Power meter output readings

The foundations for successful riding

Re: Power meter output readings

Postby Alex Simmons/RST » Sun Sep 08, 2013 7:41 pm

Arlberg wrote:Xplora, I know the hill very well, it is the Zoo Hill in Mosman. It is not long or steep but it is convenient to where I live.

By the way, DCRainmaker tested the Stages very thoroughly and was unable to influence the power output readings when actively trying to put more and less power through the LH crankarm compared with the RH crankarm. He also found the Stages to be within 2% accuracy of the SRM powermeter.


DCRainmaker has never tested anything in comparison to an SRM. Indeed he's never tested SRM. He did test the Stages relative to a Quarq and a Powertap and a Garmin Vector.

When you examine DCRainmaker's comparison data, you will see that he has quite variable left/right power output and as a result the accuracy of the Stages is quite variable for him (and at times is well outside the 2% range).
User avatar
Alex Simmons/RST
Expert
 
Posts: 3367
Joined: Fri May 02, 2008 3:51 pm

by BNA » Sun Sep 08, 2013 7:51 pm

BNA
 

Re: Power meter output readings

Postby Alex Simmons/RST » Sun Sep 08, 2013 7:51 pm

Arlberg wrote:Is it possible to ride up the same hill twice, ride it slower the second time around than the first time around, yet have a higher power output on the slower ride than on the faster ride? (Using an actual powermeter to measure the power outpuy, not a Strava guestimation).

If it is possible, how is it?

If we assume power meter accuracy, and the bike and kit and rider are the same, then the primary reason for any difference in power output when climbing at the same speed is wind velocity.
http://alex-cycle.blogspot.com.au/2013/07/windbags.html

However there can be other smaller factors as well, rolling resistance variations (small impact on power) variability in pacing (will have small impact on average power depending on the number of acceleration/decelerations during the climb), drafting other riders (some reduction in power demand, with shelter from headwind a factor). These are second order impacts after wind velocity though.
User avatar
Alex Simmons/RST
Expert
 
Posts: 3367
Joined: Fri May 02, 2008 3:51 pm

Re: Power meter output readings

Postby Arlberg » Sun Sep 08, 2013 9:48 pm

Very useful Alex. Thanks for giving your time to explain it so well.

What is your opinion on using either Zero Averaging or Non Zero Averaging on the powermeter? Which should I use, or should I say, what does everyone else use so I can compare apples with apples when comparing myself with other riders?

I only just got the Stages and currently have it set on Non Zero Averaging because I want to know what my power readings are when I apply force/speed to the pedals and my average overall power. It seems pointless to include the times when I am not pedalling (ie Zero power) in that calculation.

Also what setting should I have it on when doing an FTP test? I am about to do my first ever FTP test so I can set up the training zones.
Arlberg
 
Posts: 265
Joined: Mon Oct 03, 2011 5:43 pm

Re: Power meter output readings

Postby Arlberg » Sun Sep 08, 2013 10:15 pm

Xplora wrote:OP, I had a quick sticky on Strava at the segments near the Zoo and it's a reasonable distance. I'm guessing you were doing repeats because the area is not conducive to centuries without covering the same ground :lol:

I'm going with compensation as the explanation because it's a long hard climb if you are pushing fast; yes, I 100% agree that the Stages is accurate within 2% but the leaderboard shows people putting out 400+W average with powermeters. That comes to 8W variance of actual power, and it's hard to say if the reference is out as well... calibration is important. Anyways, the point is if you're concerned about your readings, then you're unlikely to be satisfied if you are trying to compare same day efforts and you aren't willing to accept internal variations (you pushing the pedals differently).

DCRainmaker is just one example, and he has an enormous amount more work to do with publishing his data on L/R splits to make meaningful comment. Bear in mind, that we just haven't had L/R power until recently; his data wouldn't focus on that. As a triathlete, he's concerned about pacing more than all out efforts.


Xplora, here is an actual segment, and my fastest ever time recorded a few months ago before I got the powermeter. (So the power shown here, 362 watts, is a Strava guestimation). Incidentally, no matter how hard I try, I have never been able to get within 10 seconds of this time since.

1. http://www.strava.com/activities/52376051#941978185

Here is the same ride with an actual power figure, 402 watts, from a test ride with the Stages powermeter from a few days ago.

2. http://www.strava.com/activities/80447150#1635374827

Interestingly, you can see that the power output on the second ride (402 watts) as measured by my Stages is considerably higher than the Strava guestimation (362 watts) of the first ride, even though the time/speed and effort on this ascent was considerably lower than on the first ride, when I went all out. So either the Stages is innacurate or the Strava power guestimations are innacurate. More likely the latter because if you go to the segment leaderboard you can see that all the riders who used powermeters giving real readings have consistently higher outputs than the Strava guestimations.
Arlberg
 
Posts: 265
Joined: Mon Oct 03, 2011 5:43 pm

Re: Power meter output readings

Postby Xplora » Mon Sep 09, 2013 12:11 am

Chasing cars and other people is a wonderful thing ;) I wouldn't put any stock in Strava's power at all; it's a guess which is useful against your own strava, but useless against others. The most valuable thing you get from Strava is a guess of how heavy your opponents are. ;) Alex is on the money - Strava is guessing based on wind conditions which it can't predict, weather it can't predict. It just looks at a gradient and a change in speed, both of which are inaccurate for 99% of climbs (your mosman hills segment has a short section of 30%???) so the whole thing ends up being a bit silly. You'll get much more useful and accurate times on longer segments that are reasonably sheltered from the wind, to improve consistency.

Do you ride with other people often? Or "against" them? My best commutes were done trying to bury very strong riders in the Commuter Cup. Dramatically faster times... and I always went as hard as I could without competition. I'm guessing you were paced on the fast time or had a double espresso. It's easy to gain that much time for a one off. I've had some incredible improvements on some times last week. It was crazy, and I don't know what brought the improvement at all.

Ignore Strava power, use your PM. That's why you bought it; don't get too concerned about "results" either because the power info is more useful for pacing - either ensuring you don't over or under commit in racing or training. You would use the PM data to launch an attack towards the end of the segment, rather than the PM data to work out how strong you were - no one cares if you have the biggest power, because power to weight is the name of the game on a hill. The tail must not wag the dog. It's a very very expensive toy...
Xplora
 
Posts: 5553
Joined: Fri Dec 10, 2010 11:33 pm
Location: TL;DR

Re: Power meter output readings

Postby Alex Simmons/RST » Mon Sep 09, 2013 8:09 am

Arlberg wrote:Very useful Alex. Thanks for giving your time to explain it so well.

What is your opinion on using either Zero Averaging or Non Zero Averaging on the powermeter? Which should I use, or should I say, what does everyone else use so I can compare apples with apples when comparing myself with other riders?

I only just got the Stages and currently have it set on Non Zero Averaging because I want to know what my power readings are when I apply force/speed to the pedals and my average overall power. It seems pointless to include the times when I am not pedalling (ie Zero power) in that calculation.

Also what setting should I have it on when doing an FTP test? I am about to do my first ever FTP test so I can set up the training zones.


For power, include zeros in your averaging. It's not an average when you ignore some of the data. Post hoc analysis software will (or should) include zeros anyway. I can't say what other people do, and comparing power with others is not really the primary purpose nor of much interest. The main comparison should be with yourself.

The primary reason is that the time spent not pedalling affects what you can do when you are pedalling.

If you are specifically interested in inspecting what you did power wise during a part of the ride (e.g. a hill climb, a test, or an interval effort) then mark the interval and examine that separately.

Also keep in mind the utility of Average Power - it is useful for some things (e.g. overall energy output, examining intervals) and less so for others (e.g. getting a sense of the physiological "cost" of a variable effort ride). I can think of no sound physiological or training purpose for non-zero average power. There are however other means to examine data from variable efforts that are more useful, e.g. Normalized Power, or XPower.
User avatar
Alex Simmons/RST
Expert
 
Posts: 3367
Joined: Fri May 02, 2008 3:51 pm

Re: Power meter output readings

Postby sim-o » Mon Sep 09, 2013 10:58 pm

Alex Simmons/RST wrote:http://alex-cycle.blogspot.com.au/2013/07/windbags.html


That wind speed graph is brilliant Alex. It hows just how much the effect of wind is underestimated, and the reason why drafting is so significant in cycling. And... it only goes up to "Moderate breeze" :shock:!
2011 Merida Ride 93
2012 Apollo MTB
Image
User avatar
sim-o
 
Posts: 122
Joined: Fri May 07, 2010 1:19 pm
Location: Wanneroo, Perth

Re: Power meter output readings

Postby Alex Simmons/RST » Tue Sep 10, 2013 8:29 am

sim-o wrote:
Alex Simmons/RST wrote:http://alex-cycle.blogspot.com.au/2013/07/windbags.html


That wind speed graph is brilliant Alex. It hows just how much the effect of wind is underestimated, and the reason why drafting is so significant in cycling. And... it only goes up to "Moderate breeze" :shock:!

Well keep in mind that the graph is when climbing*. The impact on flatter terrain is even more pronounced.


* the reason for that is to demonstrate the potential error in W/kg introduced by the unknown wind variable when attepting to estimate power from climbing speed, which is is often quoted with unjustified precision by those who seek to calculate such estimates.
User avatar
Alex Simmons/RST
Expert
 
Posts: 3367
Joined: Fri May 02, 2008 3:51 pm

Re: Power meter output readings

Postby Arlberg » Wed Oct 09, 2013 6:24 pm

OK. So having bought a Stages powermeter, and getting some great advice from people on here, particularly from Alex Simmons's posts and also from his website, so that I could train more effectively, I wanted to calculate my FTP so I could then train in the various zones. So according to Alex and other websites I have been looking at I needed a road, preferably going slightly uphill to minimise external factors such as wind resistance etc, where I can put in a solid 20 minute effort (and preferably longer) without having outside 'disturbances' that would affect the reading such other vehicles, intersections, traffic lights, and any downhill bits.

Not easy to find such a road in Sydney. So I went to Thailand to find one, and here is the ride:

http://www.strava.com/activities/87913151

And here is the segment meeting all the above criteria for the FTP test in particular:

http://www.strava.com/activities/87913151#1823139418

My Stages powermeter indicates an average power output of 316W for the 32 minute 23 second effort, which was effectively a 30 minute time trial. So using Alex's formula and taking 93% of that to convert it, means an FTP of 294, +/- 3%. But in my opinion this powermeter reads generally too high. I have noticed this not just on this occasion, but on many occasions when climbing around Sydney, McCarrs, Galston, Bobbin, Pie in the sky etc and comparing my actual power output readings from my powermeter with other riders actual power output reading from their powermeter. (as opposed to the Strava power guestimations). You can see that most of the riders with powermeters who have climbed this hill considerably faster than I did it with supposedly lower power outputs. So something is wrong. Either my powermeter reads too high, or their powermeters are reading too low. As mentioned already, I am of the opinion that mine is reading to high because I have seen this repeated on many climbs.

Now I understand that in the big scheme of things that the output a powermeter shows is not important, as long as the output is consistently consistent over time so that one can see if their output is increasing as a result of one's training. However, why not have an accurate reading as well? Especially when it comes to FTP. I want to be able to compare my FTP with, say, Chris Froome's FTP. So if Chris Froome's SRM powermeter is highly accurate, which I assume it is given SKY's time and money investment in cycling and in Chris Froome, then I want my Stages to be accurate too. I dont want my powermeter telling me my FTP is comparable with Chris Froome's FTP because my powermeter is grossly overestimating my power output.

So to cut to the point, is anyone able to take a look at the segment above and estimate my FTP more accurately than my powermeter?

By the way, the road up to Doi Tung was in absoloutely magnificent condition. It was like West Head in Sydney top to bottom, as are nearly all the mountain roads in Thailand. and I have been on quite a few of them. For a supposedly third world country, Thailand's roads put ours to shame.
Last edited by Arlberg on Wed Oct 09, 2013 6:53 pm, edited 2 times in total.
Arlberg
 
Posts: 265
Joined: Mon Oct 03, 2011 5:43 pm

Re: Power meter output readings

Postby michaelten » Wed Oct 09, 2013 6:29 pm

You might be putting out more power and going slower than others if you weigh more than them.
michaelten
 
Posts: 157
Joined: Thu Mar 21, 2013 5:59 pm

Re: Power meter output readings

Postby toolonglegs » Wed Oct 09, 2013 6:31 pm

You are heavier than nearly everyone else that climbed that segment with a power meter ... you can't compare your numbers with someone 5,10, 15 kilos lighter than you.
The fastest guy on that segment is an absolute machine... ripped up Alpe D'huez last year in the Haute Route in a time most top end TDF riders would be proud of!.
Image
User avatar
toolonglegs
 
Posts: 14327
Joined: Fri Jun 29, 2007 7:49 pm
Location: Somewhere with padded walls and really big hills!

Re: Power meter output readings

Postby Arlberg » Wed Oct 09, 2013 6:42 pm

Yeah, Peter Pouly. He also won the Haute Route Alps last month for the 3rd time in a row, out of the three Haute Route Alps held so far. In fact up until this year he had never even lost a single Stage of the Haute route Alps.! (This year he eased up at the line at the end of the first Stage to let his training partner Alex Destribois Cowdroy take the victory).

His weight was 60.8kg. My weight is 82kg. And here is that particular segment you refer to on Alpe d' Huez. 38 minutes 35 seconds!

http://www.strava.com/activities/19354411#349235299
Last edited by Arlberg on Wed Oct 09, 2013 7:21 pm, edited 2 times in total.
Arlberg
 
Posts: 265
Joined: Mon Oct 03, 2011 5:43 pm

Re: Power meter output readings

Postby Arlberg » Wed Oct 09, 2013 6:50 pm

toolonglegs wrote:You are heavier than nearly everyone else that climbed that segment with a power meter ... you can't compare your numbers with someone 5,10, 15 kilos lighter than you.


But isn't that only relevant when comparing a riders watts/Kg output rather than just watts output? The actual power output is the actual power output, which is totally independent of the weight of the rider isnt it?
Arlberg
 
Posts: 265
Joined: Mon Oct 03, 2011 5:43 pm

Re: Power meter output readings

Postby toolonglegs » Wed Oct 09, 2013 6:52 pm

Arlberg wrote:
toolonglegs wrote:You are heavier than nearly everyone else that climbed that segment with a power meter ... you can't compare your numbers with someone 5,10, 15 kilos lighter than you.


But isn't that only relevant when comparing a riders watts/Kg output rather than just watts output? The actual power output is the actual power output, which is totally independent of the weight of the rider isnt it?

1 watt = 1 watt no matter who puts it out ... but a heavier rider needs more watts to move uphill at an equal pace to a lighter rider.
Image
User avatar
toolonglegs
 
Posts: 14327
Joined: Fri Jun 29, 2007 7:49 pm
Location: Somewhere with padded walls and really big hills!

Re: Power meter output readings

Postby Arlberg » Wed Oct 09, 2013 7:13 pm

My output on the climb was 3.85 watts/Kg. Peter Pouly's output on the same ride was 6.08 watts/Kg.... I still have a long way to go.
Arlberg
 
Posts: 265
Joined: Mon Oct 03, 2011 5:43 pm

Re: Power meter output readings

Postby toolonglegs » Wed Oct 09, 2013 7:17 pm

Don't we all :lol:
Image
User avatar
toolonglegs
 
Posts: 14327
Joined: Fri Jun 29, 2007 7:49 pm
Location: Somewhere with padded walls and really big hills!

Re: Power meter output readings

Postby Xplora » Wed Oct 09, 2013 8:47 pm

TLL is right - you will utterly HATE yourself if you don't compare apples with apples. Watts per kilo or GTFO. The same applies in reverse - I will NEVER, even under the influence of PCP and a strict EPO and steroid regimen, be able to generate the same sprint power as a guy like TLL who weighs 20kgs more than me and is a strong sprinter. If he pumps out 1000W over 10 seconds, I will be lucky to get 600.

You have to compare apples with apples otherwise it's a waste of time. Ultimately, it's about maximising your performance. If you want to win a pissing contest, just use avg speed or top speed. It's also meaningless in the broad scheme of things. You will only know if your training is adequate by checking race results or segment times.
Xplora
 
Posts: 5553
Joined: Fri Dec 10, 2010 11:33 pm
Location: TL;DR

Re: Power meter output readings

Postby toolonglegs » Wed Oct 09, 2013 9:14 pm

For the record my sprint power is pretty crap, I sometimes ( rarely ) get away with it because ( I think ) I have pretty good 1 minute power... and I am an aggressive bastard in a sprint, but then so is everyone where I live :| . But compared to doggatas (see the race thread) in a say man on man sprint I would get crushed :lol: .
Basically power to weight is what matters ... but even then with a lump of salt unless you were riding next to the person you are comparing yourself against. Comparing your power to others is interesting but the only thing that really matters is race results ( if you race of course! ) ... compare your numbers to your previous numbers and improve on them with the right training, you can't control other peoples outputs so in reality they should be ignored as you are only psyching yourself out!.
Getting aggressive ... http://www.youtube.com/watch?v=YoffxOXjG0I :mrgreen:
Image
User avatar
toolonglegs
 
Posts: 14327
Joined: Fri Jun 29, 2007 7:49 pm
Location: Somewhere with padded walls and really big hills!

Re: Power meter output readings

Postby Ross » Thu Oct 10, 2013 9:15 am

Cool video, TLL!
Image
User avatar
Ross
 
Posts: 3668
Joined: Sat Nov 07, 2009 7:53 pm

Re: Power meter output readings

Postby donncha » Fri Oct 11, 2013 6:44 am

Arlberg wrote:My Stages powermeter indicates an average power output of 316W for the 32 minute 23 second effort, which was effectively a 30 minute time trial. So using Alex's formula and taking 93% of that to convert it, means an FTP of 294, +/- 3%. But in my opinion this powermeter reads generally too high.


In my opinion, and I'm open to correction :) , you've done it all wrong and your FTP is a lot higher than 294!

A few issues:

1. You've taken a segment out of a 4hr ride and used that as your FTP test. That is not the same as doing a proper warm-up and then doing an actual effort with nothing left at the end like a proper FTP test, so the 316W is likely lower than you're really capable of.

2. Your effort was 32mins, whereas the 93% rule Alex talks about usually applies to a 20min effort.

3. Having just ridden 3 Gorges at the weekend, I looked up your times for your best efforts on Galston & Berowra. On Galston (short) you put 45s into me and on Berowra you put >1:30 into me. I'm 85kg with an FTP of 325W, climbed Galston at 377W (4.4W/kg) and Berowra at 346W (4W/kg - didn't go hard right from the ferry) so you're already capable of performances well beyond those W/kg figures. I don't know how much 3kg is worth on Berowra, but I suspect it's a lot less than 1:30, so you'd probably beat my W figures as well.

Personally, if I had your figures, I'd start my FTP at 310-315 and only downgrade if I couldn't handle the training. Perhaps try a traditional 2x20 session at that level and see how you go? e.g:

warm-up
20mins at 90% of 310 (270W)
5-10 mins easy
20mins at 90% of 310 (270W)
warm down.

Should be a decent enough session, but not so hard that you couldn't do it again tomorrow.
Image
donncha
 
Posts: 95
Joined: Wed Mar 31, 2010 8:41 am
Location: Maroubra

Re: Power meter output readings

Postby Arlberg » Fri Oct 11, 2013 7:20 pm

Donncha, that is one of the most helpful posts I think I have ever had. Thanks so much for the time you put into it. You are spot on regarding points 1 and 2, I overlooked both those facts.

Another thing I overlooked, thanks to advice from michael10 and TLL was not taking rider weight into consideration. So maybe this powermeter is more accurate than I thought after all.

There are penty of flat road sections around here (in Chiang Rai) where I could do the 2x20 minute 'traditional' sessions. I understand the bit about the first 20 minute session (the warm up) being done at 90% (of 310) but why do you suggest the 2nd 20 minute session is also only done at 90%? Isn't the FTP test supposed to be 100% all out effort with nothing left in the tank at the end?

I will be returning to climb Doi Tung in a couple of days and will give it another crack there too. I will have to ride out there so I wont be 100% fresh for the climb, but I can ride to the start more gently next time.
Last edited by Arlberg on Fri Oct 11, 2013 7:53 pm, edited 1 time in total.
Arlberg
 
Posts: 265
Joined: Mon Oct 03, 2011 5:43 pm

Re: Power meter output readings

Postby Xplora » Fri Oct 11, 2013 7:37 pm

If you haven't read Training With Power by Hunter and Coggen, it's time. Joe Friel's Power book is good too as long as you've read the Training Bible already. All answers are there.
Xplora
 
Posts: 5553
Joined: Fri Dec 10, 2010 11:33 pm
Location: TL;DR

Re: Power meter output readings

Postby donncha » Fri Oct 11, 2013 9:57 pm

Arlberg wrote:There are penty of flat road sections around here (in Chiang Rai) where I could do the 2x20 minute 'traditional' sessions. I understand the bit about the first 20 minute session (the warm up) being done at 90% (of 310) but why do you suggest the 2nd 20 minute session is also only done at 90%? Isn't the FTP test supposed to be 100% all out effort with nothing left in the tank at the end?


The 2x20 is not an FTP test. It's a fairly common session for those who want to improve their FTP. My reason for suggesting it was that if you can do 2x20min (AT) 90% of 310W and it doesn't take too much out of you (i.e: you feel like you could do it again the next day) then your FTP is probably around 310W.

If you are going to do a 20min FTP Test, then the usual protocol is

warm-up
5mins all-out effort
10-15 mins easy
20mins all-out effort
warm-down

Then take 93% of your average for the hard 20mins. Note that that 20mins has to be all-out. You should almost need a lie-down for a while after it if you do it properly. As Xplora mentioned, now that you have the power meter, buy and read a copy of Training and Racing with a Power Meter.
Image
donncha
 
Posts: 95
Joined: Wed Mar 31, 2010 8:41 am
Location: Maroubra


Re: Power meter output readings

Postby Xplora » Sat Oct 12, 2013 10:05 am

Just some unrequested commentary... if you have the time and money to go out and buy a power meter, you REALLY need to do some serious learnings to understand how to use it - it's just too freakin' expensive to not understand the basics like TSS or FTP or etc etc etc and it's utterly useless to you without those details. The objective to provide an objective baseline for your riding, better than speed or HR or perceived effort; unfortunately that info isn't really much good unless you are capable of using it to change your riding. You might as well just use speed which is a lot cheaper to train!

It's a bit like knowing the answer to the meaning of life - if you don't actually know what to do with it, are you any better off than the person who doesn't know?
Xplora
 
Posts: 5553
Joined: Fri Dec 10, 2010 11:33 pm
Location: TL;DR

Previous

Return to Training

Who is online

Users browsing this forum: No registered users



Popular Bike Shops
Torpedo 7 Torpedo7 AU
Ground Effect Ground Effect NZ
Chain Reaction Cycles CRC UK
Wiggle Wiggle UK
Ebay Ebay AU



InTouch with BNA
“Bicycles BNA Twitter
“Bicycles BNA Facebook
“Google+ BNA Google+
“Bicycles BNA Newsletter