Using Power Curve to Identify Strengths & Weaknesses

Sure, but between using a percentage of <5 min power and a long all out effort to prescribe longer intervals I’ll use a longer effort. Friel and Coggen routinely say you can use approximately one hour races to estimate ftp, and the OP is struggling with intervals at a higher ftp. You’ll get 90+% of the physiological adaptions with intervals at this lower power since the energy systems being worked are the same

1 Like

I don’t think longer intervals are necessarily better, because (1) the element of pacing becomes very important and (2) you need to have a good ballpark estimate for your FTP. And (3) you still have the same issue that you have with the MAP test: you need to convert your 20-minute power to your FTP. Coggan initially knocked off 5 %, but for the general public that seems not enough (on average), so nowadays the recommendation is 15 % — unless you know what you are doing. Perhaps the statistical distribution is a tad narrower, but you still have a range of (at least) 10 %, which is a sizable margin of uncertainty, big enough to be significant IMHO.

Now if you are experienced, none of this is likely an issue, but then most could also use a MAP test to gauge your FTP (corrected by hand with experience to account for your personal relation between MAP and FTP).

That’s not universally correct, the easiest example I can think of are over/unders. They only target the correct energy and lactate systems if the over intervals are really over and the under intervals are really below FTP. Here, I think you really need to be in the right ballpark within 3-5 % max. Otherwise 105 % FTP-as-set (a common value for over intervals) might be <100 % of your actual FTP. Or you need to go for 110 % FTP-as-set, which could make these over/unders prohibitively hard.

The better solution is to pick your favorite FTP test, and, importantly, to not be afraid to correct it downwards. I find the ramp test better, because it is an all-out effort. That is, done correctly, you’ll be hugging your bars after it is over and panting. Yes, fixing your FTP as 75 % of your MAP is problematic, but just use common sense and double check. If necessary, tweak it. IMHO one psychological issue is that we (I’m including myself here) are much more reluctant to correct our FTP downwards rather than upwards. Once you get experienced, I find you can guess your FTP-as-tested within a few watts. That is, you develop a feel for how strong you are and how hard a particular wattage feels, which are obviously extremely valuable skills.

Like you wrote, you can verify and should whether the FTP you have set is reasonable by doing a few choice workouts or do outdoor rides. I usually like to use a threshold and a VO2max workout. On outdoor rides a good rule of thumb is that your 20-minute power on an outdoor ride is larger than your FTP, your FTP is set too low. That is assuming that you have e. g. an uninterrupted climb that is at least 20 minutes long. Or you can also use your 5-minute power PR, that should be roughly 125–135 % of your FTP (unless you really just do an all-out KOM attack-style effort).

You are looking at the wrong chart. The chart you should check is Optimized Intervals, and it gives you specific duration.



1 Like

Well. I think this is an easy assumption, but I respectfully disagree as I monitor heart rate decoupling closely and currently am in the 120-180 min range before seeing any sort of decoupling during a Z2 ride. My long rides are 3-4 hours.

I realize this isn’t the only metric for determining aerobic efficiency. A couple recent strict Z2 rides:

At the risk of further derailment, I have been doing all of my VO2max sessions (and current FTP sessions) based on a percentage of FTP (330w). So, it’s odd that I can execute a 5x5 session at 120% of FTP (which is no doubt challenging), but am finding 3x15 @ 97% abnormally difficult. RPE feels more akin to 8-min @ 105%.

I’m chalking it up to accumulated fatigue and likely poor rest/recovery. I’ve been experimenting with non-traditional rest weeks, opting for more frequent 2-3 days of rest, I wonder if it’s not working as effectively as I anticipated.


Doh. Thanks - I guess I never kept a screen shot of that one (but I did find some notes).

Do you find the recommendations change much as you add new PRs to the curve?


No need to apologize, I’m not debating what my FTP is and that wasn’t even the point of this thread, it just ended up going there.

My point was to see if anyone who has experience analyzing PD curve(s) could give some insight into what they use it for and how they use to shape training.

Again, not looking to discuss what my FTP is. FWIW, my last race effort (4 weeks ago) was 2 hour 40 min with an avg power of 293w and a normalized power of 322w. I think my FTP is well within the 310-330w range.

Depends on tte and what part of pdc you have improved. I basically do not use this chart that much for my workouts. For under FTP it is all the same for me, over ftp is always all-out in vo2 max. I do not train short power at all.

I’ve always used it to a) judge the results of training, and b) set interval targets. Can we debate your FTP instead? :rofl:



@anthonylane why isn’t that race in the 42 day curve you originally posted?

I was filtering from May 30th to July 12th, and didn’t include that race result.

Here’s my 45 day curve (race was May 29th, 45 days ago):

1 Like

I rarely use Intervals, thought it was possible to use custom dates so you could do something like 90 days.

I imagine it’s important to do some “testing” and perform a range of efforts (say 5sec, 1min, 5min, 20min) across your curve to establish a baseline? Otherwise it’s just modeling off your best efforts which could be A) not your best efforts B) old.

Can we debate your FTP instead? :rofl:

By all means. :grinning: I reckon it’s not below 300w, and certainly not above 400w.



You can, here’s 90 days. The thing it lacks is showing how you stack up amongst your peer group like the WKO chart.

I haven’t done any sort of testing since 14 April 2021 when I did some baseline measuring before starting a VO2 block. That consisted of max efforts of 5s, 30s, 60s, 4m.

It might be time to do this again in order to get clean and relevant data.

This bit stood out to me. I do understand that your 8min power in your power curve is lower than the modelled 8min power. However I don’t think that means you need to work on 8min power progressively, I think all you need is an all-out 8min test. The repeat intervalls sessions you have planned won’t give you that, because you neccessarily need the effort to be repeatable. (Obviously, if you are still below where you want to be, and 8min is a meaningful duration for your events, work on it progressively as planned.)

I think in general it is a good idea to do all-out efforts of varying duration now and then, to be able to see your actual max ability power curve. You could pick one or two points per week on the power curve where there’s deviation from the model, and do a test on those. Over time, different durations will drop out of the time frame (42 or 90 days, or whatever you set), so you get to do max efforts over the whole range. (I believe wko has a handy “residuals” chart that shows variation from the model more clearly, but you can do it with the intervals chart too).

1 Like

Hey @splash, thanks. Re: 8-min. This was just something I wanted to focus on in my second block of FTP work, with the other interval workout of the week being steady-state threshold w/bursts. I thought 8-min intervals could be fun, seeing as I’m prepping for CX and laps are around 6-8 min in duration.

The point of this block is to raise FTP from the bottom coming out of a really successful VO2 block.

FTP @ 320w (which I think is about right) is around 77% of my best 5-min power (413w or 129% of FTP). Which I think is too high, which makes me wonder if my FTP is truly closer to the 330w mark, and the recent struggles at 97% can be chalked up to just plain fatigue. Or something else? High anaerobic contribution?

FWIW, during my VO2 block in May I was targeting 390-400w for 4-5 min, which @ 320w would be 122-125%. Historically (4 years of structured training), working at percentages around 118-120% of FTP was my absolute max, and for periods no longer than 3 min.

This is quite interesting, we know that TR uses 70% of the highest 1-min power on a ramp test to calculate FTP (accuracy notwithstanding), but what if we use 5-min as a proxy. We can take 85% and for me that would put my FTP @ ~350w. Which I know isn’t possible. Interesting and possibly useless nonetheless.

Good call, this is something I should do probably bi-weekly…or more often. They usually happen organically when I’m going after a Strava segment.


this is the best way to populate your power curve without making it a chore

1 Like