# Power Data Estimates for the climbing stages

Page 80 - Get up to date with the latest news, scores & standings from the Cycling News Community.

#### Krebs cycle

Netserk said:
I'm calling BS on that statement. When calculating Froome's W/kg on Ventoux the difference between Froome weighting 65 and 70 kg is 1.38%.

So since those who have calculated it have used 67kg, they have less than 1% uncertainty in W/kg because of the weight.

Now would you be so kind and explain where you pulled the numbers 5-10% from?
how the heck did you arrive at 1.38%?

At ANY power output the difference in w/kg from 65 to 70kg is 7.1% or the other way around (ie: from 70 to 65kg) is 7.7%.

It's about 1.5% per kg. Remember that TWO riders are being compared to each other, so if you were only 2kgs out for both riders that would be a 6% error on its own, and you haven't even begun to include all the uncertainties due to wind, drafting etc etc.

Again, I'm not saying that estimating power is a totally worthless exercise. It's good for a ballpark figure, that is all. I'm not saying that Froome is clean and I get it that many people want it to be proven, but ffs don't pretend this quackery is the way to do it.

#### Krebs cycle

I can't speak for the others.

I don't think I'm debunking W/kg calculations, just encouraging people to think critically about how such information is arrived at and also what (IMO) we can and can't interpret from the information.
I agree Alex. It's not about "debunking" anything. It's about accepting the reality that if you do not have access to accurate information inputs in your model, then the error in your output will be too high to make any sort of firm conclusion.

Anyone who really wants to know the details of power output modelling in cycling should read Jim Martin's paper from 1998.

http://www.recumbents.com/wisil/MartinDocs/Validation of a mathematical model for road cycling.pdf

#### Ferminal

The absolute difference in the weights doesn't mean anything unless you want to calculate W, it's the ratio between the rider weight and equipment weight which is important to W/kg. Without getting the calculator out it is something ~1% per 5kg.

Edit: Proably confusion as this is from an estimation point of view not taking someones SRM W reading and converting it to W/kg.

#### Fearless Greg Lemond

Alex Simmons/RST said:
I can't speak for the others.

I don't think I'm debunking W/kg calculations, just encouraging people to think critically about how such information is arrived at and also what (IMO) we can and can't interpret from the information.
Okay, perhaps I misinterpreted your post. With this one I agree totally.

#### Cobblestones

Ferminal said:
The absolute difference in the weights doesn't mean anything unless you want to calculate W, it's the ratio between the rider weight and equipment weight which is important to W/kg. Without getting the calculator out it is something ~1% per 5kg.

Edit: Proably confusion as this is from an estimation point of view not taking someones SRM W reading and converting it to W/kg.

Totally agree. Some people never seem to get it. When you quote power in W, they moan about the need to take the weight of the rider into account, when you quote specific power in W/kg then they still complain about the same thing. There should be a prereq for posting in these threads such as a Phys-101 class or something.

#### Alex Simmons/RST

Fearless Greg Lemond said:
Okay, perhaps I misinterpreted your post. With this one I agree totally.

No probs. If you misinterpreted it, then perhaps others may too and I need to think a bit more about how I write, so as to avoid such misunderstanding.

#### Alex Simmons/RST

Krebs cycle said:
how the heck did you arrive at 1.38%?

At ANY power output the difference in w/kg from 65 to 70kg is 7.1% or the other way around (ie: from 70 to 65kg) is 7.7%.

It's about 1.5% per kg. Remember that TWO riders are being compared to each other, so if you were only 2kgs out for both riders that would be a 6% error on its own, and you haven't even begun to include all the uncertainties due to wind, drafting etc etc.

Again, I'm not saying that estimating power is a totally worthless exercise. It's good for a ballpark figure, that is all. I'm not saying that Froome is clean and I get it that many people want it to be proven, but ffs don't pretend this quackery is the way to do it.

I think there is a bit of cross purposes going on. It comes down to the issue of distinguishing between the physiological W/kg and the physical W/kg.

Two riders of different body mass but with same W/kg (body mass) will have very similar, but not exactly equal climbing speed on steeper climbs, due to the fact that while we measure human performance in terms of W/kg (body mass), we calculate speed based on total mass of the system.

Typically the bike/kit of a lighter rider will be a larger proportion of the total mass of bike/rider/kit, thereby placing them at a small disadvantage. There are other physical factors that may close that gap a bit (e.g. higher rolling resistance, greater aero drag for the larger rider) but the impact of the mass is usually more important on steep climbs.

e.g.

Take two riders with 5.8W/kg (body mass) to throw down:

70kg rider with 8.0kg bike+kit
60kg rider with 7.5kg bike+kit

The 70kg rider produces 406W and that goes to raising 78.0kg (5.21 W/kg)
The 60kg rider produces 348W and that goes to raising 67.5kg (5.16 W/kg)

That would result in an approximately 1% difference in speed on a steep climb (a bit less than that, but close enough for the purpose of the example).

So while the riders have same physiological ability expressed in W/kg terms, the total system mass they are lifting is not in proportion to their body mass and that results in a slightly different speed. But the difference is of the order of 1%.

Conversely, when converting climbing speed to W/kg, there does need to be an assessment of the rider mass and as a proportion of total system mass. The error then is in assessing the proportion of body mass/total system mass. But it will be of similar order of magnitude expressed above.

The differences do become greater however when both riders find themselves on bikes that are at the UCI minimum weight limit. In that case the lighter rider is penalised more.

#### Krebs cycle

Ferminal said:
The absolute difference in the weights doesn't mean anything unless you want to calculate W, it's the ratio between the rider weight and equipment weight which is important to W/kg. Without getting the calculator out it is something ~1% per 5kg.

Edit: Proably confusion as this is from an estimation point of view not taking someones SRM W reading and converting it to W/kg.
Eqn 15 in Jim Martin's paper gives power in watts. Eqns 8 and 9 require rider + bike weight. So first of all you need to know the riders bodyweight and bike weight accurately to determine absolute power to begin with. Then of course, you simply divide by the bodyweight to determine w/kg. If you don't know the bodyweight exactly then you're actually going to introduce that source of error on several occasions when expressing the output in w/kg.

As far as I can tell from Ferrari's post he made back in 2009 at 53x12, he has not created a detailed physical model as appears in the Martin paper. I could be wrong but what he seems to have done is measured power output using SRMs for many cyclists on various climbs and then fitted a linear regression eqn. It's not a bad method, but it is nowhere near as accurate because he could have been recording power data all over the shop that was being influenced by varying wind conditions, aero position, equipment etc. The coefficients included in the pVAM method are fudge factors created by Ferrari to "normalize" different climbs so that you can compare them. The altitude fudge factor isn't even a constant in reality because the effect of altitude differs for different individuals. Ferrari's method is useful for field based estimations of relative performances, but still nowhere even close to the accuracy of the model in the Martin paper for determining power output.

But hey, if people want to believe they can measure the diameter of a grain of salt to 3 decimal places with a 30cm wooden ruler then good for them.

#### Ferminal

So you no longer believe that numbers show people are climbing slower than "a few years ago"?

#### Merckx index

Alex Simmons/RST said:
As for standard stages/tests, well I just can't see such events incorporating such tests. Only place to control for the variables is a lab. That's just not going to happen. This is a race.

Again, I'll say that if cycling is serious about doing everything possible to identify doping, it will bring the lab to the race. It's time to jettison tradition and think outside the box.

There is a big difference between a test that is used to declare whether a substance actually is detected in blood/urine, and a power meter file which simple records what the rider did performance wise. The power meter file does not tell us whether a rider used a prohibited substance or method, just as climbing ADH in 40-minutes doesn't either.

As I have argued previously in this forum, using power data is a logical extension of the biopassport approach. The BP does not directly indicate whether a prohibited substance is in the body, either; it’s an indirect measure, the idea being that a certain level of physiological parameters could only be caused (i.e., with high probability only be caused) by such a substance. Use of power data simply extends the indirect approach from physiological to functional data. The premise is that a certain level of power could only be caused by some prohibited substance.

Since there is substantial normal variation in physiological parameters (and further variation caused by abnormal but not necessarily doping factors), the biopassport is always going to be less certain, fuzzier if you will, than a direct test for a substance. Still, as we have seen recently, with the help of sophisticated algorithms, it’s possible to set criteria that can be confidently used to determine whether doping occurred. Power data are even fuzzier, subject to still more variation, but I see no reason why criteria can’t also be set. Again, it just means that a lot of false negatives will get away. That is undoubtedly the case with the BP, too--probably any rider who wants to can dope a little and beat it--but it most likely reduces the degree to which a rider can dope. Riders can no longer jack up their HT with no end in sight (or even up to 50%).

The same should be true with power data. It won't identify all doping, it should be able to eliminate excessive doping.

I'm confused with your terminology. You are saying we can't sanction based on power data but it can be sufficiently indicative of being positive for doping. If it's sufficiently indicative, then surely sanction must follow? Else it's pointless.

In what way does the data tell us who should be targeted that we don't already know by the fact that they are winning/elite/professional bike riders?

OK, so what's the line? And why?
Is the line the same for say Mark Cavendish as it is for Alberto Contador or <insert any rider name here>?
If not, why not? How would you determine this?

Perhaps, but if we are going to implement a doping control regime, should it not apply equally to all?

The precursor of the BP, the off-score, was never used to sanction, either, and the 50% HT rule was used only for very short suspensions. They were never considered reliable enough for full suspensions. Even after the BP was developed, there was a long period in which the method was not considered good enough to use to sanction. Many people questioned whether it would ever be robust enough to do so. Eventually, thought, it was developed to the point where sanctioning could be done confidently.

I see the same sort of evolution with power data. These data are not going to be used to sanction riders in the beginning. As I said, they will just provide officials with a better idea of what's going on. You have emphasized all the variables that make it difficult to determine power from climbing times. Just having SRM data will clarify that a little. Actual power outputs will be known more precisely.

Now it may be that all or most of the top riders are doping, so these data will not indicate what a clean performance could maximally be. But at least they will suggest the maximum performance under current doping regimes, so if a new substance or doping method comes along--and this of course is what the history of doping has been all about--it might be possible to spot it. I'm not going to get into details here. I will just point out that the same kind of approach used in the BP--sustained deviations from baselines--can be used here. In this initial phase, the baseline will be constructed from the best riders/times, and deviations compared against that. Yes, some once in a generation rider can come along and outperform everyone will staying clean or at least not doping any more than they do--or maybe respond better to the dope--but again, there has to be some limit to this, and as data are gathered over time and riders, this limit should become clearer. And again and again and again I emphasize that the limit can be set for any arbitrary significance level, if we are willing to let many potential false negatives pass through.

This is the first phase. In the second phase, as I see it, as the program is gradually expanded, baselines will be developed for individual riders, just as they are with the BP. So in addition to a line that no one should be able to cross, individuals will have their own lines. I realize that constructing such lines will be more difficult than for the BP, but I don't see that it's impossible. Again, an enormous amount of data will be gathered by this time, we will have a very good idea of how much a rider can improve with training and time. And again, there no doubt will be a long period, as with the BP, when a lot of suspicious values will be seen, but no sanctioning, until the technology is considered robust enough to do this with confidence.

You may argue that this will never work in practice, and possibly you could be right. But we will never know for sure if we don't try it, and see how much the data allow us to refine the model and set criteria. The first phase, at least, is not terribly expensive.

And what if a rider has a "mechanical" and changes bikes? Are we going to stop the race so an independent referee can validate the new power meter's zero offset and slope setting are correct (assuming you have a power meter on the spare)? Do we ban using a team mate's bike if that became necessary?

If the rider is with others who have power meters, some kind of correlation is possible. I'm not saying problems like these won't crop up, but I don't think they're fatal to the program.

Then there is the context argument that you allude to earlier (i.e. being able to account for things like length of stage etc when deciding on what the line in the sand is). Well if you don't use the data to account for the rider's workload through the whole race from Day 1, indeed for the months leading up to the race, then doesn't that sort of negate the ability to account for such things?

This would only be an issue if we wanted to draw a very sharp line between doping and non-doping. I have already emphasized, though, that any line will undoubtedly allow for a lot of false negatives (much as the BP does). It's purpose is not to catch every doper or even most dopers, but to prevent excessive amounts of enhancement.

It's pretty clear that the main value of the BP is not that it catches most dopers or even, really, that it catches any dopers. It's main value is that it creates an environment in which riders feel they must dope less. I see power data having much the same effect. It might be that at the end of the day we can never confidently set a power value line that we can confidently use as the basis of a sanction. But merely by trying, we might well force riders to reduce their doping in fear of crossing such a line.

#### Krebs cycle

Merckx index said:
It's pretty clear that the main value of the BP is not that it catches most dopers or even, really, that it catches any dopers. It's main value is that it creates an environment in which riders feel they must dope less. I see power data having much the same effect. It might be that at the end of the day we can never confidently set a power value line that we can confidently use as the basis of a sanction. But merely by trying, we might well force riders to reduce their doping in fear of crossing such a line.
What happens if they do cross the line though? Do they get a warning? a ban? Can you see the logical outcome of this reasoning? Since power output is the dominant factor that determines velocity, then think about what that would mean for say track cycling events..... no more world records after a certain point because it is considered to cross some line. Why would anyone bother trying to break the world record then? Its like saying we are going to put an artificial cap on maximal human performance.....

Begone with you citius altius fortius.

#### Le breton

Alex Simmons/RST said:
I think ..................
........... In that case the lighter rider is penalised more.

You would make a good elementary school teacher

#### zastomito

Merckx index said:

Great post.
Maybe making power meters mandatory for 3 randomly chosen teams and rotate them yearly. All expenses regarding power meters covered by UCI. Maybe SRM would be willing to give them for free if some kind of a sponsorship deal would ensue. You could get data for various types of riders on various types of terrain, maybe not enough to form somebody's power passport but certainly enough to establish narrow area in which certain type of rider in certain type of race should be.

#### BigBoat

Krebs cycle said:
What happens if they do cross the line though? Do they get a warning? a ban? Can you see the logical outcome of this reasoning? Since power output is the dominant factor that determines velocity, then think about what that would mean for say track cycling events..... no more world records after a certain point because it is considered to cross some line. Why would anyone bother trying to break the world record then? Its like saying we are going to put an artificial cap on maximal human performance.....

Begone with you citius altius fortius.

You seem to exclude the possibility that it may be impossible to break a world record without improving aerodynamics from the past record either through pressure change, bike changes or changes in power output (through doping).

#### Alex Simmons/RST

Bring the race to the lab. never. will. happen.

Nor do I think that is necessary to demonstrate a serious commitment to anti-doping.

I have run indoor turbo racing though and it can be a blast. Handicap start time trials.

So given that it's likely that a power data method will be even "fuzzier" than the BP and never be good enough to end with a sanction, then why not put those resources into improving the BP, direct detection and/or methods that can end with a sanction?

If you have X\$ to spend on anti-doping, then are power meters a wise investment? Will they provide a better anti-doping return than using those resources in other ways.

It's not like this idea is going to suddenly add a crap load of new money into anti-doping.

And that's the crux of it - if we want to demonstrate seriousness about anti-doping, then the resources devoted to it need to increase substantially, by an order of magnitude at least.

I would suggest there are better ways to use those resources than on SRMs as dopeometers.

#### Alex Simmons/RST

zastomito said:
Great post.
Maybe making power meters mandatory for 3 randomly chosen teams and rotate them yearly. All expenses regarding power meters covered by UCI. Maybe SRM would be willing to give them for free if some kind of a sponsorship deal would ensue. You could get data for various types of riders on various types of terrain, maybe not enough to form somebody's power passport but certainly enough to establish narrow area in which certain type of rider in certain type of race should be.

It still comes down to money and choices about how to spend what little there is.

In 2012 the CADF spent a little over CHF7 million (Swiss francs) on its activities, of which the UCI contributed CHF1.1 million (16%). The largest source of income are the pro teams (pro tour and pro conti) with CHF4.72 million (67%).

Keep in mind that the UCI is not exactly a wealthy organisation with millions of bucks to toss about.

In 2012 the CADF managed to get into the black by a mere CHF200k, after several years of being deep in the red, and its annual "profit" was approx CHF150k.

OOC testing by the way accounts for CHF3.65 million (53%) of total expenditure and in-competition testing another CHF0.95 million (14%). 92% of that is spent on professional men's road racing.

So tell me where the money is coming from to suddenly have enough for power meters and technicians and specialist analyst staff and systems, if it's not going to be diverted away from the BP, OOC testing and direct substance detection?

As Jerry McGuire had to shout...
"Show me the money!"

#### 131313

Alex Simmons/RST said:
If you have X\$ to spend on anti-doping, then are power meters a wise investment? Will they provide a better anti-doping return than using those resources in other ways...

I would suggest there are better ways to use those resources than on SRMs as dopeometers.

Well, that is the question. As far as the \$\$, I think you're overstating the outlay. Many/most riders are now using PM's. A requirement for the this wouldn't be much different than the UCI's bike regs anyway, and least at the Pro Tour level. Having the teams self-report and then publicly hosting the data,say on the race organizer's web site shouldn't be that much of a chore, nor should establishing a data base for the doping management body. There are even some software programs available that make it pretty easy to review the data. For grand tours and classics, 5 SRM's could be issued to riders, and those could be managed by the governing body (calibrations/downloads). This could be done in part randomly, and partly on past results or current standings, e.g. the race leader in a grand tour is always using a monitored SRM.

As MI correctly pointed out, it's not that it will necessarily help detect doping (though as I said before, big changes in a rider could signal a big red flag), it could also serve as a deterrent. A rider will realize that a dramatic increase in power will subject him to additional scrutiny.

As to a point you made earlier, something to the effect of "if they're WT riders they're already subject to scrutiny", I'm not sure you're aware of just how few controls many of the riders actually get, particularly OOC. The three world tour riders I know best have had 4 OOC tests between them this year. So, more testing is obviously necessary. But targeting testing really seems to yield the best results, and power data could help do that--more than just race results, since racing obviously involves a lot more variables than power production.

#### Alex Simmons/RST

131313 said:
Well, that is the question. As far as the \$\$, I think you're overstating the outlay. Many/most riders are now using PM's. A requirement for the this wouldn't be much different than the UCI's bike regs anyway, and least at the Pro Tour level. Having the teams self-report and then publicly hosting the data,say on the race organizer's web site shouldn't be that much of a chore, nor should establishing a data base for the doping management body. There are even some software programs available that make it pretty easy to review the data. For grand tours and classics, 5 SRM's could be issued to riders, and those could be managed by the governing body (calibrations/downloads). This could be done in part randomly, and partly on past results or current standings, e.g. the race leader in a grand tour is always using a monitored SRM.

As MI correctly pointed out, it's not that it will necessarily help detect doping (though as I said before, big changes in a rider could signal a big red flag), it could also serve as a deterrent. A rider will realize that a dramatic increase in power will subject him to additional scrutiny.

As to a point you made earlier, something to the effect of "if they're WT riders they're already subject to scrutiny", I'm not sure you're aware of just how few controls many of the riders actually get, particularly OOC. The three world tour riders I know best have had 4 OOC tests between them this year. So, more testing is obviously necessary. But targeting testing really seems to yield the best results, and power data could help do that--more than just race results, since racing obviously involves a lot more variables than power production.
Yeah, good thoughts, just tossing stuff about...

But do we really need power meter data from races to know who to target and whose performances have improved significantly? Surely examining race results already does that and far far cheaper.

Do teams self report their internal biological testing? Can we legitimately trust teams to supply the data given they will see it first?

In 2012 there were 6,610 OOC tests performed and >90% were on men's professional road riders (world tour and pro conti) so given there are something over 1,000 riders, then yes, I'm pretty well aware of the low ratio of such tests per rider*. Mind you, it's higher than most sports.

Why do you think I am saying that the whole shebang is an order of magnitude too small? That's the real problem here. SRM data is a red herring.

Then there is the irony of those voices calling for power meters to be banned in racing

* the situation at a national anti-doping body level is even worse. e.g. take ASADA, they perform a similar number of tests per year which covers the entire professional and elite sports in the country - all 4 codes of football, all major Olympic sports, and all other major sporting codes. The test per athlete ratio is woefully low, and Australia has one of the best resourced anti-doping bodies in the world. USADA's budget is barely double ASADA's despite an economy 11 times larger.

#### zastomito

Then let's just make them mandatory on 3 random teams who would pay for it themselves and rotate on yearly bases. The point is not only target testing but to collect large data base that would allow better understanding of maximum power output in any number of conditions. So if it was started 10 years ago you would already have enough knowledge to presume with high probability where is that magic line, regardless of how 'fat' that line actually is.
Because I don't believe that there is no line. The further we go the more this story becomes Zeno's paradox, infinitely small gains by which we will never reach the line that actually is just around the corner.

#### Alex Simmons/RST

zastomito said:
Then let's just make them mandatory on 3 random teams who would pay for it themselves and rotate on yearly bases. The point is not only target testing but to collect large data base that would allow better understanding of maximum power output in any number of conditions. So if it was started 10 years ago you would already have enough knowledge to presume with high probability where is that magic line, regardless of how 'fat' that line actually is.
Because I don't believe that there is no line.
The further we go the more this story becomes Zeno's paradox, infinitely small gains by which we will never reach the line that actually is just around the corner.

But the line (however fat) will be significantly different for a rider like Cavendish than for say a rider like Froome. How do you resolve that if you are not going to monitor every ride by every rider?

And how do we calibrate our new dopeometer if we don't know whether those W/kg have / have not been achieved via prohibited means?

#### 131313

Alex Simmons/RST said:
Yeah, good thoughts, just tossing stuff about...

But do we really need power meter data from races to know who to target and whose performances have improved significantly? Surely examining race results already does that and far far cheaper.

Alex, I'm gonna answer this, tomorrow..when I'm a little less tired. It's a good question and I have some thoughts, I'm too beat to type right now.

#### zastomito

Alex Simmons/RST said:
But the line (however fat) will be significantly different for a rider like Cavendish than for say a rider like Froome. How do you resolve that if you are not going to monitor every ride by every rider?

And how do we calibrate our new dopeometer if we don't know whether those W/kg have / have not been achieved via prohibited means?

I said in an earlier post that you would be having data for different type of riders. So you are bound to have data for sprinters as well as GT contenders. In a various type of races. I am sure scientific community is able to establish some boundaries in specific circumstances based on that.

I wouldn't call it dopometer, think no one would, but agree with MI that it would be kind of a power passport that would be able (combined with ABP) to eventually diminish doping advantages to a point where they become irrelevant.

#### Alex Simmons/RST

zastomito said:
I said in an earlier post that you would be having data for different type of riders. So you are bound to have data for sprinters as well as GT contenders. In a various type of races. I am sure scientific community is able to establish some boundaries in specific circumstances based on that.

OK, I missed that. Honestly though, I can't see it making physiological, nor practical sense. Let's face it, one can train for different type of performance outcomes, and change one's power profile accordingly.

Obviously you can't turn a 100% slow twitch guy into a track match sprinter but power profile changes can and do occur, both acutely and chronically.

I'm pretty sure the scientific community will say performance is on a continuum, and that any form of categorisation is but an arbitrary one used more for convenience than anything else.

I can just see it, a rider like Sagan gets put into one category, which might class him a probable doper, but it's all green lights if he were in another category. Who makes that choice?

Then there is performance context. Environmental conditions. Race strategy/tactics. Personal motivation. Duration of climb. Chronic and acute training loads. Rest of life factors. Personal stress, medical factors and so on. The physiological is but one part of the performance story.

zastomito said:
I wouldn't call it dopometer, think no one would, but agree with MI that it would be kind of a power passport that would be able (combined with ABP) to eventually diminish doping advantages to a point where they become irrelevant.

OK. What I don't get why the focus should be to keep a lid on performance, when we should be focussed on keeping a lid on doping. The two are not the same thing.

Like I said, I'm not against it, but I still need a lot of convincing that it's really of much value / will give a better return on anti-doping investment versus other options.

Not that convincing me matters diddly squat as it's the administration of the sport that needs convincing, and this is just pub chat fodder. So far though I find the arguments to be weak/not well thought through, IMO. YMMV.

#### acoggan

Krebs cycle said:
What happens if they do cross the line though? Do they get a warning? a ban? Can you see the logical outcome of this reasoning? Since power output is the dominant factor that determines velocity, then think about what that would mean for say track cycling events..... no more world records after a certain point because it is considered to cross some line. Why would anyone bother trying to break the world record then? Its like saying we are going to put an artificial cap on maximal human performance.....

Bingo (although it's not like saying it, it is saying it).

Krebs cycle said:
Begone with you citius altius fortius.

Why do I hear Robin Williams saying that?