Critical Power

tim cronk

With all the discussion over the various testing methods for FTP. 

I am wondering does anybody use Critical Power calculations?

For the last couple of years I have put my data into Golden Cheetah and the CP numbers came out similar to my FTP usually within a couple watts.  After recently reading  Dr. Phil Skiba's books I went back to Golden Cheetah to the calculator.  You enter 2 numbers into the CP calculator ,  your best 3-5 min effort, and your best 15-60 min effort.  These are your best recent efforts and not necessarily on the same day or workout.

My current FTP test using the new EN testing protocol nets me an FTP of 222 and those numbers punched into CP calc give me a CP of 227.

Interested to hear others thoughs, experiences, and comparisons.

Rumpole

Lots of ways to skin the cat--see also this: http://alex-cycle.blogspot.com/2008/05/the-seven-deadly-sins.html



Bottom line is that you're paying to do things the "EN way"--but so long as you know your FTP, you know your ftp.

 

EDIT--sorry-should have seen the parade of stars.  I've also found that the CP method seems to be pretty close (definitely close enough for pacing interval sets appropriately).

Peter Greagg

Tim — when I use the 5 10 20 testing technique, I calculate CP using the explanation on pages 44 to 46 in Training and Racing with a Power Meter (2nd edition).

My CP is usually within a few watts of my FTP.

William Jenks

Tim- Short answer is that I keep an eye on CP using Golden Cheetah...as you say, FTP and CP are usually quite close, and the CP is dynamically updated with every ride if it's going up. Gives me an idea what to "try for" on the next test, if nothing else. I've also gone through a couple spells where I just hated the idea of testing again, and knew that if I was pushing the CP curve most of the time, I was on.

tim cronk

@Chris---No worries. Great link. I saw that before but was awesome to review. I hope others have read that lots of good stuff.

@Peter... Math ?no thanks I need the CP calculator to do it for me.

@William---Do you ever use the CP calculator in GC or do you just download your rides? Do you know how to reset the CP on GC ? Mine is stuck on the last highest CP and that is why I am using the Calculator now for some comparisons.

FWIW .... my belief.. FTP is 60 min threshold period and everything else is just an estimate , .... a 42 min test is just that, its not 60min....I think realistically a 42min test is closer to 95%FTP and a 20min test is closer to 90% FTP.... Having said that I have tested 60min , 42 min as 2x20(2), and 95% of 20 min all out, and they have all come pretty close to one another... So yes its semantics and they all work... Close enough??? For the most part yes at my baby FTP I need to be with 5watts or its too big a move , those with higher ftp can certainly vary 10 watts or maybe more. One of these days I will try a MAP test as mentioned in that link to Alex's Blog.

Back to my original post about CP. About the only thing that really stands out using CP is that is different is that you need 2 tests, But they can vary in length as long as one is short and the other is long, and they do not need to be done in the same session. This to me seems very repeatable. I know lots of people like that sound of that , specially if its accurate.

MattAaronson

Take the 227 Tim...it'll push you harder on all those FTP workouts!!

To me the repeatability is the key thing and "standalone accuracy" of the test only matters as it relates to setting your zones for racing. But I think by the time our races come up we've done so much z3 "ABP" riding that we intuitively KNOW the right target watts and +/- 3 watts of FTP isn't going to make or break.

William Jenks

 Tim-  

 

We MAY actually talking about two slightly different functions.  Because CP is a mathematical fit to your "best power" for different lengths of time, you can do it either as a two point fit (two points always define a line....) or you can use all your rides to do the fit.  This is as opposed to just doing the two tests and feeding it to the menu calculator.

 

If you load your rides in to GC and then click on Critical Power, it shows you the familiar TP/WKO type graph of your max power over all the rides for every time interval, plus the same thing for the current ride as a dotted line.  It will automatically calculate your CP from all the data from about 1.5 min to about 3 hours (or however long your longest ride is).  You can look at the linear fit by clicking on the "Y Axis shows Energy" option instead of "Y Axis shows Power".

 

One thing I really like about this graphic is that it naturally suggests Powers that should be achievable for different amounts of time based on my other data.  For example, it naturally shows how hard you should be able to go for 10 min if you've been doing short vo2 intervals and longer 20+ minute intervals.

 

Now...as far as it being stuck on the last highest CP, what you have to do is define the date interval, which is admittedly not obvious on how to do.  I am assuming you are using version 2.1.  I hear that version 3.0 should be available soon...and that will have the advantage of having NP/TSS/etc

 

• You go to the Metrics page and then (Mac version) shift-click-drag in the data window to define some arbitrary date range.  It doesn't matter what date range this is.  
• That will create a new custom date range in the right hand side window.  Ctrl-click the date range to edit it.  Rename it and put whatever date range you want.
• Quit and restart GC.
• Now when you go to the critical power window, your new custom date range will be available for you to use! 

It's obviously true that this method only shows you the highest CP during the date range specified.  However, if you think it's falling, it's easy to go in and re-edit the date range to check data for the last month or whatever.

Hope that make sense!  (and sorry for the slow reply)

 

tim cronk

William Great info. Big help. Thanks!

tim cronk

@William.... thank you again...going thru that I never would have figured it out... Hopefully there will be alot of improvemnts in the next GC version...


Follow up.... tested today 5 min effort 262 , 20 min effort 239 , 95% of 239 is FTP 227 ..... the 5 and 20 numbers plugged into GC , CP calc come out with CP of 231....

Now I am fairly confident that had I done the 5 min effort and 20 min effort on different days I would have higher numbers and then a higher CP!

Pretty convinced that CP calculator is good (specially if taken off the same ride) , but FTP tests are probably more accurate estimate . You may say that there is only 4 watts difference but when I'm on my 100% ftp interval + or - 4 watts makes a huge difference IMO.... maybe not so much for the 300 ftp guys but down here it does!

Darren Freeman

Thanks for the thread.. I did my FTP test today on my CT and wasn't sure if I should subtract the 5% as Coach Rich stated in his video on the test.  After reading your thread I'm pretty sure my 20m FTP score can stand as without the 5% adjustment.  My unadjusted CP scores today were 242 @ 5m and 191 @ 20m.  My original CP from early November was 178 @ a 1hr effort before I joined EN.  

I"m pretty happy with that and I've just begun my training with EN.  My near term goal is to get my FTP over 200. 

rkling

@Darren: Reiterating what Tim said, FTP is what your expected power would be for an all out 60 min TT. Because most of us can't focus on a trainer for 60 minutes at threshold pace without getting buggy, the original test of 2 x 20' (2') was a good estimation, because, while the 20' intervals would overestimate your FTP, adding in the 2' recovery brings the NP down to a very close estimation of FTP. Since a single 20' test would overestimate the FTP even more, that's why the 95% works.

@Tim: my understanding of why you do both the VO2 and the FTP test in the same session is to decrease the likelihood of overestimating the FTP with a single 20' session on fresh legs. Doing the 5' VO2 test first fatigues the legs enough that it better replicates the FTP. I don't use CP, so sorry if I hijacked the thread a bit.

For my purposes, FTP is a good number to know, but I'd never race an Olympic distance race at that number. While it (an Olympic distance race) is the closest estimation of the 60 min TT, the problem is that the idea with the FTP is that it's for a TT, not a triathlon, and I wouldn't run as well off the bike if I rode at FTP for 60 minutes.

tim cronk

@Darren... Keep in mind that CP and FTP are 2 different things , calculated differently , ultimately getting similar numbers.... If I understand your post correctly your 5 min interval was 242 and your 20 min interval was 191? EN - FTP calculation protocol 191 x .95 = FTP estimate of 181.45.... If I plug 252/191 into the GC CP calculator it spits out 174 critical power..... Go with 181 ;-)....also if you havent already done so read the link Chris posted to that blog .. good stuff in there as well

@ Richard.. Boy do I ever understand the theory of the 5min interval taking the wind out of my sails for the 20 min interval LOL... but that was really part of my discussion on the difference of FTP vs CP .... using CP you are not required to do both tests on the same day... Again it is my belief that whichever system you are using or correlating to, the closer you are to 1hr the more correct it will be...If its not 1hr ...Its an estimate..

Darren Freeman

Thanks for the clarification Tim.. I suspected my thinking was off and found the GC CP calculator after my post and got the same number. My power was spread over a 20 watt range for each part of the test which likely caused my FTP to go to down 174. My 1 hour power test I did in early November before I joined EN came out to 178 which is nearly the same as 181 so I'll take it..

William Jenks

This is a nice article by Phil Skiba, one of the advocates of CP and the guy behind Apollo RaceDay and xPower and the related measures that are similar to NP/TSS/etc but use different names.

There is a "theory" behind CP, but it turns out mathematically to be close enough to the "CP60" view of things (i.e., FTP) that the two numbers end up being almost interchangeable, even though they are not arrived at in the same way.

http://physfarm.com/new/?page_id=511