Power Data Estimates for the climbing stages

[IzzyStradlin]

Here's an example from a different sport - athletics. Compare and contrast the times in which 5000m and 10000m World and Olympic gold medals are won with the World Records (or even season's bests) in those events. The times are always a lot slower. (Mo Farah, for example, won the 5000m in 2012 with a time 80 seconds slower than the WR and almost a minute slower than his PB.) But I'm pretty sure everyone made the most effort they could in all those races - so why the discrepancy. Answer: It's how they proportion the effort that is key. And that's the variable that the power comparison guys ignore - the proportionality of effort.

Perfect example. 80 seconds slower than PB is not a max effort. It might be hard, he will be tired, but it is not, by definition, a max effort. He could have gone faster, but going faster carries a risk of blowing up and losing, so he didn't. Ergo, not max. Maybe there is a max 1k somewhere in there, but it is not a max effort for the length of the race

....which is why Vetoo typically focuses on segments of the climb rather than "the whole climb" to do his estimates. Eliminates early drafting or tactical riding, focuses on the steepest parts where the contenders are relatively isolated and most likely to be maxed out.

Like I said, we can never know if a winner gives a truly max effort. But the contenders who didn't win, most likely did.

Regardless, riders can not go faster than max efforts on extended climbs. So there shouldn't be too many super-fast surprises. Tactics will always cause them to ride slower. So for the most part, people do the estimates, they are 6+w/kg and we all hope it is really a max effort, cuz if it's not...

 

[ralphbert]

If you've ever tracked your own power, you would see that max efforts are very close regardless of tactics. My 20min best steady pace is within 2-3% of my best with attacks, lulls and then all out for the last 2k.

Yes that is my experience too. You are your number, trying harder or using tactics doesn't change it much.

Which is why the mountain tt used by US postal and Sky is so effective. Send the boys to the front and ride at the leaders threshold in the draft. One by one they burn up but the leader ends up going the mountain quickest possible time. Very difficult to combat against this unless you have a higher threshold

 

[ralphbert]

But when they start their maximum effort does.

....

Everything certainly does not even out.

I'm not sure I agree with you. It is counter intuitive but if you ride a section of the climb below threshold and the last section the climb at max, above threshold your avg climb ends up about same or slightly lower power than the complete climb at threshold. So where you start max effort isn't as critical as you first think.

 

[Race Radio]

But I'm not talking about 'many variables'. I'm talking about one - how the actual racing panned out. Races are not run with identical strategies, and variation in strategies has a large impact on times and power outputs but it is a factor ignored by power comparisons. My example was to show that climbs where the significant attacks were launched twelve minutes apart has a large impact on times but it is never mentioned.

Rather than trying to patronise me by telling me to read up on the topic (I have a mechanical engineering degree, I've done plenty), actually read what I have written.

I tell you to read up on it because you keep saying it is never mentioned. This specific variable was mentioned often. As were other variables, like the slow early pace of the climb. There is an entire thread on the wind direction on Ventoux. Just the wind direction.....434 posts.

Ross and Mike have written in detail about how tactics, wind, drafting, etc effect calculations. They compare their calculations with SRM files of riders in the race to insure accuracy. They have even pointed out to riders when their SRMs were not calibrated correctly . Veetoo often will post multiple calculations based on the amount of draft, the tailwind/headwind

Some do their calculations in a vacuum and ignore all the various elements, most don't.

 

[IzzyStradlin]

I'm not sure I agree with you. It is counter intuitive but if you ride a section of the climb below threshold and the last section the climb at max, above threshold your avg climb ends up about same or slightly lower power than the complete climb at threshold. So where you start max effort isn't as critical as you first think.

Exactly. If they ride up Ventoux at tempo and then do a blazing fast final 5k, it will either average out to a max effort, or it will be slower overall and not worth much statistically.

 

[Alex Simmons/RST]

The point of measuring power is to remove context. And it is very useful.

It seems that the biggest critics of veloclinic/vettoo have very little experience using power data (or have vested interests that benefit from the obfuscation of the data). Once you've trained with power for awhile you realize that max-output is very predictable.

I'm a critic, and I'd suggest my understanding of use of power meter data is at the higher end of the scale and I have no interest in obfuscation of data.

Edit just to add:

My main criticisms are:
- the persistent lack of error analysis or error ranges with the estimates of power from climbing speeds
- the tendency by some to focus on individual data points, when really all this method can properly do is provide population level analysis

 

[Alex Simmons/RST]

It appears you have not read much on this topic. All of these elements are discussed and factored in.

To be fair, this is the clinic and not the professional road racing forum.

And the main point of discussion in this thread is the use of such data as a means to establish the impact of doping.

It's correct that Ross Tucker points out that single data points should not be singled out, but that's precisely what seems to happen when such estimates are made.

 

[Alex Simmons/RST]

Tactics just don't have much effect on climbing speeds if riders are at, or near, max effort. Everything averages out.

No, it doesn't simply average out.

 

[Alex Simmons/RST]

....which is why Vetoo typically focuses on segments of the climb rather than "the whole climb" to do his estimates. Eliminates early drafting or tactical riding, focuses on the steepest parts where the contenders are relatively isolated and most likely to be maxed out

Does that then mean we end just up with a lot of data points at the short end of the power-duration curve?

 

[IzzyStradlin]

I'm a critic, and I'd suggest my understanding of use of power meter data is at the higher end of the scale and I have no interest in obfuscation of data.

What is your criticism?

And what do you think about the accuracy of estimates, Vetoo's in particular, when compared against data provided by riders?

 

[Parker]

I tell you to read up on it because you keep saying it is never mentioned. This specific variable was mentioned often. As were other variables, like the slow early pace of the climb. There is an entire thread on the wind direction on Ventoux. Just the wind direction.....434 posts.

Ok.Maybe I missed it. Please give me a link to you, Ross or Mike addressing the differences in race strategy when comparing Froome to Armstrong on either Ventoux or Ax3. Anything which mentions their point of attack. When they fully committed.

Ross and Mike have written in detail about how tactics, wind, drafting, etc effect calculations. They compare their calculations with SRM files of riders in the race to insure accuracy. They have even pointed out to riders when their SRMs were not calibrated correctly . Veetoo often will post multiple calculations based on the amount of draft, the tailwind/headwind

I'm not talking about f'ing wind. I don't care about f'ing wind. As I've said - even if the data came from SRMs, the point still applies. It's not about accuracy. My point is all about the racing - nothing else. Please understand that. Look at my 5000m examples.

Some do their calculations in a vacuum and ignore all the various elements, most don't.

And you, Ross and Mike ignore the biggest element of all - race strategy. You're not a scientist, so I'll excuse you. But when the likes of Ross start sucking up to Twitter characters like Digger, alarms bells should ring about the intentions of their 'science'. It's now about profile with them.

 

[Alex Simmons/RST]

I'm not sure I agree with you. It is counter intuitive but if you ride a section of the climb below threshold and the last section the climb at max, above threshold your avg climb ends up about same or slightly lower power than the complete climb at threshold. So where you start max effort isn't as critical as you first think.

The mean maximal power-duration relationship is not linear, and neither is the rate of recovery from supra threshold effort, so how you attack a climb very much impacts what power you end up with.

 

[IzzyStradlin]

No, it doesn't simply average out.

Have you ever seen anyone significantly exceed their wattage PR with tactical riding? As opposed to steady effort? On a 20min+ climb?

 

[IzzyStradlin]

I'm not talking about f'ing wind. I don't care about f'ing wind. As I've said - even if the data came from SRMs, the point still applies. It's not about accuracy. My point is all about the racing - nothing else. Please understand that. Look at my 5000m examples.

Ultimately, you seem to be arguing that tactical riding, ie going slower and then attacking, can cause riders to have a higher speed than what is understood to be their personal max (speed/power). Right?

If so, I disagree. And most anyone who has spent much time with a power meter probably would also.

 

[Parker]

Ultimately, you seem to be arguing that tactical riding, ie going slower and then attacking, can cause riders to have a higher speed than what is understood to be their personal max (speed/power). Right?

Ultimately my argument is that climbs aren't time trials - they're races. And how they pan out is very varied. And that's a factor that the power guys don't acknowledge.

 

[Race Radio]

Ultimately my argument is that climbs aren't time trials - they're races. And how they pan out is very varied. And that's a factor that the power guys don't acknowledge.

If you have read anything guys like Ross Tucker write you will see that this is not only clearly acknowledge but it is often central to the discussion.

 

[Alex Simmons/RST]

What is your criticism?

And what do you think about the accuracy of estimates, Vetoo's in particular, when compared against data provided by riders?

I went back and edited an earlier post to add those.

Mostly it's to do with an assessment of errors and the tendency by many to single out the data points.

As for comparing with power data, I'd want to know a lot more about the validity of the data from the power meter as well (there are many factors to consider here).

I would prefer data were presented with error ranges.

But then I don't think having everyone's power meter data will make a lick of difference to the issue at hand, i.e. who is and is not doping, or that it will do anything about reducing doping's insidious influence on the sport.

 

[Alex Simmons/RST]

Have you ever seen anyone significantly exceed their wattage PR with tactical riding? As opposed to steady effort? On a 20min+ climb?

Yes, but not often, and of course that then ends up revealing something about the state of their fitness.

More often of course the value is lower, which is the problem when one tries to assess such climbs in context of a rider's overall capability.

 

[Parker]

If you have read anything guys like Ross Tucker write you will see that this is not only clearly acknowledge but it is often central to the discussion.

I've read lots of it. Now show me an actual link where they acknowledge race strategy is an overarching problem with analysing climb data, with discussion of how they provide balance for this situation (rather than an excuse for a low figure). Because as far as I can see, they just chuck out the figure.

 

[IzzyStradlin]

Ultimately my argument is that climbs aren't time trials - they're races.

And I would argue that if you spent much time with a power meter you would realize that climbs are closer to TTs. (Given that they are longer than 20min and steeper than 5% or so.)

 

[Alex Simmons/RST]

If you have read anything guys like Ross Tucker write you will see that this is not only clearly acknowledge but it is often central to the discussion.

I agree and think he has done a fair job in making this point in more recent times. I was pretty critical at one time.

I don't agree with the notion of a clear indication doping or sans doping plausibility ascertained via power output though (e.g. suggesting there is a magical number one crosses). The line is just too fat and blurry and individually variable as far as I'm concerned.

 

[Parker]

And I would argue that if you spent much time with a power meter you would realize that climbs are closer to TTs. (Given that they are longer than 20min and steeper than 5% or so.)

Neither you nor I are riding Grand Tours. Experience with power meters is not relevant. Race experience at that level is.

 

[Alex Simmons/RST]

And I would argue that if you spent much time with a power meter you would realize that climbs are closer to TTs. (Given that they are longer than 20min and steeper than 5% or so.)

That's true (a power trace on a long climb is a lot smoother than on flatter terrain), but there are still variables in play.

The question is the level of impact of each of the variables.

As I've said before, this is mostly pub chat fodder, and a bit of fun.

It's when the estimates are used inappropriately, or that people represent them as being more precise than they might be, that I show a modicum of concern, well as much concern as one posting online might, which is not much.

 

[ralphbert]

Yes, but not often, and of course that then ends up revealing something about the state of their fitness.

More often of course the value is lower, which is the problem when one tries to assess such climbs in context of a rider's overall capability.

In my understanding no, not possible, going over your threshold power will always result in lower average. Ie if your P20 is 400w and you ride at 450w for 10min your next 10min will always be less than 350w.

Regardless this is not really a problem it just means a person's capability is higher than their performance in most instances. If their performance is already startling and the pacing is not ideal then can say there capability is even better.

 

[ralphbert]

Neither you nor I are riding Grand Tours. Experience with power meters is not relevant. Race experience at that level is.

Eh. These guys are still human, physiology still applies them and us alike.