20 min FTP test route?

Alex_Simmons/RST

jibberjim wrote:

amaferanga wrote:

My understanding of the CP algorithm is that if you haven't done a truly maximal effort of a few minutes as well as a longer maximal effort then your CP will be overestimated.

The AWC would be understated yes, but remember CP is not equivalent to FTP, as on top of the CP there's also the AWC spread out over the 60minutes such that FTP must be larger than CP.

But of course CP is just a model, it may not be accurate.

When you use good data*, it is a pretty good model for durations within the range of a few minutes to an hour.

It is my experience however that CP tend to be a better estimate of FTP than the predicted 60-min power.


* data needs to be from (at least) two maximal tests of suitable durations within about a week or of each other. Cherry picking data from rides and using data from efforts that are months apart is a "sin of sins".

doyler78

Alex_Simmons/RST wrote:

When you use good data*, it is a pretty good model for durations within the range of a few minutes to an hour.

It is my experience however that CP tend to be a better estimate of FTP than the predicted 60-min power.


* data needs to be from (at least) two maximal tests of suitable durations within about a week or of each other. Cherry picking data from rides and using data from efforts that are months apart is a "sin of sins".

Is that for those using the standard 95% of 20 min power or is that even for those that have done the 1 hour tests as well and therefore have a good idea of what percentage of FTP their 20 min power lies at?

Alex_Simmons/RST

doyler78 wrote:

Alex_Simmons/RST wrote:

When you use good data*, it is a pretty good model for durations within the range of a few minutes to an hour.

It is my experience however that CP tend to be a better estimate of FTP than the predicted 60-min power.


* data needs to be from (at least) two maximal tests of suitable durations within about a week or of each other. Cherry picking data from rides and using data from efforts that are months apart is a "sin of sins".

Is that for those using the standard 95% of 20 min power or is that even for those that have done the 1 hour tests as well and therefore have a good idea of what percentage of FTP their 20 min power lies at?

It is also my experience that 95% of 20-min mean maximal power over estimates FTP for the majority.

doyler78

Alex I was really asking about those that have used more reliable tests to confirm their FTP but who use 20 mins efforts as well so know pretty well where their % of 20 min lies in relation to FTP. I was wondering whether you had seen much variabillity here too?

I personally have found that I can pretty much now rely on 20 min tests (though I always confirm anyway but confirm is all that they ever do) to set new FTP figures. (I should say at this point that always use NP rather than perfectly paced efforts as I just don't have terrain to permit it or at least I can't pace myself on my terrain so given NP range of error can be out by 5% I could just be finding myself in the margin of error but then again to do that consistently seems unrealistic to me).

jibberjim

Alex_Simmons/RST wrote:

[It is my experience however that CP tend to be a better estimate of FTP than the predicted 60-min power.

Interestingly that for me following what you say there gives me the value I generally do believe is my FTP (in november when I did some good tests it gave me 323 CP and my AP from 50-60 minute rides was taking me towards an FTP of 320.

However I've always had the problem that I can produce NP efforts in the ~340 range without too much difficulty on a few courses / races so by some measures I could call that my FTP. But I don't as no matter how motivated I can't break 320 for the hour. Or even steady state %ages on shorter durations that might make me think 340 for an hour is possible (I only do 345 on a 10mile TT for example)