Power Data Estimates for the climbing stages

[Dear Wiggo]

lol someone doesn't know the difference between joules and calories

Fail.

I am sure if you ask around, someone can explain it to you.

 

[Alex Simmons/RST]

Ten Dam said there was a headwind, as did the commentators (Phil and Paul). I suggest we add 0.5W/kg to the estimates.

Phil and Paul can't even get the rider's names right, let alone what the wind is doing. Only time I heard them talk about the wind was in reference to the final hundred metres.

The fact that we have such varied reports simply means that such estimates will have quite large errors in their calculation.

I looked back on video of the Armstrong/Pantani Ventoux ascent, and it seemed to be quite strong headwind for large sections, could be easily a 1W/kg difference for same time depending on the wind.

 

[karlboss]

I suggest noting what I wrote.

Whilst quoting my estimate you wrote this.

Greg Henderson tweeted he had a strong tailwind for the climb.

Assuming that's the case, I suggest dropping estimated W/kg numbers by at least 0.5W/kg, if not more.

If you don't mean in response to my estimate, why quote me?

 

[the sceptic]

Phil and Paul can't even get the rider's names right, let alone what the wind is doing. Only time I heard them talk about the wind was in reference to the final hundred metres.

The fact that we have such varied reports simply means that such estimates will have quite large errors in their calculation.

I looked back on video of the Armstrong/Pantani Ventoux ascent, and it seemed to be quite strong headwind for large sections, could be easily a 1W/kg difference for same time depending on the wind.

Maybe you should take a look at the climb today too, since your estimate is the only one that doesnt have a large error in it

 

[luckyboy]

Ten Dam said there was a headwind, as did the commentators (Phil and Paul). I suggest we add 0.5W/kg to the estimates.

About the wind - http://forum.cyclingnews.com/showpost.php?p=1301201&postcount=5667

 

[LongSprint]

Ten Dam said there was a headwind, as did the commentators (Phil and Paul).

A headwind is always noticeable, a tailwind not so much.

 

[Alex Simmons/RST]

Whilst quoting my estimate you wrote this.

If you don't mean in response to my estimate, why quote me?

I was referring you to where I said "Assuming that's the case,...", meaning we can't really and coupled with power from speed estimation based on a windy climb is folly, or comes with largish error bars.

The fact that we have a power file (assuming it's correct as well) from a rider that was ~20-minutes slower and finished with a group of 10 riders may help to get us to narrow the error bars a bit but I think there are still too many unknown variables to be pinning a number that precise.

 

[karlboss]

I was referring you to where I said "Assuming that's the case,...", meaning we can't really and coupled with power from speed estimation based on a windy climb is folly, or comes with largish error bars.

The fact that we have a power file (assuming it's correct as well) from a rider that was ~20-minutes slower and finished with a group of 10 riders may help to get us to narrow the error bars a bit but I think there are still too many unknown variables to be pinning a number that precise.

Do you know what the error bars are in this type of estimation? If you do I'm happy to add them. I thought adding "I'm not claiming any sort of accuracy with this method, I already know of a number of them." would cover the lack of accuracy.

Interesting that it is inside the range given by veloclinic. I found it more interesting that it was at the upper end, rather than the lower, as this method should estimate the power a little low when estimating from lower speed to higher and vice versa.

This is of course providing the SRM file is accurate.

 

[Alex Simmons/RST]

Do you know what the error bars are in this type of estimation?

I'm not sure we can really say to be honest, since there are unknowns with the errors as well.

The same site seemed to confuse the impact of an error in CdA (which would have a relatively small, albeit not insignificant, impact on estimation of power from VAM) with an error in the wind velocity assumption (which would be a very significant source of error in estimated power from VAM).

Like I've shown before with ADH climbing times, you can have a spread of 1W/kg in an estimation of power from VAM with only a marginal wind variance of +/- 2.5m/s.

2.5m/s is not a particularly strong wind. On the Beaufort scale that's rated only as a light breeze, only 2 out of 12 on the wind scale.

Wind matters a lot when determining power from speed, or speed from power, no matter the gradient.

An error in wind assumption can lead to a large error in either speed or power derived from the other.

Put it this way, if you can see flags fluttering on a mountain, forget about doing VAM to W/kg calculations.

 

[Le breton]

2.5m/s is not a particularly strong wind. On the Beaufort scale that's rated only as a light breeze, only 2 out of 12 on the wind scale.

To feel a 2.5 m/s wind at bike level, the wind 10 meters higher must be blowing at about 5 m/s.

Vehicles on the side of the road, plus spectators lined up all the way up the climb tend to shield the cyclists considerably, further reducing the apparent wind felt on the bike.

What I don't have a good feel for is the advantage drawn from the intermittent presence of (sometimes many) motorbikes a short distance ahead of the leaders, plus cars/ motorcycles overtaking the racers. They must superimpose a constant uphill draft, but I don't know how strong it's likely to be depending on circumstances.

 

[Alex Simmons/RST]

To feel a 2.5 m/s wind at bike level, the wind 10 meters higher must be blowing at about 5 m/s.

Vehicles on the side of the road, plus spectators lined up all the way up the climb tend to shield the cyclists considerably, further reducing the apparent wind felt on the bike.

What I don't have a good feel for is the advantage drawn from the intermittent presence of (sometimes many) motorbikes a short distance ahead of the leaders, plus cars/ motorcycles overtaking the racers. They must superimpose a constant uphill draft, but I don't know how strong it's likely to be depending on circumstances.

Yep, I get all that. Simple fact is, we just do not know how much wind these riders actually have to deal with but even a light breeze introduces a large error in W/kg estimates. e.g. 2.5m/s is insufficient to extend light flags yet such a breeze can throw out the W/kg estimate by nearly 10%!

Yet it was clear the flags were flapping in strong winds for much of the climb.

Here's the blog post to explain.

 

[MizunoMX20]

Original data is irizar's SRM, so no adjustment needed using this method at least for Irizar. Are you saying linearly interpolating from this data I should knock off .5 as there was a tailwind? That seems odd. That would imply it affected Froome but not Irizar

I use Strava myself as well. However, I do not use Strava as I ride, I use a different app (cyclemeter). Then when return home, I upload the CycleMeter GPX-file to Strava.

I think the Strava algorithm then transforms the data slightly, because I alsways get slight differences in for instance distance.

Therefore I would not automatically assume that Irizar's SRM numbers are identical to his Strava numbers (assuming he uploaded his SRM data into Strava).

 

[Bumeington]

Ventoux:
Irizar from SRM 324W at 76kg 15.67kph
Froome 22.86kph
Assuming power to weight is directly proportional (324/76)/(15.67/22.86)=5.95Wkg

Take from it what you will, I'm not claiming any sort of accuracy with this method, I already know of a number of them.

Where did you get the speed numbers from? I make it Irizar finishes at +19:20, so with Froome at 59:00 for the entire ascent, and I'd go with a power of 1.25 in the proportionality to get 6.08W/kg for Froome. With a direct proportionality it would only be 5.66W/kg.

Also, (324/76)/(15.67/22.86) = 6.22 W/kg

 

[Eshnar]

Power to weight is not directly proportional, not even power to speed is...

 

[Tyler'sTwin]

Now, assuming the aerodynamic efficiency to be equal for all riders (something that IS NOT...), considering the square speeds, the difference in power would be represented as follows:

Froome: 54.0 Km / h - 2916 (v2)
Porte: 52.3 - 2738, -6.5%
Mollema: 51.3 - 2635]

http://www.53x12.com/do/show?page=indepth.view&id=138

 

[Netserk]

Greg Henderson tweeted he had a strong tailwind for the climb.

That was sarcasm.

 

[acoggan]

https://strava.zendesk.com/entries/20959332-power-calculations

Note the date of that blog entry.

 

[Le breton]

Here's the blog post to explain.

I put your blog among my favorites

 

[karlboss]

Note the date of that blog entry.

Forgive me for assuming the info on the website no matter when it was posted was still relevant.

Where did you get the speed numbers from? I make it Irizar finishes at +19:20, so with Froome at 59:00 for the entire ascent, and I'd go with a power of 1.25 in the proportionality to get 6.08W/kg for Froome. With a direct proportionality it would only be 5.66W/kg.

Also, (324/76)/(15.67/22.86) = 6.22 W/kg

Speed numbers come from the SRM file where it says speed, and speed for Froome comes from the average that is posted for a similar interval, they are not the same, Irizar's is less,

Power to weight is not directly proportional, not even power to speed is...

Correct, but uphill they should be close.

I'm not sure we can really say to be honest, since there are unknowns with the errors as well.

The same site seemed to confuse the impact of an error in CdA (which would have a relatively small, albeit not insignificant, impact on estimation of power from VAM) with an error in the wind velocity assumption (which would be a very significant source of error in estimated power from VAM).

Like I've shown before with ADH climbing times, you can have a spread of 1W/kg in an estimation of power from VAM with only a marginal wind variance of +/- 2.5m/s.

2.5m/s is not a particularly strong wind. On the Beaufort scale that's rated only as a light breeze, only 2 out of 12 on the wind scale.

Wind matters a lot when determining power from speed, or speed from power, no matter the gradient.

An error in wind assumption can lead to a large error in either speed or power derived from the other.

Put it this way, if you can see flags fluttering on a mountain, forget about doing VAM to W/kg calculations.

I was referring to looking at a Power file and multiplying to estimate another riders power. I would guess, at best depending on drafting and wind around, 5%.

In your estimates of what wind can do to power on a climb which direction does the wind come from, to you just add 2.5m/s tailwind, because on a climb like ale d'huez that obviously can't happen.
I'll read your blog when i have a little time, looks interesting.

 

[Eshnar]

Correct, but uphill they should be close.

I highly doubt that (though I don't have data so I can't say I'm sure).

Power is directly proportional to weight only in vacuum, but air drag accounts for a consistent part of a real power output, and it is not related to weight at all.

For istance, if I take for true froome's average speed of 22.86 kph (--> 6.35 m/s) above posted, the power output due to weight-lifting only is simply (roughly, assuming the average gradient as 7%) Pw=0.07*69*9.8*6.35=300 W

which is quite low, hence all the remaining power Froome produced (how much? 100 - 150 W ) is due to drag resistance (for the most part) and therefore not proportional to weight.

 

[131313]

http://www.53x12.com/do/show?page=indepth.view&id=138

I can't see the link, but I'm glad to see at least one person is asking the right question.

All of this "wind speed", "power calculation" stuff is complete nonsense. Unless there's a rider in that front group with known weight and a well-calibrated power meter (two very big "ifs"), determining w/kg is going to be fraught with enough error that it's completely useless to make any determinations regarding doping. This is particularly true on a climb like Ventoux that has high, variable winds and is exposed. Plus, strategy and race situation always play a part.

The real question is "how much more power did Froome have to put out compared to the other guys". If Ferrari's calculations are correct, then the answer is "a lot". Enough of a difference to be plenty suspicious, particularly given Froome's pedigree. >10% over the next non-Sky rider isn't exactly "marginal", it's the advantage you'd get with a very well-orchestrated doping program.

 

[red_flanders]

I can't see the link, but I'm glad to see at least one person is asking the right question.

All of this "wind speed", "power calculation" stuff is complete nonsense. Unless there's a rider in that front group with known weight and a well-calibrated power meter (two very big "ifs"), determining w/kg is going to be fraught with enough error that it's completely useless to make any determinations regarding doping. This is particularly true on a climb like Ventoux that has high, variable winds and is exposed. Plus, strategy and race situation always play a part.

The real question is "how much more power did Froome have to put out compared to the other guys". If Ferrari's calculations are correct, then the answer is "a lot". Enough of a difference to be plenty suspicious, particularly given Froome's pedigree. >10% over the next non-Sky rider isn't exactly "marginal", it's the advantage you'd get with a very well-orchestrated doping program.

Good post. I think you state the situation well.

 

[karlboss]

I highly doubt that (though I don't have data so I can't say I'm sure).

Power is directly proportional to weight only in vacuum, but air drag accounts for a consistent part of a real power output, and it is not related to weight at all.

For istance, if I take for true froome's average speed of 22.86 kph (--> 6.35 m/s) above posted, the power output due to weight-lifting only is simply (roughly, assuming the average gradient as 7%) Pw=0.07*69*9.8*6.35=300 W

which is quite low, hence all the remaining power Froome produced (how much? 100 - 150 W ) is due to drag resistance (for the most part) and therefore not proportional to weight.

My estimation rates power to weight as being proportional to speed. It is so far as gravity, rolling resistance and friction are concerned, however you have correctly noted not so far as wind resistance is concerned.
This method assumes things that for accuracy you can't assume, the two riders are the same size, were riding in an equal draft the same amount of time, and experienced identical wind conditions. We know none of these things are true. However as you note aerodynamic drag still makes up a large proportion. Let's assume it is different by a massive 30%, then that is still less than 10% of the total number.

This is a pub chat number.

Edit: If we look at sutherland's vs Horner AToC on both Sierra road and Mt Baldy, this method is accurate to 2-3%, low both times as predicted. If anyone has more SRM data from same day same race I could check more of these.

Edit edit edit: Sorenson vs Horner Tourmalet 2010, going by sorenson to horner, actual numbers 5.2 and 5.6, 11 minutes apart at the finish. Predicted power for horner from sorenson's data 6.2. Major cause of error drafting, Sorenson pulled early on for 11minutes at 6.6 while Horner maintained pace at 6.3, then sorenson dropped back through the field to the top, while horner drafted to the top behind the gesink group.

So depending on drafting I'm happy to say power meter, then using speed could be out by as much as 10%

 

[Alex Simmons/RST]

The real question is "how much more power did Froome have to put out compared to the other guys". If Ferrari's calculations are correct, then the answer is "a lot". Enough of a difference to be plenty suspicious, particularly given Froome's pedigree. >10% over the next non-Sky rider isn't exactly "marginal", it's the advantage you'd get with a very well-orchestrated doping program.

>10% more power than next non-Sky rider?

Quintana finished 29 seconds down on a climb that took nearly an hour. That's less than 1% difference in speed. That is most definitely not going to result in a more than 10% difference in power.

I can't see that linked website, so can't really comment or not if that's what they actually claim, but if it is, it's nonsense.

 

[Eshnar]

>10% more power than next non-Sky rider?

Quintana finished 29 seconds down on a climb that took nearly an hour. That's less than 1% difference in speed. That is most definitely not going to result in a more than 10% difference in power.

I can't see that linked website, so can't really comment or not if that's what they actually claim, but if it is, it's nonsense.

It was referring to the ITT, not the Ventoux stage.
edit: see post #1881