Depends on the test. If you take a traditional 20 minute test, you take 5% off. I have seen one on youtube that just took the 20 minute value as the ftp, but there were some major sprint efforts before the 20 minutes, so I guess it was taking that into account.

You take the number as-is if you do the test for roughly an hour, give or take a few minutes.

What is/was the duration of your test?

---

Brian Stover USAT LII
Accelerate3 Coaching
Insta

20 minute test

---

"see the world as it is not as you want it to be"

From a 20mn test, the general rule is indeed "remove 5%" to the value measured in order to get estimated FTP.

I have not done the Sufferfest 4DP test before but from what I can remember the 20 min effort in this protocol takes your average 20 min power and uses this with no correction as your FTP, I think the theory being that because you have already done significant efforts as part of the test then your 20 min power more closely resembles actual FTP from the fatigue etc induced.

Most other protocols as far as I can gather take 95% of your 20 min average power.

Yes, if you do something significant just before, apart WU, it is not any more a "20mn test" but a "20mn portion out of a longer test", and of course the relation between power estimated during this "20mn portion" and FTP can be very different : instead of approximately 5% higher, it can be 3% higher, or equal, or lower than FTP. Depending what you have done before this 20mn portion ...

FWIW, I know almost nobody I follow on Strava that trains formally that could do 95% of their 20min number for an hour. Totally anecdotal.

Now, I only follow folks up to probably Cat 2 or 3 road racers so the variability index and their power duration curve is going to be a sharper drop than a TT or triathlete.

I'd expect the longer the effort durations that are your specialty, the closer to 95% you'd get.

I think it works for training zone purposes to use about 92% for me. 92% is much closer if you tend to train anaerobic more (bike racer) and 95% if you train aerobic more (triathlete).

Yeah, I've never actually tried to ride for a full hour to see if it works. But what I have done is found my supposed FTP, then that was used to find what I "should" shoot to ride at for 56/112 miles then trained to hold that for the necessary time and it's worked out okay. So I could see training for the specific 1 hour fitness attempt might make it possible. But I know that just because the 20 minute test says my FTP is xyz, that I can't hold that truly for roughly an hour unless I'm trained for that effort.

But I think the OP was trying to take that further and state that some people are not even taking 5% off the 20 minute test any more. I remember now where I saw it....GCN had one of the presenters do the Sufferfest test that another poster mentioned. He did some intense efforts before the 20 minute test, so he was starting already significantly tired. So that was the reason to take the 20 minute test at full value.

I've started using Trainer Road's ramp test. Like you said, I never could truly hold the 20 minute test anyway, so I just use that value to set training intensity. I wouldn't trust it nearly as much as a good 20 minute test to go out and pace a time trial purely from that info, but I didn't ever do that anyway. From the training based off that ramp test, I really learn what I can hold for a race. I prefer the ramp so far since it feels more repeatable. I had many bad 20 minute tests where I maybe didn't feel that great or just did a bad job at pacing and the whole test was useless.

It probably takes consideration for the type of training efforts you intend to do, as you mentioned. I've always used a warmup, couple VO2 max efforts, brief recovery, then 20-min test for FTP testing. I've also typically used 95% of that value and then rounded up to the nearest five or ten. But then I've had days where I've held above my supposed FTP on rides for longer than an hour because I was feeling really good that day... I've also had days where I couldn't seem to hold FTP for 10-minutes because of the opposite...

If the purpose is to set training zones, the difference in taking 100% or 95% of a 20-minute test is likely small. Might it be of value to take the higher number just to train harder?

---

Blog | Strava

95% of a 20-min test still way overestimates my FTP, even with a 5-minute all out VO2 effort done just minutes before. It depends on the individual. Likewise, a 30-minute test still overestimates my FTP. I like to do a 40-minute test, which I can mentally break it down into four 10-minute TT intervals to make it a great workout.

As a Triathlete (not bike racer)...yes, in-season I can hold 95% of 20mTT for an hour (or longer). But, I work up to those specifically during the build up to race-season following a progression like:

2x20 -> 3x15 -> 3x20 -> 2x30 -> 1x45 -> 1x60 -> 2x45 -> 1x90

all at ~95% FTP.

It just takes some time to adapt to the fatigue. Typically, I do a (2-3)xYY @ FTP on Tuesday, and a 1xZZ @ 95% on Saturday (in the middle of my long ride). Usually, its the same total duration of "ON" time for Tuesday and Saturday (eg, 2xYY = ZZ).

  
Lack of base endurance and sweet spot training, may be, leading to a steep CP curve ?

I've done 40+ min Zwift races averaging 105-110% of FTP (as calculated based on a 20min test, with the correction). Similarly I did a 1:12 ride on the weekend (including alpe du zwift), averaging a few watts over my FTP for the first hour (not surpsingly Zwift detected a new FTP value a few watts higher...), before the descent...

So what you mean to say is you're not good at the 20 minute interval of that particular testing protocol?

My understanding is the "20mn test" was always the "longer test" that included v02max depletion interval beforehand, and the 5% subtraction is still used. Doing simply a 20 min test without the depletion interval and only subtracting 5% is a misinterpretation or mischaracterization of the test http://www.hunterallenpowerblog.com/.../01/what-is-ftp.html

What was your 20 min test protocol

That sounds HORRIFIC. Just do a ramp test and get it over and done with.

I've been using the ramp test for nearly a year (maybe 6 tests in that time), and have found it to be great. Seems to be consistent with results from my previous estimates using either the 20' or 2x8' TR tests. It's changing in the way I'd expect it to change depending on training load. And also seems to track pretty well with actual performance over efforts of ~1 hour outside.

Re the latter, I suspect the reason many people would struggle to achieve their estimated FTP over an hour is simply that there are very few opportunities where you have both the conditions and motivation to do this. It takes a huge amount of motivation for most people to do a 1 hour max effort, it also takes a course where you can apply steady power throughout e.g. a mountain climb or a flat non-technical TT. I did a couple of bike races last year which were basically perfect for FTP testing - flattish section to start followed by a big mountain climb and a summit finish. ~40 and ~65 minute climbs respectively. The 65 minute effort was pretty much bang in line with ramp test results from a few weeks early - ramp test gave me 320W, climb was 315. 40 minute climb was at 322W, should have been higher for that duration but the flat section before the mountain was more technical and there were a lot of riders trying to break away and give themselves an advantage before the climb started so I had to work quite a bit harder before the climb started.

Can't see myself ever bothering with anything but the ramp test again, in my case it's certainly more than accurate and consistent enough for measuring progress and setting training zones. And it's so much better than the 20 minute test both in terms of the level of mental toughness you need to summon before attempting it, and the time after to recover.

couple of things to keep in mind

1. the 20 min test should have a 5 min vo2 max effort in front of it
2. the 20 min test is valid for ~ 60% of people.
3. Are you in the 60% or the 40%?

---

Brian Stover USAT LII
Accelerate3 Coaching
Insta

1. I did the Zwift 45 min test which I think has at that in it
2. I wasn't aware of that
3. I don't know and wouldn't know what to substitute for it instead.

---

"see the world as it is not as you want it to be"

Yes, we use 100% of the 20 min result as part of our Full Frontal test. Here is why (from: https://thesufferfest.com/...l-fitness-test-faq):

Q: My FTP from Full Frontal is lower than from other FTP tests. What’s going on?
First, let’s get this out of the way: you are not your FTP. Threshold power is only one way you put out power on the bike. It only tells you how well you can generate power for long, steady-state efforts. Once your power goes north of your FTP, all bets are off.

Now, bear with us: here comes the science. After testing hundreds of athletes in his lab at the Boulder Center for Sports Medicine, Coach Neal Henderson found that for many athletes, their lab-determined FTP did not match up to the traditional 20-minute field test. The assumption had been (and continues to be with conventional FTP tests) that 95% of an athlete's 20-minute power is equal to their FTP. Unfortunately, it’s not that cut-and-dried. Two athletes can have the exact same 20-minute power, but produce that power differently. One may produce 97% of it aerobically, while the other may produce 93% of that power aerobically. Assuming 95% of an athlete’s 20-minute equals their threshold power is the same as assuming everyone who is the same height has the same inseam. Not only does this make the 20-minute power test imprecise, it also misses out on other important aspects of a rider’s physiology.

If you are a rider who is only producing 93% of your power during a 20-minute test aerobically, then that means you are drawing on other energy systems—either Maximal Aerobic Power (MAP) or Anaerobic Capacity (AC)— to generate that power. Instead of measuring your threshold power, the 20-minute test is really measuring multiple energy sources. That results in an FTP that is artificially high. By measuring four distinct ways cyclists produce power, Full Frontal will be able to tell us where that additional power is coming from. A rider who is only generating 93% of their power over 20 minutes aerobically will often have 5-minute power (MAP) or 1-minute power (AC) that is well above average . Since the 20-minute test in Full Frontal comes after both the 5-second sprint efforts and the 5-minute test, you will not be able to hold the same 20-minute power in Full Frontal that you would in a conventional 20-minute test like Rubber Glove. Because of this, your FTP from Full Frontal is based on the full average of your 20-minute effort, rather than 95%. But to be accurate, you have to go all-out on the 5-minute effort that precedes it.

While this might tempt some people to go easy in the 5-minute effort to save something for the 20-minute effort, you need to keep in mind that the 4DP™ power targets for workouts are based on all 4 metrics (and the relationship between them), not just FTP. If you sell yourself short in one area of the test to improve another area, the training you do afterwards won’t be accurately tailored to your true abilities, and you will miss out on the real fitness improvements you are looking for!

Lastly, it’s important to remember that training is a means to an end. You train and suffer today so you can be faster tomorrow. Fixating on a single number at a single point in time isn’t productive. What point is there in having an artificially-inflated FTP if you’re not getting the right training stimulus out of a session? Since 4DP™ takes all four metrics into account and personalizes all of your targets, you are ultimately going to get a much more effective workout, even with a “lower” FTP.

---

David McQuillen
Founder & CEO of The Sufferfest

That’s a really excellent post.

The only thing I’d add to the OP is that my experience is “newer” athletes tend to get overestimated results from the 20 minute tests because it’s far easier, mentally, to pound out 20 minutes hard than it is to ride a 40k or 1 hr TT as hard as possible.

It’s kind of like marathoning....anyone can do a great 18-22 miles at their “goal” pace....but can they sustain that when the body really starts to rebel and all the senses are telling them to slow down. Same thing in a TT. For me it starts about 45 minutes in. It’s not “can” I hold the pace, but WILL I hold the pace for the last k’s?

That’s why you see a lot of older athletes still turning out fast efforts. They’ve made friends with that discomfort zone.

The most important guidance that you'll find is not to use FTP for prescribing workouts. So if you're only using FTP to gauge improvement in ability to perform a test then it doesn't matter because you can do any test you want. If you're using FTP so that you can plug it into some generic SW/plan/model then you've already lost. You're much better served by learning about load monitoring and then figuring how have YOU can do "more" with the same effort.

---

Interval Design Studio
YouTube | SoundCloud

It depends a lot on what your goals and resources are. Not everybody has the interest, ability and/or time to teach themselves enough to do a better job of setting their training plan than a generic plan. Particularly amongst triathletes who are already juggling 3 sports as well as what's already going on in their lives. And who also, it needs to be remembered, will do all or most of their racing at sub-FTP levels. There's a reason why triathletes focus on FTP a lot more than road cyclists, and it's because when you're racing at solo steady efforts below threshold, it's the single best metric to track. Pure cyclists need to take a more balanced view across metrics at different durations, because they need to be able to sprint, accelerate over short hills, do multiple pulls above threshold in a breakaway, etc.

They're not perfect, but generic SW plans like TR or Sufferfest are both cost-effective (much cheaper than a coach) and time-effective (much less time than putting together your own plan), and do a pretty decent job of getting good results from any given effort.

You make some good points but I'd also urge people to consider that the relative simplicity of the bike leg compared to cycling makes it even easier to understand and self-coach; once you have a budget within which to work. But yeah I get it that some can't be bothered and there's nothing that can be done about that. Doesn't change that they'd get better results by individualizing their training, though.

---

Interval Design Studio
YouTube | SoundCloud

I can't speak for others, but I find that my FTP ends up being ~92% of a 20 minute test. I validate my FTP using HR as I know my FTHR.

That's not particularly helpful.

1. Its a bit of hyperbole to say "you've already lost"---as if to imply that someone will not improve following such a methodology.
2. You provide no definition of "load monitoring", or practical advice on how to use it to "do more with the same effort".

So, perhaps you convince the audience NOT to use "FTP" for designing workouts, but then you don't replace it with anything equally practical---just an esoteric declaration of "use load monitoring, instead."

Additionally, while digging around I found your "winter training program" pdf, which specifically prescribes workouts using % of FTP.

Long run, it doesnt really matter. Do some workouts with the FTP that are given. If they are too hard, scale down the FTP a little bit. Too easy, scale it up a little bit.

Ok i thought it was me missing something. I need some standard to start with as a baseline and for now FTP is it.

---

"see the world as it is not as you want it to be"

I'm a bit confused. How should one prescribe workouts?
I don't use a generic plan but I do use FTP as a metric to judge intensity of my efforts. I don't know what you mean by "load monitoring", are you referring to training stress score or similar as calculated by say Training Peaks? If so, does it matter what sessions are used to accumulate TSS or anything will do? When you say I should figure out how I can do more with the same effort, what are you referring to? More of what? TSS? What do you mean by the same effort? Do you mean work, intensity, TSS, perceived effort, or something else?
What has ones budget got to do with self coaching?
How do I individualise my training? I can make up my own plan (and I do) but how do I know if it's better than a generic plan? Why is individualised necessarily better, without knowing in what way it differs?

Help! I don't know how to follow your advice.
However, I'm glad to hear it's simpler than cycling training and can be done by anyone as long as they're bothered. That's reassuring.


Were you actually trying to be helpful? If so, I suggest you provide some solid, meaningful, followable guidance and rationale. I'm sure that would be appreciated. I'd be interested to hear what you have to say.
However, as it stands, the above quoted posts come across as a load of unhelpfully nebulous posturing waffle, that pours scorn on some of the most common training methods without providing any rationale/justifications or explaining any alternatives.

Do not believe and trust anything you read from the average Joe on the internet. Only believe and trust things you read from well known, reputable sources.

Or that my FTP result is out of date, and I need to re-test... But yes, I find it's always much easier for me to push myself in a race setting than it is in a workout type setting... While I suffer like a dog doing 30s intervals at 700+watts with easy recovery, I can produce a number of those types of efforts to avoid getting dropped from a breakaway, with riding at around FTP in between...

That was rather my point. Not sure I got the sarcastic tone quite right in that post ;)

We made a *personalized* 20min > 60min FTP calculator last year as an app to take into account whether you did a protocol, are a regular or beginner and depending on your starting power. The power-time curve looks different if you are a beginner vs pro and hence the calculation is not the same 95% for everyone. See https://goo.gl/jS78B8 (ps credit to FFT for the underlying data analysis).

Just circling back to this thread and didn't read the rest of the replies.

What I may do if it were me and frankly I enjoy the suffering associated with racing & testing, I may do a 5 or 6 min all out test and then a 60 min TT. Or do the 5-6 min test after about 60-90 min of riding then the next day do a 35-45k TT where I pinned a number on. Get in a 20-30 min warm up, hit the TT, ride 20- 30 min easy after. You'll have more data than someone doing a 20min test & more importantly you'll have better data. (plus it's not too shabby of a workout either)

---

Brian Stover USAT LII
Accelerate3 Coaching
Insta