I think that on quite steep climbs where aero drag is very small, and where the rider's mass and that of his bike is known precisely, power can be easily calculated. At least to a margin of error such that one can see 'unbelievable' performances fairly easily. So, are they doping? Nothing is absolutely certain, but are you a betting man?

---

Advanced Aero TopTube Storage for Road, Gravel, & Tri...ZeroSlip & Direct-mount, made in the USA.
DarkSpeedWorks.com.....Reviews.....Insta.....Facebook
--

Going up a hill, you can quite accurately determine minimum watts/kg figures for the bike+rider, since you can easily determine how many meters they climbed and how fast they did it. And if the hill is sufficiently steep, then you can assume that your minimum watts/kg value is fairly close to the true watts/kg value, since aero and rolling-resistance will be close to zero compared to the amount of energy required to climb the hill.

Riders tend to lie about their weights (or at least, they have competitive and doping-suspicion incentives to), so it gets harder to determine absolute wattages.

---

STAC Zero Trainer - Zero noise, zero tire contact, zero moving parts. Suffer in Silence starting fall 2016

It can be very accurate. Probably the best at it is @ammattipyoraily on twitter--Vaughters has even said he's very good at it and has shared power files from his pros with him fwiw. He's typically within the margin of error of a power meter with his estimates.

http://alex-cycle.blogspot.com/...er-to-body-mass.html

That assumes there is no wind. The speed-power relationship even on steep climbs is still pretty sensitive to air movement. So unless you know the wind vectors along the course, and level of draft assistance (which is greater for headwind climbs), then there will be a sizeable margin of error.

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

Chris Froome is climbing for 30-40 minutes at around 400 watts.

Is there a point at which the actual number means he is using PEDs? No

Yeah, that guy is pretty good. Always enjoy seeing his tweets.

How can someone put a number on human performance?? Has it been proven that a baseline of 6-6.2 watts/kg for example is the highest anyone will ever go?

And the whole argument about the Tour being the fastest it's ever been, even faster than the EPO (90-2000's). What about the fact that "hopefully:" teams in general are training smartly nowadays, with all teams wanting to be strong and at the front, rather than just following the USPS or T-Mobile bus. More so the fact that aerodynamics and equipment have improved immensly, where every team is doing the maximum on every stage. So why wouldn't the AVG speeds get faster in general across the board?

Thoughts?

---

Carson Christen
Sport Scientist / Coach
Torden Multisport

Actually you can estimate a ceiling for human performance. The guys already mentioned in this thread who do the analysis have been extrapolating what these w/kg outputs require in terms of physiology and efficiency. Physiology in terms of VO2 in ml/kg/min and the % of that VO2 over a period of time, and efficiency in terms of cycling efficiency which has had multiple studies and established a fairly solid 'range' of efficiency numbers. So for example when you see performances that mean an output of say 90% of VO2 max at 98 ml/kg/min for 1 hour at an efficiency that is well outside the pre established typical range you are either looking at some mutant/freak of nature or a doper.

I think the actual numbers are to a large degree irrelevant in the PED debate. If someone is capable of blitzing a field of known (i.e. have been busted) drug users and putting up times comparable with the performances of the past which we KNOW were fueled by some very, very effective PEDs then it probably doesn't take much of a leap of faith to question the athlete. When said athlete was at the age of 25 unable to keep up with the peleton (DQed from the Giro in 2010 for holding on to a motorbike after being badly dropped) but is now the best at the game only five years later based on "superior training techniques" the questions are impossible to ignore.

When not only the GC contender but also his team mates are besting the best is it likely that the team are simply training better than everyone else and is full of super-responders? Or is it more likely that a sport with a long and ongoing history of doping might just have its next US Postal?

The speed-power relationship even on steep climbs is still pretty sensitive to air movement.

Sure, but there are a lot of weather stations around. All the better if winds are mild. For drafting it's best to look at just the part where the rider is solo.

I'd be surprised if the error in W/kg was more than a couple percent in that case.

There's already 145 pages on this over on the Clinic: http://forum.cyclingnews.com/...3537a9c7c16d8a0c0735

---

Professional Athlete: http://jordancheyne.wordpress.com/ http://www.strava.com/athletes/145340

Coaching Services:http://www.peakformcoaching.com/

Do you have the time and distance data for when a rider is solo? Most don't. They only have the timing for the specific timing points on the climb.

Local town weather reports of wind are inadequate. Air movement is variable in and around the immediate topography of a climb. The shape of mountains, where trees are, the nature of retaining walls all affect wind direction and flow where the riders are.

Put it this way, an unaccounted for very slight wind that you cannot even feel is enough to create a greater than 2% error in estimated power made from climbing speed. If you can see a flag moving, then the potential for error starts to get quite large.

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

What a good reason to not read that forum. I wonder how many pages was devoted to Cancellara's bicycle motor.

Local town weather reports of wind are inadequate. Air movement is variable in and around the immediate topography of a climb. The shape of mountains, where trees are, the nature of retaining walls all affect wind direction and flow where the riders are.

I live near a long switchback climb so I have a lot of experience with this. What you say is certainly true in spots, but it tends to cancel out over the whole climb. The best is a general tailwind going in the direction of the switchbacks, but it isn't a large benefit. Next best is no wind at all. And wind from any other direction is nearly a wash. Not a lot of difference from best to worst.

Estimates could be improved with an adjustment factor calculated using actual data of riders other than the target rider that is applied to the target's estimate.

you're not for a moment suggesting that any team in this years Tour are going "Postal" are you?

cycling is absolutely clean... as clean as any other professional sport in the world... and I'm not just suggesting the main Euro & USA pro sport either...
speed skating - clean
MLB - clean
NFL - clean
Athletics - clean
Football(soccer) - clean
weight lifting - clean
all the combat sports - clean
cycling - clean - well as clean as those above

Wish I could find that study which examined climbing power estimates versus actual...

Edit: found it:

http://www.fredericgrappe.com/...s/2015/01/Millet.pdf

Quote:

Conclusions: Aerodynamic drag (affected by wind velocity and orientation, frontal area, drafting, and speed) is the most confounding factor. The mean estimated values are close to the power-output values measured by power meters, but the random error is between ±6% and ±10%. Moreover, at the power outputs (>400 W) produced by professional riders, this error is likely to be higher. This observation calls into question the validity of releasing individual values without reporting the range of random errors.

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

Sure, that can be one means to reduce error, depending on the rider and how similar their climb was to those you are seeking estimates for.

For a moment let's imagine that we have perfectly accurate power data for all riders, or at least for the ones that matter.

Now what?

How will knowing power data result in a doping sanction?

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

FWIW to you, both the Ferrari model and the CPL model have been repeatedly validated in cases where we have known the power output as measured.

To directly quote Ross Tucker:
"The estimate for Froome ranges between 6.1 W/kg (Ferrari method) and 6.2 W/kg (CPL method) for the 41 min climb.
For the sake of comparison, the estimates compare favorably to estimates of Robert Gesink’s power output. As per his MEASURED power output reported on Strava, he produced 409 W (5.8 W/kg if we go with 71kg, halfway between his Strava-listed and team-website listed mass), while the CPL method estimates 408.8W (5.8 W/kg) and Ferrari’s formula 5.86 W/kg. Conservatively then, I’m once again confident that the estimate for Froome is not outrageously wrong (I’d say it’s exceptionally accurate), and that he was at least at 6.1W/kg (calculations & times, as always, provided bythe excellent Ammattipyoraily)
Quintana, incidentally, at 5.9W/kg, was just slightly above what we’ve seen from the Tour winners for the last four years, for the length of the climb, while everyone else was down on estimates, so a howling tail-wind won’t cut it this time."

It was also noted that the modelling of Froome's 2013 Ventoux climb produced an estimated power output that was only a single Watt above/below (I don't remember which) the actual number reported in the hacked/leaked power file this week.

Everybody will have different opinions on evidence of exceptionally high power output being used as evidence of guilt, but I think we have reached a point where we can at least trust (within limits) the accuracy of the models when applied during often used climbs.

...which FWIW is exactly as I concluded when people first started going down this path. That is, such estimates are pretty good on average, individually generally w/in a few percent, but occasionally off by as much as 5-10%.

Now, ask me who is going to win the Republican nomination for president. ;)

You mean our model:

http://www.cyclingpowerlab.com/ModelValidation.aspx

Cool. Can you post the data and methodology used for verification?

Also according to Tucker this week, 6.1W/kg for ~40-minutes equates to having a VO2max in the mid 90s. That's Vayer-esq maths.

For a scientist, I think he is substantially over confident in the precision of these power estimates.

Gesink was mostly solo, no? Froome was not.
Who tested the validity of Gesink's Pioneer power meter data?
Who measured the wind along the course?


Impressive. But just because two methods came up with the same number doesn't mean they are right. There may have been a similar bias or even random error in both. e.g. Froome used non circular rings, known to inflate SRM power meter readings, even more so on a climb. So how come the estimate matched the power meter? Does that not suggest a little more rigour might be worthwhile?

Whatever one might think of what is plausible/implausible power output, a dopeometer it ain't.

I don't, and so far the only actual published peer reviewed science testing these very models on climbs with proper controls doesn't either. Well to be fair, they attained good accuracy over a large number of samples, but precision was not so flash with much random error to rely on a single data point.

In science you don't just go looking for data that support your theory, you also rigorously look for why it might be wrong and see if these reasons can be legitimately eliminated, or if the theory needs adjustment, or that its validity applies with certain conditions or caveats. Until that's done, I think scientists like Tucker should be careful in over claiming the precision of such power estimates.

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

Rider X, Y or Z may be doping or may not. I don't know because either is physiologically plausible and this approach is just not going to tell us either way. There are of course those who think variations in performance (a power passport) will flag suspicions. Again, since doping and non-doping reasons are equally plausible explanations for changes in performance, I fail to see how.

And do what, tell us that anti-doping should be focussed on the riders they are already focussed on?

So I come back to my key question:

How will knowing power data result in a doping sanction?

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

I agree with you re: power data being used as evidence of guilt.
My view is that it is NEVER evidence of guilt - It is not a source of answers, but merely a source of better/more targeted questions.

Of course there is a greyscale of physiological plausibility associated with higher and higher Wattages, but I'm in no position to put a line in the sand over what is plausible/implausible for a clean rider.

My original point was never that a power file can lead to a sanction or that rider X or Y is doping because they did xxxW. It was that, IMO, the models have reached a point of accuracy that allows us to have an acceptable degree of confidence in them.
An error of 1-2% is perfectly acceptable as far as I'm concerned. Not many power meters claim an accuracy greater than 2%.

With regards to your comment about Gesink's power file:
I accept the possibility that it wasn't calibrated properly, although I find it unlikely.
You surely must attribute some weight to the idea that if we have two methods of estimating power that give results within ~10W of each other (sure, nothing special in itself) and then we add to that a figure that was actually measured by a power meter on that rider's bike, and it falls within that same small window, then the methods of estimation have been somewhat validated?

I'm putting a point across that the models are doing a good job. That's it.