How does the error of estimate work, actually?
Is it just a description of how well the model (i.e. the calculated FTP and FRC) fits to the MMP-curve, or is more to it?

Just picking a link at random, see point 3:

http://stat.ethz.ch/~stahel/courses/cheming/nlreg10E.pdf

So how does FTP relate to the anaerobic threshold, which this paper found to be associated with a TTE of 60 minutes?
https://www.ncbi.nlm.nih.gov/pubmed/2242751

On average, FTP, "anaerobic" (i.e., ventilatory) threshold, MLSS, IAT, the iEMG threshold, the NIRS breakpoint, etc., etc., etc., all correspond to (essentially) the same exercise intensity. In fact, FTP is a blanket concept that explicitly recognizes this fact.

However, whether you see exact correspondence in any given individual, or even in a study of a small number of subjects, will depend in part upon both the exact methods used as well as random variability.

The above is so well-established by the existing literature that additional studies simply comparing the various "thresholds" are, IMO, a waste of time. To further advance the field, investigators should be focusing on why any small discrepancies might exist, i.e., are they entirely due to random factors, or are their some underlying, cause-and-effect physiological relationships?

Note that some inroads into such questions have already been made, e.g., two decades ago it had already been established that interval training tends to have a bigger effect on VT than LT, whereas the opposite is true with steady-state endurance training.

Unfortunately, many younger scientists seem to be unaware of the numerous previous studies in this area, which really developed starting in the late 1970s. As a result, they spin their wheels repeating experiments that have already been performed, or designing studies the results of which are entirely predictable.

Note that the final comment also applies to this study:

https://www.ncbi.nlm.nih.gov/pubmed/29203319

That is, it is already well-established that the range of test durations used as well as the formula chosen has a significant impact on the calculated CP (and W'). Thus, while it is handy to have all of the comparisons in one place, and the paper may be publishable for that fact alone, at the end of the day nothing new has really been learned as a result of the authors' (and subjects') efforts. Their time and energy would have been better spent pursuing a more important/novel/unanaswered question.

The PD curve is a strictly decreasing monotonic curve. Therefore, FTP/CP cannot be identified as a point along the PD curve. There is no basis for choosing one point over another unless you start considering first derivatives. Besides, since each point on the curve is derived on different days, albeit different weeks or months even, one can't really argue that the rate of change of the PD curve aligns with some physiological phenomenon.

One can use the CP/W' model to yield a consistent and repeatable measure for threshold power. When you consider the amount of work available above threshold (W'), this work can be consumed above threshold, independent of intensity. (Note that in Xert we further constrain the ability to use up this available work across all durations. This is our concept of MPA. For example, you can use up more W' at 20 minute power than at 5 minute power. This is a main difference in how Xert works vs. how CP/W' works.)

This principle does not hold true when you consider lower values for threshold. Let's look at some PD curves to understand why:

https://www.xertonline.com/calculator?power_0=315&duration_0=1200&power_1=359&duration_1=300&power_2=1000&duration_2=1&action=calculate


High Intensity Energy is analogous to W'. The calculation yields an HIE of 18.6kJ and an FTP of 300W.


https://www.xertonline.com/calculator?power_0=315&duration_0=1200&power_1=364.4&duration_1=300&power_2=1000&duration_2=1&action=calculate


This yields an HIE of 21kJ and an FTP of 298W. We kept 20 minute power and 1s power the same for convenience.


So for a 2W difference in FTP, we see a 2.4kJ increase in W'/HIE which would then imply a 5.4W increase in 5 minute power. If indeed FTP was 2W lower then the additional W'/HIE would then be available for use at intensities above FTP. If this is not the case and 5 minute power is 359W, then FTP must be 300W and not 298W.


This example is a simplification of how Xert does this and how it is able to consistently and precisely measure FTP. It will pick up small variations in FTP that can then be tracked and managed with training.

---

Armando Mastracci, Founder of Xert, an advanced data analytics and training platform. Blog, Podcasts

You're as bad as Trev. Worse, actually, since you are just here to shill for your product.

As for "choosing one point over another", our bodies certainly seem to think that it can be done, e.g.:

https://www.ncbi.nlm.nih.gov/pubmed/28302463

Given results such as these, only a fool would buy into your argument above (especially the last bit).

Actually, we have more business than we can manage at the moment. Just here to provide folks a unique perspective they may not have heard before.

Going to bow out now and let you folks get back to where you were.

---

Armando Mastracci, Founder of Xert, an advanced data analytics and training platform. Blog, Podcasts

Just for fun. Who agrees with this?

Unless you're a Trumpian, you can't really disagree with established facts (lyrrad was referring to the WKO4 model).

That is the point. The model does not reflect reality.

You could ask a timetriallist to perform a 20 minute effort at their 20 minute power and the result would be just as accurate
You could ask a timetriallist to perform a 30 minute effort at their 30 minute power and the result would be just as accurate
You could ask a timetriallist to perform a 40 minute effort at their 40 minute power and the result would be just as accurate
You could ask a timetriallist to perform a 50 minute effort at their 50 minute power and the result would be just as accurate

Whatever duration you ask them to perform, they will pace it well, because almost all racing cyclists worth their salt can do this over any duration in which they have some experience racing at.

Clearly this study in no way validates that there is anything special about MLSS

"When power exceeds FTP, fatigue will occur much sooner, whereas power just below FTP can be maintained considerably longer." This statement, apart from being extremely vague is misleading, because it sets FTP apart as if it is the only power at which this occurs. You could say the same for any duration on the power duration curve beyond the first few minutes.

Why do you perpetuate this threshold power myth?

When power exceeds any power, fatigue will occur much sooner, whereas power just below that power can be maintained considerably longer.

THis is no more true for LT than it is for any other power, as Armando just implied (and then got insulted for saying it)

Functional Threshold Power is an estimate of the point where we transition from using a mix of carbohydrate and fat to solely using our finite CHO stores. This point is observable when doing ramp tests or the MLSS test.

MLSS is better estimate of this true threshold point than a ramp test although you can observe the transition using lactate, heart rate, NIRS, and expired gas analysis.

Because of the issues with performing ramp tests, where we see a lot of guessing where the inflection point is, if not obvious, and the challenge that performing tests in the lab is not specific to where we actually perform. Plus, the costs and logistics of doing lab test. Also, our physiology changes very rapidly, so, metrics determined from a test have a very short period of usefulness.

So we estimate threshold power from data from the bike in the field. Not just the FTP, but peak power (PMAX), sprint power (PMAX/FRC), track time trial power (FRC), track pursuit power (FRC/FTP) and time trial power (FTP) and long term power (Stamina). To help us plan our training and to teach riders to learn what performance at a given level looks like.

---

Hamish Ferguson: Cycling Coach

No one is trying to educate Trev as he's not interested in learning and he has no genuine interest in understanding the topic. That's evident from his persistent repetition of nonsense over many many years.

Any responses to Trev which happen to be educational are done so for the wider audience, not for Trev the Troll.

If anything, Trev's attempts to deliberately sow misunderstanding has consistently backfired.

Trolls are a part of online life unfortunately, so we can either ignore them or use their trolling an an opportunity for learning.

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

It's not a myth. There is a power output above which the physiological responses are markedly different from those at even slightly lower intensities. Akin to the renal threshold for glucose spillover in the urine, there is some "splay" or blurring at intensities close to this point, but describing it as a threshold is still a convenient (and widely accepted) paradigm.

Armando is as much of a shill as Trev is a troll.

As for your statement, try maintaining a power, say, 100 W above vs. below your FTP, then come back and talk to us.

You are saying, we have to maintain a power that the rider would NEVER race at to prove your point ? We are only interested in locating where FTP is, and for a rider with a 270w 25 TT power we are looking in a window between their 10TT power at around 290w (cp20) and their 50TT power at around 250w (cp120). What on earth would an effort at around 3 minutes or at around 48 hours give us ? All we are trying to prove here is that 270w is more special than 250w or 290w, because you are saying it is. I maintain that it isn't more special.

Have you ever examined any physiological test data or perhaps read published scientific papers on such things?

I ask because you (and Trev) are claiming the concept of threshold does not exist.

You do so based on the flawed assumption that if it's not obvious just by looking at a PD curve then it doesn't exist (Trev does so because he's a troll). But we've been over this and you still have not got it. Threshold is a physiological phenomenon and the very shape (or lack of shape) of the PD curve in that region is a direct result of the threshold behaviour exhibited by our metabolism.

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

No, you and AC are claiming that FTP is special because a small increase in power leads to a much shorter TTE, and a small decrease in power leads to a much longer TTE. That has been presented many times as an identifying feature of threshold. However the same assertion is also true of every intensity which we sensibly race at - from well before CP20 right to well beyond CP120. Therefore that assertion should not be used to describe threshold.

i.e. I'm just repeating what Trev said.

BTW I don't have a particular affinity to whether a threshold exists or not in terms of lab data - Noakes says no, others say yes. I just think its time people stopped claiming that "FTP" is some kind of special intensity where you can naturally pedal away until the cows come home and its so absolutely noticeable that almost any rider can ride at it on demand.. All intensities have the same features and any rider can ride at 10 pace, 25 pace or 50 pace on demand and get it roughly right every time - they are not likely to get 25 pace more right than 50 pace or 10 pace and AC's quoted study above where he claims this is demonstrated - it absolutely does not demonstrate this whatsoever since they only tested one intensity - other intensities would have given the same finding.

There are none so blind as those who will not see...

To reiterate: "threshold" refers to the physiological responses, which essentially cap your sustainable power at an upper limit. It is the very flatness of the power-duration relationship (where power declines relatively slowly over a wide range of time, i.e., hours) beyond the first few minutes of exercise that results from this fact. The reason for focusing on a particular approximate duration is NOT because of any inflection point in that region, but because power at distinctly shorter (e.g., 20 min) or longer (e.g., 2 h) will be significantly above or below the maximal steady-state intensity.

So then you must agree that this statement is misleading:

"When power exceeds FTP, fatigue will occur much sooner, whereas power just below FTP can be maintained considerably longer."

Because,

"When power exceeds CP20, fatigue will occur much sooner, whereas power just below CP20 can be maintained considerably longer."
"When power exceeds CP40, fatigue will occur much sooner, whereas power just below CP40 can be maintained considerably longer."
"When power exceeds CP60, fatigue will occur much sooner, whereas power just below CP60 can be maintained considerably longer."
"When power exceeds CP120, fatigue will occur much sooner, whereas power just below CP120 can be maintained considerably longer."

Like Trev, you clearly don't understand the signifance of the study I mentioned. It isn't the fact that the subjects held a steady power for 30 min, but the fact that *they honed in on their MLSS while doing so* (and more closely than the CP model).

IOW, perceived exertion indicates that there is indeed a "threshold" intensity, as any experienced athlete (or even untrained subject...just go look up all of the studies of self-regulation of exercise intensity in various training studies) can tell you.

With all that said, there clearly are a few "perceptual idiots" in the world.

No, it is not misleading, as it is an accurate description of reality.

Again, try sustaining a power measurably above vs. below your FTP for as long as you can, then report back to everyone here.