Or CP vs W’ if you want to be a stickler. I did several tests of varying lengths to hone in on my CPs and FTP. My best were from 3 dates;
8/5/2025
Seconds Watts
A 233 364
B 992 316
8/7/2025
C 1217 314
8/12/2025
D 653 330
Using HighNorth gives
CP/FTP W Prime
A+C+D 311/299 12.3
B+C+D 294/282 23.5
A+B+C 301/289 14.5
A+B+D 305/293 13.9
CP and Wprime are physiological phenomena (albeit not completely understood) which we try to estimate using mathematical calculations and what you’re describing is ‘normal behaviour’ of the critical power model. Generally speaking, training that increases your critical power lowers your Wprime and vice versa. With more than 2 tests there will always be an error of measurement, which you can calculate.
As with all tests, the results are protocol dependent. If you’re relatively in-tune with your body, I would use the tests and results that most closely correlate with what you’re feeling.
I’m 61 and back to cycling for almost 2 years after a several year lay off. I MTB raced from the early 90s up to middle 20teens. I had a good stint racing single speed MTB in my mid 40s. Early I used HR at did trainer ramp tests to estimate LTHR and later had power taps on all my bikes. I haven’t been doing any formal FTP or interval training, except for the tests. Just a variety of riding. Some very hard single speed and some hard geared MTB rides. Endurance riding on the Canal system in Mesa AZ and some road rides with hills. I was surprised at my FTP (at my peak it was 330 Watts) and judging from what I read my W Prime seems higher than a regular guy training. I am pretty much gassed when that 21.7kj runs out and not good for more efforts even after it has been theoretically replenished around 10 minutes later.
First off, it’s just a model. The model is sensitive to the time durations and if you really went max. Your results are not surprising.
HighNorth recommends
3-min
5-min
12-min
with a fourth effort around 20-min to help calibrate the output of the 3/5/12-min CP estimate.
A was 3.9 minutes
B was 16.5 minutes
C was 20.3 minutes
D was 10.9 minutes
Essentially, B, C, and D are long efforts. So you have a shorter effort and 3 longer efforts.
Beyond 2 minutes I find it can be difficult to truly go all-out and maximize power for a duration. For example going out a little too hard on a 3+ minute max effort will reduce the power for the 3 minutes.
In other words, there is some skill involved in testing.
I wouldn’t overthink it. Pick an estimates and adjust based on your training.
Not that I haven’t been accused of over thinking it. And I kind was set at the 299. Seemed reasonable. Then on 08/26/2025 I did 3 SST intervals between 19 and 28 minutes and felt toasted the next day. I now think that could have been fueling and or the heat.
What got me started, was playing around with W Prime, over the last month or so.
The estimate showed several kj low compared to the Drop I got from the Intervals.icu chart. I adjusted it and went about my business. Repeating the drop a couple times. But after those SST intervals I was concerned that I over estimated my FTP. I will alter my fueling and it will not be so hot here soon.
And yes it’s just a model, but it was very predictive of the Watts that I could pull from durations that I had not done yet, based on the few more moderate attempts leading up to these final ones.
Every system has there thoughts and one thought 3,5 and 12 were too short.
yes, when set accurately, it does a great job of prediction. However I’ve also been able to exceed it when I thought it was set accurately. Interesting for post-ride analysis and tracking changes in fitness, not good for displaying on a computer and trying to ride by it.
Short version; Is Intervals over estimating my W Prime drop?
Long version;
So I have been reading up on Xert and started to question what I previously read and seems to be commonly believed, that CP was higher than FTP. If CP is indefinite (Theoretically of course) and FTP is an hour then, Yada yada yada. Then came across the post below.
I know it is over a year ago and your views may have evolved since then, but I found
This part particularly eye opening “CP model overestimates the tipping point between stable and unstable” Leading me to consider the longer test may be closer to my to reality.
Leading me to try to clear up my original question.
If I drop my FTP to 282 the W Prime cranks to 34kj in Intervals.icu, but only 23.5kl in high north (which is much closer to my observations)
I’ve compared my self-assessed FTP to GC’s Extended CP and they are within a watt or two of each other. As far as I’m concerned, FTP and CP are essentially the same if you properly feed the CP model. Otherwise if you only use something like 3 minutes, 6 minutes, and 12-14 minutes, in my experience, the CP model overestimates the tipping point between stable and unstable."
That’s the major problem with the W’ models. They are using a large population average for replenishment functions. And that can be:
Totally off in all conditions
Be quite reproducible if you deplete it in one go
Be quite reproducible if you deplete it in a strict protocol of hard efforts
But, for myself, I have never seen a number from a W’ test being replicated even moderately close during an outdoor freeride. But the exact same interval session on an indoor bike, gives me acceptably close results.
I’m convinced that the time constant and amplitude for the replenishment is the reason for this.
Have you ever tried to use both models? There’s an integral and a differential model. The absolute number for W’ is quite different but for some people, one gives better results then the other when comparing depletion over different ride conditions.
This topic was discussed on the Time Crunched Cyclist podcast, obviously with a bias to TrainingPeaks/WKO5. They do refer to the two linked concepts of FRC/FTP and W’/CP.
On the bike I use a garmin 540 with a widget. I used the recommended one. I will have to look. But it matched Intervals almost exactly and I can repeat it on 10-20 minintervasl at the max Watts for me on those intervals. So the tank seems right if I keep my FTP at 299 Intervals.icu give me about a 21.7 kj W Prime drop. But it and High north estimate My W Prime to be much lower. But if I drop my FTP to 284 say Intervals.icu shoots up to 34kj and HighNorth goes to 23.5 on that same interval. I seriously doubt I have a 34kj tank! and once I drain to zero once I am done for higher zone work.
Thanks, that pretty much confirmed what I thought, FTP, FRC, CP and W’ are. Was there a specific take home that you wanted me to, well take home?
I did find it interesting that it was 50/50 on atheletes with FTP higher than CP and vice versa. HIGHNORTH and Interval.icu both give me and FTP of 299 Watts using 4, 11 and 20 minutes on HIGHNORTH and Intervals.icu using a 20 minute effort from August 7th, with a 21.3kj drop during the interval. HN estimated 12.3 kj
I use the differential option on the Garmin W Prime widget. Apparently that is what Intervals uses and I don’t know if you can change it. I started with this after reading.
