• The Cycling News forum is looking to add some volunteer moderators with Red Rick's recent retirement. If you're interested in helping keep our discussions on track, send a direct message to @SHaines here on the forum, or use the Contact Us form to message the Community Team.

    In the meanwhile, please use the Report option if you see a post that doesn't fit within the forum rules.

    Thanks!

Critical Power Study of GT Winners

Page 6 - Get up to date with the latest news, scores & standings from the Cycling News Community.
happychappy said:
My racing license reads "elite". What does yours say? The very point I was making was that average cyclists CAN compare themselves in terms of "muscular strength". Anyway, I wasn't comparing myself to them in terms of performance, I was giving a reason why larger muscles don't necessarily mean more power in endurance cycling (I'm not talking about sprinting, completely different).
An average cyclist (such as myself) can generate the power of Cancellara's outputs in a TT effort. The difference being the duration Mr. Average can maintain that power - or can't. That proves it isn't a "strength" issue.

I specifically said elite at discipline, because I didn't know your status, but your own admission pretty much indicated you didn't consider yourself an elite TTer. I won't repeat what I said before, which you don't seem to have understood, but maybe my next comment will make the point clearer.

Steroids in road cycling aren't used to gain muscle mass, they're used for faster recovery from aerobic endurance training so that they can go out and do more aerobic endurance training.
Gaining muscle mass in the legs isn't going to make you a faster Time Trialer.

Yes, it will. You are putting out more power in terms of watts. I think you are fixated on endurance, which is basically watts/kg. I’m not saying that isn’t crucial for a TTer, but it isn’t the entire story.

If it were the entire story, you wouldn't see these relatively small riders who are excellent climbers, but not elite TTers. They obviously have elite level endurance, yet they are relatively poor at TTng. Pantani wasn't a great TTer, but his drug-fueled endurance was one of the highest ever. As we have discussed upthread, he recorded some of the fastest climbs, and estimated watts/kg values, in history. Why was he not better at TTing? Because he lacked the muscular strength, which translates into more watts.

red_flanders said:
One, Moncoutie, is the outlier in the TT. The other is Froome, who somehow manages to match the times of Armstrong, Virenque, Contador and Schleck...clean.

And RR claimed or implied once that Moncoutie was not clean, though AFAIK he has refused to discuss the subject any further.

Ferminal said:
So P*t = AWC + CP*t but somehow P when t = 0 is not equal to AWC? That is exactly how they get the number in their table...

So by dump instantaneously, you mean use all of it. Yes, there is the equation P x t = AEC + CP x t, but as I said before, it breaks down at t close to zero. You can’t “dump” all of your AEC instantaneously. If you could, P would approach infinity as time approached zero.

Science is full of very useful equations that are highly accurate within certain limits, but which break down beyond those limits. Think of Newtonian physics at the quantum level, or all kinds of space-time singularities. It’s quite possible to define some parameter as power at time zero, even if power at time zero does not actually, experimentally equal that parameter. The problem does suggest this is not the ideal equation to use to model power output, but that does not mean that it can’t be very useful within certain limits. Quantum mechanics implies some phenomena that make no sense to us, but it also makes some of the most highly accurate predictions of any scientific theory.

If the results produce numbers that do not make any sense and may even be physiologically impossible then I will haggle over them...

What numbers don’t make any sense? What numbers are physiologically impossible? I mean the numbers he actually published as values in the paper.
 
Using Usain Bolts 100m time, it predicts a 1:36 marathon
(Using a Jack Daniels Vdot calculator)

Go Bolt, lets do THAT.


Using anaerobic power to estimate aerobic efforts is junk. So is the reverse.

The whole premise of the originally posted study is heavily flawed.

Now using say a 15 min effort to predict a threshold/60 min effort is much more sound.
 
Merckx index said:
If it were the entire story, you wouldn't see these relatively small riders who are excellent climbers, but not elite TTers. They obviously have elite level endurance, yet they are relatively poor at TTng. Pantani wasn't a great TTer, but his drug-fueled endurance was one of the highest ever. As we have discussed upthread, he recorded some of the fastest climbs, and estimated watts/kg values, in history. Why was he not better at TTing? Because he lacked the muscular strength, which translates into more watts.

But did not Pantani actually perform OK in the TTs when he was juiced to the gills?

Heras is also an example of someone who, when fully blood boosted, not only climbed insanely fast, but also did the TTs in the top 5.
 
deviant said:
I dont think it's as black and white as that, the argument that a GT winner is either a living legend, the best of all time etc etc....or a doper....doesnt take into account so many other variables.

First, that's a false dichotomy.
Second, the substantial marginal gains of low-dose Testosterone, recovery peptides, and documented use of blood transfusions while never testing positive. There is little hope for a clean-ish rider hitting a podium at a modern grand tour.
Third, the federation is still enabling the doping. Modern grand tours are a combination of entertainment wrestling and full-juice body building.
 
deviant said:
I dont think it's as black and white as that, the argument that a GT winner is either a living legend, the best of all time etc etc....or a doper....doesnt take into account so many other variables.

The discussion isn't about "a GT winner", it's about Froome. No it isn't as black and white when applied broadly, but it's pretty black and white when applied to Froome.

He was not any kind of GT contender for years. Then he's not just a contender, which would be startling, but he's dispatching the best (and likely doped) GT riders of the generation with ease.

That doesn't happen clean. Full stop. It's frankly remarkable even if you knew he was full on the gas for sure. Stunning transformation, the likes of which we've not seen, and we've seen some doozies.

The point of the "He'd have to be the best clean rider of all time" discussion is that for anyone to do what he's doing clean, that person would have to be the most gifted rider we've ever seen. 99.999% of pro riders are not that guy. Merckx is that guy. Froome wan't within shouting distance of being that guy, then all of a sudden he's crushing everyone. Makes no sense. Very black and white.
 
Jul 5, 2009
2,440
4
0
Visit site
Froome is ridiculous. Between the Tour of Romandie, mid-April 2011 and the Tour de Suisse, June 2011 his FTP permanently increased by almost 20%. This comes from analyzing his TT performances (relative to the rest of the peloton) from 2008 until now.

Choosing only non-prologue and non-mountain time trials, I found that Froome became 4.7 seconds per kilometer faster from mid-2011 until today. If you assume an average TT speed of 50 km/hr, the speed increase works out to ~20% more power.

It's a sudden and permanent increase with no possible explanation other than doping. And that must be some impressive dope, because that is an *enormous* increase.

John Swanson
 
Jul 21, 2012
9,860
3
0
Visit site
If the peloton is clean-ish, then doping becomes even better, so I dont see how the peloton could become clean by itself. There will always be someone that will take the extra risk and go for the big payday.

In Froome case it seems obvious he is someone like that, rather than a clean rider suddenly shining in the "clean-ish" era.
 
Mar 31, 2010
18,136
4
0
Visit site
Race Radio said:
It has been well reported and discussed that Horner's W/kg and VAM were higher then Froome/Quintana on multiple climbs in the Vuelta. His climbing times also set new records on Peña Cabarga and the Angrilu.....not by a little but by a lot

again with the bull**** horner didn't beat the angliru record for sure
 
Mar 31, 2010
18,136
4
0
Visit site
none of them are ridiculous. the 2003 could still happen with clean athletes or close to that level for a bit. what can't happen anymore is seeing this after 3 weeks anymore. this alpe dhuez one is early in the tour I know but from 2003 tour I can show videos of late stages where the same happens.
 
Ferminal said:
Yes



So P*t = AWC + CP*t but somehow P when t = 0 is not equal to AWC? That is exactly how they get the number in their table...



Where have I challenged that? I'm not that much of an idiot. That does not make critical power models any better than a rule of thumb (or even descriptive) approach to guessing the P-t relationship, especially when adding another layer compounds any errors in the initial data. People get worked up enough over individual power estimations yet somehow one at 20mins and one at 60mins and we have cracked a rider's physiology, to the point when we have even decomposed power into aerobic and anaerobic sources. If the results produce numbers that do not make any sense and may even be physiologically impossible then I will haggle over them...

Every model has its limitations.

Predicting short duration power outputs with the CP model is prone to large error and/or completely unrealistic predictions. You just don't do it.

Such predictions get better/worse depending on the duration of inputs used to calculate CP and AWC (W' / AEC), and also on the nature or the rider's physiology (e.g. proportion of various Type II fibres). The CP model is pretty sensitive to the power-durations used as input.

If these models did not use a maximal effort of a handful of minutes as one key input, then all bets are off.

Keep in mind that in such models, the slope of the power-duration line (CP) isn't nearly as sensitive as the intercept (AWC), and using data points a long way from the intercept makes this problem worse. Hence why it should include a short duration maximal power effort of a handful of minutes.

Small changes in calculated slope (CP) can cause large changes in AWC. Such large changes that they are physiologically nonsensical. If you plotted error bars on the input power numbers (and these are power estimates from ascent times), then you'll quickly see how AWC has a massively wide margin of error.


Then using power estimates from climb times, that are not controlled, and don't include appropriate durations for the model to be reasonably valid over durations from ~ 3-60 minutes (but definitely not 0-2 minutes), I'd be very cautious in drawing any conclusions from such analysis.
 
Ryo Hazuki said:
none of them are ridiculous. the 2003 could still happen with clean athletes or close to that level for a bit. what can't happen anymore is seeing this after 3 weeks anymore. this alpe dhuez one is early in the tour I know but from 2003 tour I can show videos of late stages where the same happens.

So...performances which are proven to have been done fully doped are not ridiculous in your opinion and can be matched clean...for a bit.

Good luck convincing someone else on that one. Sort of darkly curious...do you really believe what you're saying? It seems hard to imagine.
 
Dec 18, 2013
241
0
0
Visit site
ScienceIsCool said:
Froome is ridiculous. Between the Tour of Romandie, mid-April 2011 and the Tour de Suisse, June 2011 his FTP permanently increased by almost 20%. This comes from analyzing his TT performances (relative to the rest of the peloton) from 2008 until now.

Choosing only non-prologue and non-mountain time trials, I found that Froome became 4.7 seconds per kilometer faster from mid-2011 until today. If you assume an average TT speed of 50 km/hr, the speed increase works out to ~20% more power.

It's a sudden and permanent increase with no possible explanation other than doping. And that must be some impressive dope, because that is an *enormous* increase.

John Swanson

I'm guessing he may have been tested once or twice during his emergence as a GC contender, if his doping is that enormous wouldnt we be hearing his rivals crying foul?, seeing some failed drug tests?, anything?....instead its the armchair scientists who think they have all the answers, if his doping is that blatant wouldnt WADA love to get their hands on him?....i'll wait until/if he tests positive before resorting to the clinic's default position of claiming doping to explain every decent performance.
 
May 26, 2009
4,114
0
0
Visit site
deviant said:
I'm guessing he may have been tested once or twice during his emergence as a GC contender, if his doping is that enormous wouldnt we be hearing his rivals crying foul?, seeing some failed drug tests?, anything?....instead its the armchair scientists who think they have all the answers, if his doping is that blatant wouldnt WADA love to get their hands on him?....i'll wait until/if he tests positive before resorting to the clinic's default position of claiming doping to explain every decent performance.

May want to check out 'Omertà' and the control it has over the peloton.
 
I just KNEW someone would argue that Pantani was actually a great TTer. It doesn't change my point in the slightest. He was not nearly as good as, e.g., Ulle, and in any case, there are many other climbers with very high watts/kg values who can be used to make the same point. Muscle does matter.

Alex Simmons/RST said:
Every model has its limitations.

Predicting short duration power outputs with the CP model is prone to large error and/or completely unrealistic predictions. You just don't do it.

Yes, for the reason, among others, I discussed before.

If these models did not use a maximal effort of a handful of minutes as one key input, then all bets are off.

I don't know how you define handful, but fifteen minutes is the lowest time used.

Keep in mind that in such models, the slope of the power-duration line (CP) isn't nearly as sensitive as the intercept (AWC), and using data points a long way from the intercept makes this problem worse. Hence why it should include a short duration maximal power effort of a handful of minutes.

That's backwards. CP is the intercept, AEC is the slope (in the 1/t plots used to determine these parameters). And it is a long duration effort that matters, because these are the data points that are closest to the intercept (1/t approaches zero as t increases).

Small changes in calculated slope (CP) can cause large changes in AWC. Such large changes that they are physiologically nonsensical. If you plotted error bars on the input power numbers (and these are power estimates from ascent times), then you'll quickly see how AWC has a massively wide margin of error.

Again, it is AEC that is slope and CP that is intercept. I think both slope and intercept are somewhat sensitive to errors, but CP or intercept errors are somewhat limited because CP is not that much lower than power at one hour (again, it is long duration that matters), and the graphs all include a data point at or fairly close to that. E.g., Contador in 2009 had according to this study a very high AEC or slope, with a calculated CP of about 5.4. If the slope is lowered, the CP increases, but having a data point at about 6.0 puts a limit on how much it can be increased. A greater slope could potentially lower CP substantially, but the AEC here is already the highest in the study.

A good example of the problem, though, is provided by Horner's data. He had a very low determined AEC, and a high CP (which the author implies might suggest doping). But if you raise the slope or AEC just slightly, the CP can be lowered to the same range as that of the other riders, or maybe even lower.

Then using power estimates from climb times, that are not controlled, and don't include appropriate durations for the model to be reasonably valid over durations from ~ 3-60 minutes (but definitely not 0-2 minutes), I'd be very cautious in drawing any conclusions from such analysis.

Certainly agree that the power estimates can be debated, but this is a limitation of any analysis, given the absence of SRM data. And when data do exist, such as Froome's Vuelta ITT, it does suggest that the climbing power estimates are not his maximal efforts.
 
Merckx index said:
I don't know how you define handful, but fifteen minutes is the lowest time used.
Too long as the minimum duration input, but fine as an intermediate or upper duration input depending on how you are implementing the CP model.

For the purposes of estimating say 1-hour power (what you can sustain for a long time), then you'll want tests of like 3-6 minutes + ~20-25 minutes, and maybe another in between if you like. For traditional CP in literature, then upper duration tests are shorter, but they provide a CP value somewhat higher than 1-hour power or power that's sustainable for a long time.

If you already have an estimate of 1-hour power, then CP is redundant, and attempting to infer AWC from it is not sensible.

Using input durations that are too long (e.g. > 30-minutes) can also compromise the model (the assumption of linearity doesn't hold in reality).

Merckx index said:
That's backwards. CP is the intercept, AEC is the slope (in the 1/t plots used to determine these parameters). And it is a long duration effort that matters, because these are the data points that are closest to the intercept (1/t approaches zero as t increases)..
well yes it depends on how you flip the equation around, I was referring to the more conveniently presented equation in the linear form that Ferminal showed.
y = mx + b, where:
y = total energy output
m = critical power (slope of the work-duration relationship)
x = time
b = AWC (intercept)


Merckx index said:
I think both slope and intercept are somewhat sensitive to errors, but CP or intercept errors are somewhat limited because CP is not that much lower than power at one hour
CP is almost always higher than 1-hour power, sometimes significantly so. How much higher depends on the input durations chosen.

Merckx index said:
(again, it is long duration that matters), and the graphs all include a data point at or fairly close to that.
Using long duration inputs can invalidate the CP model linearity assumption.

Merckx index said:
Certainly agree that the power estimates can be debated, but this is a limitation of any analysis, given the absence of SRM data. And when data do exist, such as Froome's Vuelta ITT, it does suggest that the climbing power estimates are not his maximal efforts.
And if you don't have accurate data from true maximal efforts of appropriate durations, and collected from around similar periods of form, then then to be frank, it's just all GIGO.
 
Mar 12, 2009
2,521
0
0
Visit site
deviant said:
I'm guessing he may have been tested once or twice during his emergence as a GC contender, if his doping is that enormous wouldnt we be hearing his rivals crying foul?, seeing some failed drug tests?, anything?....instead its the armchair scientists who think they have all the answers, if his doping is that blatant wouldnt WADA love to get their hands on him?....i'll wait until/if he tests positive before resorting to the clinic's default position of claiming doping to explain every decent performance.

Lance -I never tested positive - says hi.
 

TRENDING THREADS