Froome's SRM data on Ventoux

[King Boonen]

Re: Re:

You are naming the riders who won the most in cycling. The greatest was born in Lakewood, CA.

Nope, not just the winningest. Hinault was the coolest, hardest, greatest rider who ever competed. He's quite possibly the only person I would ever get star-struck meeting

 

[M Sport]

Re: Re:

What I think Sky should have reported yesterday was how the Left Right balance effected the power. They overstated the effect of the rings but neglected the balance issue. That in my opinion could swing it back closer to zero.

Sky will know that variance.

Sky may have an idea of general or typical trends for a rider, but power imbalance is variable and multifactoral. It varies with position on the bike, fatigue, power, gear/cadence, inertial load (e.g. flat v climbing), any acute injury or niggle and so on. It varies during a ride and from ride to ride. So if only measuring left leg (although they may be using a two sided measurement device not yet available to purchase) then the error will be variable and unknown.

Asymmetry is also normal.

In highly trained riders I would expect that imbalance would be less variable than in untrained or less trained riders, I'm not sure if there are are studies or not other than what's available using less trained riders. Typically in right handed or right side dominant riders the Stages under read.

The reason I made the comment "Sky will know" is that they were using SRM's in training in conjunction with their Stages, so I would expect some sort of pattern or trend emerged. In the spirit of transparency I thought it should at least be mentioned.

 

[M Sport]

Re:

Here is the thing I dont get. Wiggins and Froome are two guys renowned for staring at their power meters as they cycle. They are riding to a wattage.
But they using osymetric chain rings. But the power meter fails to display accurate wattage if you are using osymetric chain rings.
What wattage are they then riding to? What are they staring at?

They can be reasonably precise (1.5% - 2%) but not accurate.

 

[Alex Simmons/RST]

Is there no way to calibrate a power meter taking the error margin into account?

No, because as I've said several times already, the error is not constant but variable.

 

[Alex Simmons/RST]

Re: Re:

What I think Sky should have reported yesterday was how the Left Right balance effected the power. They overstated the effect of the rings but neglected the balance issue. That in my opinion could swing it back closer to zero.

Sky will know that variance.

Sky may have an idea of general or typical trends for a rider, but power imbalance is variable and multifactoral. It varies with position on the bike, fatigue, power, gear/cadence, inertial load (e.g. flat v climbing), any acute injury or niggle and so on. It varies during a ride and from ride to ride. So if only measuring left leg (although they may be using a two sided measurement device not yet available to purchase) then the error will be variable and unknown.

Asymmetry is also normal.

In highly trained riders I would expect that imbalance would be less variable than in untrained or less trained riders, I'm not sure if there are are studies or not other than what's available using less trained riders. Typically in right handed or right side dominant riders the Stages under read.

The reason I made the comment "Sky will know" is that they were using SRM's in training in conjunction with their Stages, so I would expect some sort of pattern or trend emerged. In the spirit of transparency I thought it should at least be mentioned.

Pedalling imbalance is not unique to non-elite riders.

And pedalling imbalance is not consistent for an individual rider either. It varies due to a range of factors, during a ride, from day to day and also over longer time periods. It's a known unknown. Unless using a good quality independent crank/pedal power meter.

 

[red_flanders]

Is there no way to calibrate a power meter taking the error margin into account?

No, because as I've said several times already, the error is not constant but variable.

What is the range of possible error in percent?

 

[franic]

Funny blooper by Gallopin.
https://www.youtube.com/watch?v=RmiXBulLdlw&feature=share

By the way, am I the only one scared when a rider say “I think the peloton is really clean” while smiling?!

 

[Dear Wiggo]

Re:

Funny blooper by Gallopin.
https://www.youtube.com/watch?v=RmiXBulLdlw&feature=share

By the way, am I the only one scared when a rider say “I think the peloton is really clean” while smiling?!

Kerrison reading from a script. It is laughable.

Around 67.5kg.

Since when do you say "around" and then provide a decimal place for weight in kg.

Laughable.

 

[Franklin]

Re: Re:

Kerrison reading from a script.

He looked extremely uncomfortable. Then again, so would I if there would be a camera pointed at me, so that's more a bad choice by Sky as a spokesman than an indiction.

 

[Alex Simmons/RST]

Is there no way to calibrate a power meter taking the error margin into account?

No, because as I've said several times already, the error is not constant but variable.

What is the range of possible error in percent?

Even the error range will depend on various factors. For a power meter (upstream of BB that assumes constant crank rotational velocity) with circular chain rings, even these can have an error of several percent during a short period of hard acceleration. Here for an explanation:
http://alex-cycle.blogspot.com.au/2015/01/accelerating-sins-crank-velocity.html

Non circular rings exacerbate error this further.

For steady state riding, it will depend mostly on the nature of the eccentricity of the ring, then ring set up, inertial load (e.g. flat versus climbing), seated vs standing, and the rider's pedalling "style"*. So for steady state riding, an error variability range of a few % is possible, and of course shift the range higher for accelerations.

If you have a highly eccentric ring (they exist) then the overall bias error will be higher, and can be higher than 6%. Dan Connelly explains this:
http://djconnel.blogspot.com.au/2013/09/simulation-of-power-error-from-constant.html

You could decide to apply a greater slope value to your SRM data, which would lower power readings by whatever % you increased slope by. All that will do is move the variable error to oscillate about a different mean, there will still be a different bias error for different riding situations.

Of course if you use a power meter that cannot have its slope adjusted (which is pretty much all of them except SRM and Quarq), then that's not an option available to you.

* lay speak for the torque application profile of the rider

 

[dwyatt]

Sorry if i missed this already, but would a motor on the BB not show on a power meter at all?

 

[Saint Unix]

Re:

Sorry if i missed this already, but would a motor on the BB not show on a power meter at all?

It would show as a difference between the estimated value and the value on the power metre if the power metre value is lower than the estimate.

In other words, if the released value of 5,78W/kg is fresh from the power metre and not tampered with, an explanation for the difference between that value and the estimated value of 6,09W/kg, aside from power metre error, could be a motor putting out an average of about 20W for the duration of the climb.

It's an additional force that isn't accounted for in the formula. It also doesn't show up on a power metre, since the power metre measures torque on the crank arm, whereas a motor would just spin the crank from within the frame.

At least I think this is correct. I'd gladly be corrected by someone with more experience on the matter.

 

[MikeS369]

Is there no way to calibrate a power meter taking the error margin into account?

No, because as I've said several times already, the error is not constant but variable.

I know that but they could calibrate them at the average of the known error margin.

 

[The Hitch]

In the Vayer interview on off the ball he says that when he spoke to Kerrison and Dave Brailsford in 2013, they told him the opposite about the 6% to what they are saying now.

 

[Saint Unix]

I get the feeling that the oval ring error margin thing is being somewhat misinterpreted by some, including Vayer. I doubt there's any sort of noticeable performance gain from using them, or everyone would be going oval ring crazy, but that's not where the error margin comes from either.

It seems to be that some are assuming the ring is just magically creating a certain amount of extra watts out of thin air. However, if I've understood this correctly, the error margin comes from a problem with the power metre's ability to give an accurate estimate of the actual power because the shape and characteristics of the ring makes it harder to create an accurate formula for power (power=torque*angular velocity), since the oval ring will act as, say, a 56 tooth and a 52 tooth ring. as well as everything in between, at different points of the pedal stroke, instead of just being a constant 54. I guess there would be a way to get the correct readings if the cadence and power was constant, but that's borderline impossible to maintain during a 41 minute climb, and thus there will always be a small error in the readings.

Of course, the 6% error margin seems to be pulled out of thin air anyway. Lord knows where Kerrison got that.

 

[Tonton]

When I read Tyler's data giving full *** EPO a 10% advantage, studies and posts that suggest micro-dosing is 3-5%, and some put the oval chainring at 6% (legal), then I wonder why the entire peloton isn't using the oval chainring. Who would not take advantage of a legal 6% advantage? The answer is that it is not true. Maybe 1% with chances of losing the chain and pedaling in semolina (French expression). BS.

 

[ulrichw]

[...]I wonder why the entire peloton isn't using the oval chainring. Who would not take advantage of a legal 6% advantage? [...]

You're making the same mistake Saint Unix attributes to Vayer (and Greg Lemond made on the air for Eurosport).

The rings don't give a 6% advantage, and that's not what Sky (or anyone who knows what they're talking about) are saying.

Due to the idiosyncrasies of how most crank-based powermeters calculate power (which Alex Simmons references in his post a little earlier), the power meter will indicate an artificially high number with oval rings. So, for example, when it says 318 watts, the rider is actually only generating 300 watts.

Sky is arguing that Froome's actual power was less than the powermeter indicated, due to the error introduced by the oval rings.

 

[hfer07]

Blame it on the Chainring

I love it!!!

when that theory gets busted, then what? the Clothing?

 

[WildspokeJoe]

Two questions:

Has anyone asked Lance Armstrong to release his power numbers? I don't know if it would add any relevance to the discussion but it would give 'experts' a chance to study the data of someone who cheated and dominated for 7 years. He has admitted to doping so he has nothing to hide. Perhaps the UCI could use it as a bargaining chip. You give us all your data will will reduce your lifetime ban.

My second question is: Is there any data or information that would convince you that Froome is 'clean'? Yes or No. If yes, what. If no, why?

 

[Alex Simmons/RST]

Is there no way to calibrate a power meter taking the error margin into account?

No, because as I've said several times already, the error is not constant but variable.

I know that but they could calibrate them at the average of the known error margin.

Sure, but the average would only be a long term one, and not applicable to any one ride or parts of a ride. Think of it as precision v accuracy. It might improve overall accuracy (over say a season) but the precision would still be less than with circular rings, such that the level of error will still depend on all those factors I mentioned.

Or put another way, the more homogeneous is the riding you do, the better this approach would work, and of course the more variable are the type of rides you do, the more variance in the error.

Of course it's not a solution that will work with a Stages (or many power meters) since you cannot adjust the slope calibration like you can with an SRM.

 

[Alex Simmons/RST]

Re:

Blame it on the Chainring

I love it!!!

when that theory gets busted, then what? the Clothing?

No one has blamed anything on a chain ring. All that has been pointed out is people need to be aware of the very real error that such non-circular chainrings introduce to the power reported by most power meters.

Who knows with the Stages used by Froome? There may well be other sources of error we don't know about with this measurement set up (it might be less or more). Normal Stages are left leg only, and that's a variable accuracy red flag right there, although it appears Sky may be using a prototype dual sided model. Beta product. No one can tell how well it performs since no independent testing of the meter exists.

Then there are those relying on Gesink's and Ten Dam's data, yet they are from new meters (Pioneer) with very limited verification of accuracy. In a recent test by DC Rainmaker they were reporting substantially different numbers to other meters, e.g. differences of 9-12% on road rides to a Garmin Vector! Now the Pioneer might have been on the money, or it may not. But anyone worth their salt should be checking this sort of stuff before taking it at face value.

Power meters, like most cycling equipment, vary in quality. In the case of power meters, the most important consideration is the quality of the data they provide, everything else is a feature. Pro teams are advertising billboards for power meter brands just like they are for bikes, wheels, helmets etc.

 

[Alex Simmons/RST]

Re:

My second question is: Is there any data or information that would convince you that Froome is 'clean'? Yes or No. If yes, what. If no, why?

Perhaps if Froome (or any athlete) were the subject of a "Truman Show" style of permanent hour by hour publicly viewable scrutiny then one could say for certain. Else, proving one is not a doper is probably impossible.

 

[42x16ss]

Is there no way to calibrate a power meter taking the error margin into account?

No, because as I've said several times already, the error is not constant but variable.

What is the range of possible error in percent?

How long is a piece of string? There's so many factors that come into play it's crazy.

Temperature shifts can have a large effect, especially when the top of a HC Mountain is 10-12 degrees c cooler. Wind, tyre pressure (which can be affected by long stints on poor roads), surface conditions, rider weight (I know I can lose 2kg on a 5-6 hour ride, even when eating and drinking). Sadly it's not as straight forward as it could be.

 

[42x16ss]

Re: Re:

Blame it on the Chainring

I love it!!!

when that theory gets busted, then what? the Clothing?

No one has blamed anything on a chain ring. All that has been pointed out is people need to be aware of the very real error that such non-circular chainrings introduce to the power reported by most power meters.

Who knows with the Stages used by Froome? There may well be other sources of error we don't know about with this measurement set up (it might be less or more). Normal Stages are left leg only, and that's a variable accuracy red flag right there, although it appears Sky may be using a prototype dual sided model. Beta product. No one can tell how well it performs since no independent testing of the meter exists.

Then there are those relying on Gesink's and Ten Dam's data, yet they are from new meters (Pioneer) with very limited verification of accuracy. In a recent test by DC Rainmaker they were reporting substantially different numbers to other meters, e.g. differences of 9-12% on road rides to a Garmin Vector! Now the Pioneer might have been on the money, or it may not. But anyone worth their salt should be checking this sort of stuff before taking it at face value.

Power meters, like most cycling equipment, vary in quality. In the case of power meters, the most important consideration is the quality of the data they provide, everything else is a feature. Pro teams are advertising billboards for power meter brands just like they are for bikes, wheels, helmets etc.

Yep, it's easy to forget that the most important function of a power meter is the consistency of the measurement, not the actual number. As long as the same effort always registers as more or less the same number, all is well. About the only power meter I'd come close to taking as gospel would be an SRM ergo in a controlled lab.

Consumer power meters are great as training and racing aids but for now there's too many variables to use as doping detectors by themselves, unless there's an absolutely insane performance.

 

[franic]

Re: Re:

Blame it on the Chainring

I love it!!!

when that theory gets busted, then what? the Clothing?

No one has blamed anything on a chain ring. All that has been pointed out is people need to be aware of the very real error that such non-circular chainrings introduce to the power reported by most power meters.

Who knows with the Stages used by Froome? There may well be other sources of error we don't know about with this measurement set up (it might be less or more). Normal Stages are left leg only, and that's a variable accuracy red flag right there, although it appears Sky may be using a prototype dual sided model. Beta product. No one can tell how well it performs since no independent testing of the meter exists.

Then there are those relying on Gesink's and Ten Dam's data, yet they are from new meters (Pioneer) with very limited verification of accuracy. In a recent test by DC Rainmaker they were reporting substantially different numbers to other meters, e.g. differences of 9-12% on road rides to a Garmin Vector! Now the Pioneer might have been on the money, or it may not. But anyone worth their salt should be checking this sort of stuff before taking it at face value.

Power meters, like most cycling equipment, vary in quality. In the case of power meters, the most important consideration is the quality of the data they provide, everything else is a feature. Pro teams are advertising billboards for power meter brands just like they are for bikes, wheels, helmets etc.

What is really suspicious about Froome is that all these tiny errors add up in the same direction proving he’s not doping. I would expect powermeter measurement errors, wind, rolling resistance, weight etc… cancel one each other in the long run.
Regarding Gesink and Ten Dam, I guess people are confident about them because they are consistent with the models we have to estimate power on climbs. Now you can call it confirmation bias, but when I see smoke there must be fire somewhere.