CTL/ATL and stagnation
- Thread starter kmavm
- Start date
Yep, and as long as it's applied with an eye on the composition of the training load, which is not something the PM shows, it makes for a killer ap.frenchyge said:
An interesting chat.
thought you were questioning something as fundemental to the understanding of the model as whether TSB was a surrogate for performance... which it is.. but yes like everthing else in PMC it's only relatively so...frenchyge said:
i would tend to think of CTL as just another measure of a member of the composition of fitness... the part that will indicate how well addapted you are to long races and multi-day races... i would put it on the same level as FTP, AWC, NMP... and also CTL... not above it or in a separate category... depending on event, for some CTL will not be as important... and thinking of it that way you don't lose sight of the importance of the composition of fitness.. in fact it might be nice to put CTL on some of the other graphs in order to track one's progress... for that matter, i know that it's not changing every day or anything, but wouldn't it be good to be able to track FTP or give some indication of where you've changed it?Alex Simmons said:
Have you read the article you yourself linked?doctorSpoc said:
From the "science" article:
So, TSB is and has always been a surrogate for "freshness" not for "performance"... I don't know where your "misconceptions" came from, but I know the papers you linked aren't the source.
Ken
no matter how you slice it, in PMC, TSB IS the surrogate for performance in the impulse response model... in precisely the way Dr. Coggan uses the word surrogate... if you are saying it isn't you are wrong! notice in the quote you gave he doesn't use the word surrogate? you used surrogate as a surrogate for indicator... and you wrongly think that analogous is analogous to surrogate... read again.. I'm sure you'll figure it out eventually...sugaken said:
"subtracting the area under the the ATL curve from the area under the CTL curve gives you a value (TSB) that turns out is very representative of changes in potential performance" and as i later said in response to frenchyge query, only relatively so... want to call that freshness.. sure... freshness is probably a good word... but unlike Dr. Coggan i think it is painfully obvious that "fitness" is not the direct equivalent of CTL or even necessarily varies directly with changes in CTL directly except in a few circumstances... CTL is only an indicator of a contributor to fitness... its contribution varying depending on your event and the important performance metrics for that event... so i consider most of the rest of what he says in relation to CTL as being some what off base...
man... in PMC TSB is not a surrogate... it is THE surrogate for performance in the impulse response model... that doesn't mean it's equal to it, that doesn't mean if varies as it... it is though absolutely, 100% the surrogate for it though... look at the formula...sugaken said:
does this help...
acoggan said:
TSB is 100% the surrogate for full adaptation to a given training load if you look at the formula or once again from the article you linked:doctorSpoc said:
Using the definition that you posted:
you can't really say TSB takes the place of or is a substitution for performance when the article defining the origin and basis clearly states that performance depends upon both TSB and CTL.
I've really lost track of what we're debating here. Frenchyge posts his thoughts that you can't seperate TSB from CTL when predicting performance. You reply that no one has bothered to read the background articles describing the Performance manager. Several of us point out that those very articles support Frenchyge's post and now we're debating the use of the word surrogate???
What exactly is your point? Are you arguing that TSB predicts performance without regard to CTL? Are you arguing that your use of the word surrogate is valid regardless of the TSB/CTL/performance relationships? Are we talking physiology, adaptation, and cycling are have we moved on to a more general discussion of the english language?
-Dave
That's a little restrictive. It works just as well for a track individual pursuiter.... Indeed, one of the very first examples of its usage by Andy was for a women's elite national pursuit champion. But that would be the "it's an aerobic sport, dammit" line again...doctorSpoc said:
If you're saying that TSB is the surrogate for the impulse response model's "performance", but *not* for performance in the PMC concept, then I agree with you. I misunderstood you to say that TSB was the performance indicator for PMC.doctorSpoc said:
Impulse-response: Performance = k1*Fitness - k2*Fatigue
PMC: Performance = k3*Fitness(CTL) + k4*Freshness(TSB)
doctorSpoc said:
It seems to me that you are the one who is misinterpreting things here (or perhaps I should say overinterpreting them). That is, I don't know how to say it any more clearly than this:
"Thus, within the logical constructs of the Performance Manager, performance depends not only on TSB, but also on CTL (in keeping with saying that “form equals fitness plus freshness”). The “art” in applying the Performance Manager therefore lies in determining the precise combination of TSB and CTL that results in maximum performance. To put it another way: in the Performance Manager concept, an individual’s CTL (and the “composition” of the training resulting in that CTL – see more below) determines their performance potential (at least within limits), but their TSB influences their ability to fully express that potential. Their actual performance at any point in time will therefore depend on both their CTL and their TSB, but determining how much emphasis to accord to each is now a matter of trial-and-error/experience, not science. "
You also note that I went to great lengths in the article from which the above quote is drawn to emphasize that, despite the use of equivalent time constants, ATL and hence TSB behave differently than fatigue and performance as predicted by the original impulse-response model. I therefore don't see how/why anyone would conclude that the two approaches are directly analogous.
rmur17 said:
I tried to answer you as well.
Functional threshold power is just that: your functional threshold power. So, regardless of the time-course of physiological adaptations that contribute to this performance ability, I'd expect the time constant of the fitness component of the impulse-response model to be a good estimate.
If that logic isn't clear, try thinking about it this way: the impulse-response model has been studied extensively, and it is amazingly robust. Even so, it is a bit of an artificial construct, i.e., the model is really as much or more empirical in nature than it is based on first principles. In particular, the impulse-response model assumes a linear, non-plateauing relationship between training load and fitness, which is clearly wrong. This inherently invalid assumption, or perhaps others, may account for the discrepancy between the changes in, e.g., blood lactate during exercise at a constant power vs. changes in actual performance ability. Nonetheless, the impulse-response model still provides a quite reasonable fit to the data.
i asked if anyone had read the article... i was pointing people to the article to look at how the model was formulated, not to just blindly accept it's conclusions... and i would hope they would read it in the context of what was being discussed in this thread...daveryanwyoming said:
lets look at the point i was making that triggered frenchyge's query:
i was making the point that... and i'll try to be more explicit:
- PMC is based on the impulse response model
- the impulse response model is a black box model who's output is performance
- TSB is literally the surrogate, it takes the place of performance in the equation (no gain terms, no P(0) so only a relative measure)
- when using a black box model it's not wise to draw any conclusion from the terms inside the "black box"
- fitness and fatigue terms are inside the "black box" in the impulse response model, so it is not wise and is not proven that the fitness term actually accurately represents actual fitness or that the fatigue term accurately represent fatigue.. but this is not necessary... the only thing that matters in a black box model is that the output of this equation accurately predicts performance (experimentally verified)
- in PMC CTL is the surrogate for fitness and ATL is the surrogate for fatigue.. again literally the surrogate for the terms in the equation...
- the fitness term in the impulse response model has not been shown to be an accurate measure of actual fitness and so i would conclude that a simplified surrogate for it (CTL) is likewise not been show to be an accurate measure for for fitness and in fact, in this thread has been shown to be inaccurate in many instances
- it has been demonstrated that the performance term (the output of the impulse response model) to be accurate and is really the only validated term in this equation and so i would conclude that its surrogate (TSB) even though is only a relative measure of performance to be the only somewhat valid term in the PMC model
misinterpreting my use of the word surrogate... but also completely missing my point and ignoring 4 pages of post showing that it's probably not wise to use CTL in that way... so the article may support frenchyge's assertion but i don't believe the model does and much of the discussion in this tread doesn't.. so when i read his question i was like... ???frenchyge said:
I didn't misinterpret the use of 'surrogate', but rather which model you were applying it to (impulse-response, where it's fitting, vs PMC, where it's not). I intentionally ignored your point and the other several pages of discussion because I had no desire to debate your views on CTL and fitness components.doctorSpoc said:
My only reason for commenting on a discussion in which I otherwise had little interest is because I thought you to be stating that TSB was a surrogate for performance *in the PMC model*, which of course would be incorrect. Or, as you later noted, I was indeed.....
.... as it related to the PMC model.doctorSpoc said:
In the PMC model: Performance = CTL + TSB, so of course substituting TSB in place of Performance would be incorrect.
