Originally Posted by An old Guy .
The discussion has nothing to do with training with power. The discussion has to do with the concepts of NP, TSS, and others that are related.
You say there has been no contrary evidence. How is this:
acoggan has posted that NP is not valid for times under 20 minutes. NP gets overestimated until one gets rid of a lot of glycogen.
I think the point of this disclaimer is to take into account that during the initial 20 minutes of of a ride, you are essentially warming up. This statement applies to the first 20 minutes, not ANY 20 minute period. During this time, your energy is produced from fat stores and lactate accumulation would not accurately follow the 4th order relationship on which the NP algorithm is based.
Now, acoggan is saying that TSS achieved below 75% IF is not the same as TSS achieved above 75% IF. (It is really hard to know what his objection is. He seems to be unwilling to share what TSS is modeling except that it is "some process" that is not glycogen depletion.
Unless TSS is accumulated under conditions at least similar to the conditions on which the algorithms are based, the TSS calculation is not correct. AC apparently determined that an IF of approximately 0.75 was the threshold for the accurate application of his model. Since the r2 value for the regression analysis of the blood lactate values was something like 0.88, the 4th order calculation (the exponent is actually 3.9xx) will in fact over estimate at one end of the curve and under estimate at the other because the relationship is not precisely a 4th or 3.9xx order relationship. This is not an indictment that the model is wrong, but rather a recognition that the model has a range in which it operates most accurately.
(A current anecdote: There is a current thread I read earlier today on this site. It seems some people think you need to be doing L4 for at least 8 minutes before it becomes L4. If so, NP of the first 8 minutes of any L4 interval is over estimated. And nothing special about L4 so every interval is overestimated. And that is why you are all above average.)
It doesn't matter what some people think about how long an L4 interval should or should not be. If someone hops on their bike and pounds out 20min L4 intervals with no warm up, yes, the first 20 minutes will indicate a TSS that is biased on the high side. I don't know anyone that does not warm up prior to their interval workout - do you?
---
You seem to think that anecdotes are bad. But both acoggan and Andy Coggan offer less than anecotes. Just a story that was concocted in a couple hours. acoggan says his theory cannot be disproven. It would be easy to disprove any of his claims if he wouild say what is necessary for him to accept a proof.
A PhD thesis is hardly a conversation concocted in a couple of hours. Actually, it would be quite easy to disprove AC's theories. Do the work and show that the blood lactate curve is wrong and does not follow the relationship presumed in the NP algorithm.
---
I find it amazing that people tell anecdotes about their training and make claims that they cannot maintain 300TSS for many rides. Going from 15-20 hours a week to 35 hours a week is a big step, but 7% (more riding time) a month gets you there in a year. Not too many people set high TSS as a goal. They simply want high FTP. But ...
Joe Friel has a new commentary out. He thinks that more riding time produces higher FTP. (You can find his site.) Post a link - his blog site does not seem to have this information as this would seem to be a significant departure from his recent trend towards the time crunched training approach.
---
So what proof do you want.
Post your files supporting your claims