Training Analysis

by old_man_biking

Downloads: 2,939 (12) • Reviews: 3

Version:

3.0.5857

Updated:

Jan 14, 2016

Normalized Power lower than Average Power?

Normalized Power lower than Average Power?

Postby rikicarra » Sat Feb 26, 2011 3:15 pm

Did a circuit race today, spent some time in the peloton (sometime drafting and not pedaling). Analysing the activity with Training Analysis I found that the Normalized Power is 281 W, lower than the Average Power (287 W). Is that due to the fact that Average Power is discarding periods at 0 power and the Normalized Power is not? Comparing apple to apple, I know that Normalized Power should always be higher than Average Power (particularly for an activity like today, a lot of power variability :P ).
rikicarra
 
Posts: 73
Joined: Sun May 03, 2009 8:05 am

Re: Normalized Power lower than Average Power?

Postby PissedOffCil » Sat Feb 26, 2011 3:27 pm

I see no other explanation. NP is, by definition, always higher than AP.
Developer of the Garmin Fitness, Filtered Statistics and Distribution plugins
PissedOffCil
Donated!
Donated!
 
Posts: 1822
Joined: Sat Jul 05, 2008 9:18 pm
Location: Chicoutimi, Qc, Canada

Re: Normalized Power lower than Average Power?

Postby old_man_biking » Sat Feb 26, 2011 4:26 pm

The plugin displays two different average powers: one - like ST calculates - ignoring times with 0 power value and the other one including 0s into the calculation.
I always felt that the ST way of calculating power discarding 0s is a bit strange. Performing 10 minutes at 500W and coasting 10 minutes at 0W will give an average of 500W.
That's why I introduces the "average power including 0s". And 've based NP on this.
Does this explain the data that you see?

Cheers,
OMB

PS: Of course NP must be highter than avg power, as it's calculated as (Avg(X^4))^1/4.
old_man_biking
Donated!
Donated!
 
Posts: 6766
Joined: Mon May 28, 2007 8:07 am
Location: Munich, Germany

Re: Normalized Power lower than Average Power?

Postby rikicarra » Sat Feb 26, 2011 6:59 pm

old_man_biking wrote:The plugin displays two different average powers: one - like ST calculates - ignoring times with 0 power value and the other one including 0s into the calculation.
I always felt that the ST way of calculating power discarding 0s is a bit strange. Performing 10 minutes at 500W and coasting 10 minutes at 0W will give an average of 500W.
That's why I introduces the "average power including 0s". And 've based NP on this.
Does this explain the data that you see?

Cheers,
OMB

PS: Of course NP must be highter than avg power, as it's calculated as (Avg(X^4))^1/4.

You are 100% right. I simply didn't see the "average power including 0s" field. Now I checked, it correctly indicates 230 W. I was mislead also by the fact that my Edge 800 at the end of the race gave me an average power of 287 W, I have now checked the setup of the Edge and infact it ignores the 0 power times. I should probably change that setting, as it is misleading.
In case you would include the Variability Index in your plugin (pag. 120-121 of Allen&Coggan book) it should be the ratio between NP and the "average power including 0s". In my case today it would have been VI=281/230=1.22

Thank you.
Riccardo
rikicarra
 
Posts: 73
Joined: Sun May 03, 2009 8:05 am

Re: Normalized Power lower than Average Power?

Postby FranR » Mon Sep 15, 2014 3:53 am

Hi OMB,

I am having a hard time understanding my numbers in TA. I got the following in a recent ride:

1- AP (excluding 0´s)= 304W
2- NP=276W
3- VI=0,91 (less than 1!)
4- AP (including 0's)= 152W

According to previous posts, NP should be calculated based in AP including zeros, and then VI should be the ratio between these two values. But my VI seems to be 2/1 instead of 2/4, giving me a value below 1... This consistently happens in all my rides. Obviously I am missing something here.

If it helps, I am currently using GPS2Power plugin until my new groupset and Stages PM are delivered...
FranR
 
Posts: 62
Joined: Wed Nov 07, 2012 3:21 pm

Re: Normalized Power lower than Average Power?

Postby texmurphy » Mon Sep 15, 2014 8:37 am

FranR wrote:...If it helps, I am currently using GPS2Power plugin...

Do you have Power=0 when Cadence=0?
texmurphy
Donated!
Donated!
 
Posts: 2185
Joined: Wed Jul 05, 2006 7:38 pm
Location: Maryland, USA

Re: Normalized Power lower than Average Power?

Postby FranR » Sun Sep 21, 2014 2:12 pm

Sorry for may late reply, had a very busy week...

I only can select this opton once I have created the power file, in the same menu where the "calculate power track" and "clear power track" are... No way for me to make sure this option is enabled.

Anyway when I hit this, I see my average power slightly change, but the same issue still applies (VI still in 0'9s)
FranR
 
Posts: 62
Joined: Wed Nov 07, 2012 3:21 pm

Re: Normalized Power lower than Average Power?

Postby old_man_biking » Mon Sep 22, 2014 1:35 am

Hello Fran,

could you please export this activity to .FITLOG format, zip this file and attach it to your post.
Please add a screen shot of the "Performance Indices" tab of the "Training Analysis" page.
And please what is your setting for Settings:Display:Analysis:Power cutoff .

Cheers,
OMB
old_man_biking
Donated!
Donated!
 
Posts: 6766
Joined: Mon May 28, 2007 8:07 am
Location: Munich, Germany

Re: Normalized Power lower than Average Power?

Postby FranR » Mon Sep 22, 2014 8:15 am

Thanks OMB,

I finally found the problem, right there in my ST disply cyclist settings, I had my watts setup with an older FTP value. Changed it to 0, and use a cadence cutoff of 20, and now VI values make sense. Interestingly, average power values (with and without zeros, as provided by TA plugin) are within 1% difference now.

As a newbie on a bike, I am still trying to fully understand these concepts. I have done my homework, read my Friel and Coggan, but I still have to sink that in and figure out many things... More rides and more accurate FTP figures should help for sure!

For example, got a handful of rides already, and Normalized power ranges from 10% to more than 30% more than average power, well beyond my FTP estimation... I can see no trends in course profile or perceived effort in these data...but surely there has to be an explanation. :roll:

thanks for your support!
FranR
 
Posts: 62
Joined: Wed Nov 07, 2012 3:21 pm

Re: Normalized Power lower than Average Power?

Postby Burnt Toast » Tue Sep 23, 2014 8:39 am

Make sure your pm is calibrated if you have one.

With GPS2Power, the weather info that is imported can make a huge difference in the power numbers. If the winds are high, but you are riding among lots of trees or buildings, the wind that is actually affecting your cycling could be much different then that of the weather station.
Burnt Toast
Donated!
Donated!
 
Posts: 666
Joined: Thu Apr 29, 2010 9:21 am
Location: Minnesota, USA


Return to Training Analysis

Who is online

Users browsing this forum: No registered users and 2 guests