucker should really take note), I think it should be illustrative that using charts with single data points taken from stages years ago to show that the peloton is "clean" or "dirty" is fraught with a lot of error.

We have a couple of very methodical, rigorous individuals estimating the power output and power files from the actual race...and yet there's huge disagreement in even the time and length of the climb! That doesn't give me a lot of confidence in estimating VAM or power output from a stage that happened 20 years ago...

Looking at Horner's file, it appears to be 5.7-5.8 w/kg depending on his weight on the day. That would probably put Schleck around 6.1-6.2 and Contador a little under that.

That's certainly a lot of power this late into the race, but then again these are the best two guys in the world, so I wouldn't say it's impossible to do clean. That would probably put their rested power around 6.4 w/kg, doing some more handwaving.

There is so much info in those SRM charts that it should be possible to check for consistency and maybe draw some additional conclusions. If I find the time I'll look at it closely.

 

[acoggan]

We have a couple of very methodical, rigorous individuals estimating the power output and power files from the actual race...and yet there's huge disagreement in even the time and length of the climb! That doesn't give me a lot of confidence in estimating VAM or power output from a stage that happened 20 years ago...

Me either.

That's certainly a lot of power this late into the race

So let's think about this...

Chris Anker Sorensen threw down his best 10 min power of the entire race at the base of Tourmalet.

You produced your highest 5 min power on the last day of the Tour of California.

At least some riders have had stellar performances (e.g., Bradley McGee) soon after completing a Grand Tour.

The Aussies have demonstrated a positive correlation between changes in training load (measured using TSS ) and changes in hemoglobin mass in elite (female) cyclists (see http://www.ncbi.nlm.nih.gov/pubmed/20058020).

Could there be a connection? Even if there isn't, where is the evidence that a cyclist's power output must decline during a 3 wk stage race?

 

[Le breton]

...........
We have a couple of very methodical, rigorous individuals estimating the power output and power files from the actual race...and yet there's huge disagreement in even the time and length of the climb! That doesn't give me a lot of confidence in estimating VAM or power output from a stage that happened 20 years ago...
.

You have different cases, some well documented, other not so well documented.

Some climbs never change, like Alpe d'Huez, Ventoux from Bédoin ( where there actually is a reference line of stones on the road)... etc and other do change, as the Morzine-Avoriaz climb - as recently documented by Frédéric Portoleau in cyclismag.com.

In well documented cases you can confidently go back 20 years.

 

[Dekker_Tifosi]

According to Cyclismag they did Ax-3-domaines at the insane wattage of 438
Vino did the first few kilometers at 460W

That being said, it's a short climb, being only 8km, so it won't say everything.

 

[Le breton]

Chris Anker Sorensen threw down his best 10 min power of the entire race at the base of Tourmalet.

Is it surprising? He knew it was his most important contribution of the whole 3 weeks and he looked like he was about to die. I suffered just watching him.

 

[acoggan]

Is it surprising? He knew it was his most important contribution of the whole 3 weeks

Only if you believe that a rider's ability to produce power aerobically must decline over the course of the 3 wk.

 

[acoggan]

We have a couple of very methodical, rigorous individuals estimating the power output and power files from the actual race...and yet there's huge disagreement

Just to drive that point home: somebody posted this link in a comment in reply to Ross and Jonathon's blog:

http://www.members.aon.at/o.n/bergfahrer.html

Note:

1) the huge discrepancy between many of these numbers compared to others that are bandied about, and

2) the fact that they seemingly got Indurain's sustainable power very close to what lab and wind tunnel studies around the time of his hour record indicate that he could maintain (i.e., at least Indurain's numbers in this table seem accurate).

 

[Cozy Beehive]

It's 5.6 W/kg actually as the other values were given for 64 kg.

5.65 W/kg average.

 

[Mr.38%]

5.65 W/kg average.

Yo, superbrain:

Thread title [1]: "Power Data Estimates..."
SRM page [2]: "377 watts (5.9 w/kg)" -> 63,9 kg
SRM page [2]: "347 watts ... (5.4 w/kg)" -> 64,26 kg
Pubmed [3]: "mean error scores for SRM ... were 2.3 +/- 4.9%"
Pubmed [3]: "power output readings were noticeably influenced by temperature (5.2% for SRM)"
Pubmed [3]: "During field trials, SRM average and max power were 4.8% and 7.3% lower, respectively, compared with PT"

Pretending pseudo-accuracy sucks.



[1] http://forum.cyclingnews.com/showthread.php?t=8839
[2] http://www.srm.de/index.php?option=com_content&view=category&layout=blog&id=112&Itemid=260&lang=en
[3] http://www.ncbi.nlm.nih.gov/pubmed/15235334

 

[acoggan]

Pubmed [3]: "mean error scores for SRM ... were 2.3 +/- 4.9%"
Pubmed [3]: "power output readings were noticeably influenced by temperature (5.2% for SRM)"
Pubmed [3]: "During field trials, SRM average and max power were 4.8% and 7.3% lower, respectively, compared with PT"

Pretending pseudo-accuracy sucks.

[3] http://www.ncbi.nlm.nih.gov/pubmed/15235334

Note that such errors are the result of not periodically checking and adjusting the zero offset whenever the temperature changes markedly, as recommended by the manufacturer.

Then again, they may actually be representative of the accuracy of the data obtained from professional cyclists "in the heat of the battle", especially when large swings in temperature are most likely going to result from climbing or descending a mountain (with neither situation being particularly conducive to coasting long enough to check the zero offset, much less unclipping from the pedals if one believes that is necessary).

(Of course, now that the SRM records the temperature periodically during a ride, rather than just writing it into the data header once when a new file is first created, you might be able to correct the data in a post-hoc fashion to obtain accurate results. Doing so, however, would require knowing the exact relationship between the zero offset and the temperature for that particular power meter, something that varies considerably from one unit to the next, and something that only the most compulsive of individuals <g> would bother to measure.)

 

[Cozy Beehive]

Yo, superbrain:

Thanks.

The number is from the SRM file for Horner.

 

[smaryka]

Is it surprising? He knew it was his most important contribution of the whole 3 weeks and he looked like he was about to die.

Only if you believe that a rider's ability to produce power aerobically must decline over the course of the 3 wk.

That doesn't really follow, does it, unless you know what his rested peak 10 min power was? How do you know he wouldn't have been putting down a bigger number had that "most important contribution" happened in the first week?

 

[Mr.38%]

Note that such errors are the result of not periodically checking and adjusting the zero offset whenever the temperature changes markedly, as recommended by the manufacturer.

Then again, they may actually be representative of the accuracy of the data obtained from professional cyclists "in the heat of the battle", especially when large swings in temperature are most likely going to result from climbing or descending a mountain (with neither situation being particularly conducive to coasting long enough to check the zero offset, much less unclipping from the pedals if one believes that is necessary).

(Of course, now that the SRM records the temperature periodically during a ride, rather than just writing it into the data header once when a new file is first created, you might be able to correct the data in a post-hoc fashion to obtain accurate results. Doing so, however, would require knowing the exact relationship between the zero offset and the temperature for that particular power meter, something that varies considerably from one unit to the next, and something that only the most compulsive of individuals <g> would bother to measure.)

That's why I hate to compare my data or anyones data with one another. I have a pretty good "feel" for my different SRM cranks and *think* I got them calibrated pretty good in sync.

"Hey, I gained 3 W @ FTP this week and lost 0.21 kg"

Yeah, riiiiight! Before or after the morning toilet?

 

[Le breton]

...........

So clearly, that 18.6km to go banner was not correct. I would estimate that it was about 1 km late. Which is why I agree more with Ron's calculation that the 18.4km took about 53 minutes and not the 49 min that I timed from that banner.


[/B]
.

I JUST WATCHED THAT SECTION OF THE RACE here
http://tour-de-france.sport.francetv.fr/videos?date=20100722&type=revoir-etape

I had not seen that 18.6km banner on the RHS of the road yesterday. You are right when the leading group reaches it the GP indicates -17.8km!!!

 

[acoggan]

That doesn't really follow, does it, unless you know what his rested peak 10 min power was? How do you know he wouldn't have been putting down a bigger number had that "most important contribution" happened in the first week?

His maximal power for 10 min when fully rested is relevant only if you believe that he hasn't been equally motivated during previous stages. Me, I think that "cherry-picking" values from throughout the 3 wk of racing should provide you with a good feel for any downward trends in an individual's performance, which at least in his case don't appear to exist.

(BTW, has anybody attempted to estimate Contador's power from the l'Alpe de Huez stage of the Tour de Suisse? I have a very good idea of what it was, so it's interesting to compare that value - produced at the end of a much shorter stage race - to the estimates in this thread for when he and Schlek went mano y mano after 3 wk of racing.)

 

[Cozy Beehive]

BTW, has anybody attempted to estimate Contador's power from the l'Alpe de Huez stage of the Tour de Suisse? I have a very good idea of what it was, so it's interesting to compare that value - produced at the end of a much shorter stage race - to the estimates in this thread for when he and Schlek went mano y mano after 3 wk of racing.)

On stage 13 of the 2008 Vuelta a Espana, Contador had around 6.3 W/kg +/- 0.1 W/kg on the massive Angliru. I watched it with great interest.

I doubt VAM or W/kg is a complete indicator of performance and fatigue as a function of racing duration. What would be more interesting to see are heart rates (something hardly anyone talks about on the forum as a constant power output does not mean exercise intensity is constant), power outputs, CE, GE and a knowledge of the level where a pro-cyclists' stress level hormones are at after each stage, and on recovery days. This will pin down the entire map of what we term "performance" and "fatigue". Such data from a rider like Contador is a fairytale wish. Until then, we can keep speculating.

 

[acoggan]

On stage 13 of the 2008 Vuelta a Espana, Contador had around 6.3 W/kg +/- 0.1 W/kg on the massive Angliru.

Interesting.

a constant power output does not mean exercise intensity is constant

Not to nitpick, but technically, it does. That is, in the field of exercise physiology exercise intensity has long been defined on the basis of either the velocity/external power or the resultant metabolic rate. I take your point that efficiency could vary throughout a 3 wk Grand Tour, but if anything you would expect it to go up, not down.

 

[131313]

(BTW, has anybody attempted to estimate Contador's power from the l'Alpe de Huez stage of the Tour de Suisse?

I think you mean the Daupine? They tend to do the Tour de Suisse in, well, Switzerland.

I have a very good idea of what it was, so it's interesting to compare that value - produced at the end of a much shorter stage race - to the estimates in this thread for when he and Schlek went mano y mano after 3 wk of racing.)

You mean based on the power file of Mr. "40W in 2 weeks"? I'm supposed to believe that dude is re-doing his zero offset?

Kidding aside, do you have any insight into the rigor those guys apply to such things? The "professionals" I know seem pretty amateur when it comes to things such as data collection. Several seemed to not even understand the difference between "including zeroes" and "non including zeroes". One guy said "I don't include them because it's always higher that way". True story.

Anyway, I have no data from the climb but if I can dig it up I'll proffer a calculation.

To your greater point about power needing to drop over a 3 week stage race, I'll say that like you on Lance's alleged doping, I'm agnostic. As you've pointed out, I've seen personal best numbers during periods of high fatigue. However, a drop in HCT has been documented (Basso's blood values, anecdotal comments from team physicians, comments from LeMond ((I honestly don't know if this is supposition on his part or actual test results)), the published study on the MTB stage race guys). Of course, your link seems to demonstrate that it's not absolutely necessary.

Likewise, I guess it's possible that an increase in efficiency could overcome a drop in HCT. And lastly, motivation may simply be a key component, i.e. actual ability may be underestimated in the beginning. Short answer: I don't know. I still think it's a reasonable question to ask, though.

 

[Cozy Beehive]

Interesting. Not to nitpick, but technically, it does.

When I used to be into heart rate monitors, I discovered the simple principle of the slow component of VO2. On several flat 70-100 mile rides where my goal for the day was to maintain constant speed (or effort) just to enjoy the day, I was surprised to find that heart rates kept climbing up steadily until towards the end of the rides, my lungs were heavy and I was beyond 70% of my max HR. I wondered if I had some kind of heart problem. Why could I not get my HR to be constant in spite of trying to bike at submaximal level. There is a definite relationship between exercise intensity indicators of VO2 and HR. 55% VO2 max corresponds to about 70% max heart rate. This is why I wrote that maintaining a constant power output or speed does not necessarily mean your HR is steady. It will climb as your body needs more oxygen to process for sustaining the same workload, in other words...body becomes less efficient.

I take your point that efficiency could vary throughout a 3 wk Grand Tour, but if anything you would expect it to go up, not down.

How do you know efficiency goes up throughout a 3 WK grand Tour?

 

[Parrot23]

I was surprised to find that heart rates kept climbing up steadily until towards the end of the rides, my lungs were heavy and I was beyond 70% of my max HR. I wondered if I had some kind of heart problem. Why could I not get my HR to be constant in spite of trying to bike at submaximal level.

That's standard "cardiac drift": stroke volume, I think, declines with dehydration/heat. Heart compensates for lower stroke volume by beating faster. Doesn't mean you are working harder. One reason why Lemond wished he had an SRM in his day.