@james_a_c sounds good. Go ahead and use it.
All I’m saying is that both your premises are provably false in my experience. So I use something else. It’s ok with me if you want to set up your training using critical power testing protocols.
Somebody once told me that critical power proponents were like the elite circle of people who had enough experience to truly understand human exercise physiology. 
I guess I’m not there yet!
The 3-min Critical Power test gives you two things:
So there are some things to like about it, primarily that it gives you anaerobic capacity and a target for sub 30min efforts.
Agreed - this shows that a popularized FTP test protocol is not that good. Just like there are different MLSS testing protocols that will yield different results, there are different FTP test protocols that will yield different results.
My take is that FTP == power at MLSS, and that this was/is the intended definition. (If this is not the intended definition, then it is divorced from underlying physiology, and much, much less interesting.) Due to varying TTEs, I do not think that it is possible to have a simple test definition that “is” FTP. What I think we are all looking for is an easy enough, accurate enough, repeatable enough test protocol to make repeated testing and training adjustments regularly possible.
The bit you mention about doing 30 minutes (Friel’s approach) is pretty much a “hammer on the head” moment. “It’s a Progressive thing”.
Start at 30 minutes and progress the amount of time you can do. Back in 2017 I ended up just over 60 minutes.
@james_a_c while I’m waiting on some calibrations to complete I spun up my critical power model. It says I should be able to hold my best 40minute power PLUS 18 watts for 3600 seconds. I don’t think that I can! (if only) Could I even do that for 20 minutes? Maybe. But, anyhow, there’s my criteria #2, out the window!
I guess in general that’s the case: CP overestimates what you can do for extended periods of time. Or at least I can never get it to match reality. And that’s for a 5 point model! The problem gets worse for a two point model (so there’s my criteria #1 out the window, too!)
Just the same, I’m sure there are many athletes that find the CP model just fine for their purposes. Just like there are many athletes that have FTP ~75% of MAP & find the Ramp Test suits their purposes very well. For those athletes I think they should use the tool that suits them & not give it a 2nd thought.
I recently read this paper “ Relationship Between the Critical Power Test and a 20-min Functional Threshold Power Test in Cycling”
Which kind of agree with what your saying here. TTE of CP is ~40minutes(in one study), and sits about 20W above MLSS.
The paper goes to mention another study that redefines CP and it’s place in the energy systems - but I don’t understand it to be honest.
What they do mention is the role of central fatigue affecting TT performances over ~11minutes, hence CP protocols having short TTs so the line fits well.
It’s interesting to think about what testing protocols we use, and are they suitable tools? CP has a lot of value on the track, I think, and also for shorter efforts where sitting above MLSS is ok, <30minutes TTs for example.
Well I don’t think CP is right for me…on the other hand I bet it would work pretty good for @HLaB In the end a test that is simple & consistent that has good utility to set training structure is the correct test to use. Whatever that is, you should use…regardless of anybody else’s dogma.
This is absolutely worth a read if you’re interested in the FTP to CP relationship.
Not based on a single users experience.
https://physoc.onlinelibrary.wiley.com/doi/full/10.14814/phy2.14098
From this single users experience, they are both very useful. The anaerobic component of the CP protocol is superb. It is often overlooked by the current FTP obsessed masses.
It gives you the highest possible steady state, and most of the shorter ftp estimations are closer to cp than mlss. A cp test shows you that asymptote on your pd curve, but does not tell you your tte.
I’ve been pretty conflicted before about what number to use… if actually using IF to pace and watching tss as your metric using something closer to 1 hour makes the most sense to me. But if you are trying to target specific metabolic targets, cp may be the better choice. For instance, 4x8 sessions I found were not above cp and most likely reason they are not a huge vo2 stimulus for me, and doing 6x4 5-10 watts above cp got me right where I wanted to be. Now i mostly track minutes and tiz by hr instead.
My delta between cp and 1 hour tte is about 20-30 watts, but the cp testing I did was outside and on an incline which both tend to produce more power for same effort level.
The three minute cp test is relatively easy to execute imo, the ramp test actually crushes me for the rest of the week so I’m not sure it fits parameter 1 for me.
I’m a TT type rider. The last time I compared the ramp test underestimated my FTP by about 10%. I think for my power profile, the longer the test the more accurate the result.
Ramp test estimate: ~300w (~400w MAP)
Alpe de Zwift: ~335-337w for 40ish min.
I try to keep things simple now. Full gas up Alpe de Zwift avg power is close enough for me to estimate my FTP. I think its also good to pay attention to how you feel when doing the longer tests. If in the 30-40min range I feel like I cannot go any harder but can keep going at the current power, I’ve paced it right.
I also do a full gas 5-7min effort to estimate pVO2max for above threshold power targets.
BRWM compared hour, 20 minute, & ramp test results in this video:
272W (hour), 282W (20minute), & 297W (ramp). So ramp test results were even higher for him than what I experienced. He did comment that riding for an hour in erg mode was harder than he thought. All I can say is: that was my experience as well! Maybe another interesting experiment for me to do is hour outdoor vs indoor.
The “what is the right FTP test” stems from the wrong assumption. The “right” FTP is always, always, always only useful relative to a training program which uses that number for its intensities. So, the “right” FTP is one that’s matched to your training program. Period. Everything else is just a number coming from testing protocol, technology, human condition, environment, etc. If TR gives you the Ramp Test and its program (based on that outcome) makes you fast, then everything is good. You have the “right” FTP. So relax. It’s all about the FTP to training program relationship and given the normal curve, TR must be assumed to optimize their program based on their Ramp Test. Due to human variability, you can then tweak THEIR Ramp Test for THEIR program to optimize YOUR results over time by using their tools to vary your workouts. That’s it. FTP is only useful relative to a program. Otherwise it’s just a meaningless number that clinicians might be able to assess in physiological terms but which assessment is also meaningless to you as the athlete. The entire TR program is based on getting that FTP-to-program relationship right for the vast majority of athletes and I trust them to do that. The rest I trust can tweak the FTP number to either get themselves into the programmatic sweet spot (not hard) or to leave. What’s annoying is the ones who should just leave hang around and complain, diminishing the trust we should have in the program and its adaptability. For what it’s worth, I’m in their sweet spot. Ramp Test provides me a great program. I get faster if I stick to it. The program keeps getting better (e.g., Adaptive Training). I get faster. Therefore, I don’t really care whether the Ramp Test FTP number translates to one hour or 40 minutes or whatever. I only care whether it allows me to carry out their program and get faster. V/R Ennis “Enn” Ovone (N=1)
Andy coggon would agree with you
Not really. Correct FTP is crucial for anything below FTP as zones there are tied to FTP. This guarantee you that your SST work is SST work. It is not that if your FTP is overestimated you won’t improve but rather the case that in the long term you can dig yourself a big hole, especially when you follow the percentages of the workouts blindly. If you do HV training, as an example, 4xSST per week is vastly different training than 4x threshold. Then suddenly quite easy sst workout like 3x20 becomes 3x20@FTP. Not to mention beloved by TR suprathreshold workouts when 105% can be your 110% and doing that for 8-10 min intervals makes a huge difference in recovery and stress to your body. So no, FTP is not tied to the program. If it’s overestimated/underestimated you will get into problems sooner or later. The ramp test works for many people because TR isn’t focused on working on your TTE and workouts in plans have quite short intervals so it compensates a little for the issue but then a lot of people simply burn out on this, only because their FTP can be hugely off.
Not to mention that with FTP from longer test sometimes you will find that you can go…longer? So suddenly 8 min intervals in sst are not enough and you start to go longer and improve your durability and endurance more. FTP from ramp test can work or can not work for some users. FTP from long test works for every user - simple as that.
You are making my point while disagreeing enthusiastically! FTP is only useful relative to a program–your whole discussion about how FTP relates to a program is, of course, A PROGRAM. So yes, an FTP will relate to the program workouts you are suggesting in a particular way, and if that way is wrong, then either the program might be too weak or strong, or the FTP measurement device associated with that program might be too weak or strong. So, in the end, FTP only relates to a program. It’s a rather non-controversial statement so it’s always interesting to watch people disagree–mainly they are just not reading carefully or I’m not writing clearly. So, again, for those that missed it, your FTP is only something peculiar to you, your power measurement tools, and your protocol and its only meaning is relative to your (or a particular) program. The FTP could be 20 points off another FTP protocol and you could be riding an A+ program if they are well-matched for you. The real issue is FTP person/tool/protocol MATCH with a program. That’s what TR provides–a matched program that they can tweak to match better by using lots of big data from lots of people riding their program. So, take a big, high-FTP breath and quit believing that FTP is some sort of independently confirmable unicorn in the sky. Humans are way too complicated for that. The key is getting the right FTP-to-Program match.
I’m always curious by those that make your argument, as it goes against what physiologists and coaches say.
The head coach of TrainerRoad, the guy that created the program, has a blog post that was recently updated:
and it still says this:
TR workouts are based on Coggan classic power zones.
Sustainable power at threshold, your FTP from an accurate field test, is not useless and not about matching it to a ‘a program’
About 5-6 years ago Coggan introduced iLevels to replace classic levels. The motivation is shown in this power-duration chart that has 200 different athletes power curves on a chart:
Which shows that sustained power as % of FTP is basically the same for athletes above say 8-10 minutes, and short power is highly variable between athletes.
If that image is anywhere close to scaled correctly, there’s still like a 20% spread in power at those endurance-type durations. Besides, I’m not sure how prescriptive my own curve would be. I haven’t ever gone all out for the 60 +/- 15 minutes or whatever that my FTP is supposed to measure. I’m just some guy who likes to ride bikes; I don’t need to do that to myself. (But maybe next year…) You still need some sort of test that approximates your prolonged sustainable power. Any test except the “go all out for an hour” test is going to be an approximation, as shown on that graph.
What I really want to know is how that one rider’s power curve go out past 100k seconds - 27 hours!
Forest and trees. Forest and trees.
Keep it simple and do 30+ minutes in time-trial mode. I worked up from that to 70 minutes, over 3-4 months, doing only one longer threshold ride a month. Long FTP testing is a combination of mental and strength endurance. Kolie Moore’s ftp and new protocols blog post on TrainingPeaks is worth a read.
27 hours? I know several folks that got into randonneuring and have done longer 400km and 600km brevets.
I forgot about brevets. That checks out.
I think the problem of estimating FTP is interesting. As you mention here (and elsewhere) it’s not just a physiological limit. If your mental game isn’t on point, you’re not making it a full hour at the target power. Heck, the same applies to the ramp test. I hate those last couple minutes! There’s so much that goes into getting the “right” protocol that works for most people and has high compliance. It’s tricky.
