I’m sure what most differentiates the two groups is their actual performance.
I was just imaging that mitochondria have a duty cycle, and that increasing the count extends that duty cycle by reducing load. Then, I was thinking, maybe very high outputs pressure the mitochondria leading to damage, and that ultimately the ‘rate limit’ is about preserving mitochondrial function.
Then I did a quick search, AMPK came up, repairing mechanisms, etc. And so the 
appeared. I can’t vanish into it today.
But it did make me think, you know if you have more then the load is lower, so higher ftp = durability. et voila.
edit but now I’m thinking it’s the fractional percentage of your ftp potential that matters, not your raw ftp. This is maybe where volume matters. Where you are on the diminishing returns arc… Maybe that’s the coffee talking. 
Similar =/= same.
More specifically, keep in mind that every ~1% difference in power corresponds to a ~10% difference in the duration.
IOW, could he beat up on you and others in longer races simply because his muscular metabolic fitness was greater, but the difference was small enough that it was only really obvious under those circumstances, but not readily detectable in shorter events, even though it was really there?
Yes, very noticeably. Minor variation leads to huge outcome differences.
That actually looks like quite a lot of variation, the choice of axis scales just makes it less obvious out towards 1+ hours.
Like it eyeballs out to 20-50% variation in value power among whatever sample size this was at that time range, which is quite large.
Do you have the standard deviation for that data? Or maybe even r^2 between FTP and various >1 hour power points? Or something that quantifies the spread of the sample values rather than just relying on subjective “eyeballing” of the data?
I think this all gels together then. That suggests that FTP is the dominant predictor of endurance performance (even for very long events), which I think people agree with. But that it doesn’t perfectly predict it, meaning there is room for other independent variables to explain the remaining unexplained variation in performance.
Here’s the correlation between mFTP and power as a function of duration. The wonkiness at the far right is again due to more and more drop-outs. Out to not quite 6 h, the R is >0.90.
Or is the residual variability between individuals due to 1) errors in estimating FTP, and 2) lack of maximal efforts at longer durations?
ETA: These data/this discussion is more than 10 y old.
That I don’t know, I’m sure you’re correct. Obviously there was something significantly different about our physiology that was not easily detectable by our fresh power numbers.
Also a sincere thank you for engaging with us (and me) by the way. I bought your Training and Racing with a Power Meter book in 2008, along with WKO and a powertap. I attribute the majority of my early racing success to your book and training principals.
That graph looks to me as if differences in FTP are irrelevant for durations of over 300s (approximately). Surely there is some sort of detail that is getting lost.
That’s patently a false statement.
Would you disagree with this explanation of fatigue ?
Fatigue at the cellular level would be explained by leakage of electrons (or other inefficiency) at any step in the electron transport chain?
I would expect huge differences, in TTE and RPE, for people riding sweetspots based on their endurance/durability, otherwise people would just start their SST progression at 95% for 1h+.
Yes, those things could definitely cause this.
I think the other thread that’s talking about this “durability” concept contains references/reasoning that would argue that mismeasurement or lack of data isn’t the only cause of this though.
Like how at a world-level, performance is better predicted by variation in ability to consistently perform repeated intense efforts, rather than aerobic capacity alone. And that there was a large amount of variation in this ability between athletes. Which also suggests that down-stream PDC isn’t perfectly predicted by FTP alone (at least at this level).
My take away from reading on all of this, everyone’s comments here, and reflecting on it all is that “durability” or whatever you’re going to call it is likely real, but probably only a minor or trivial factor for most riders unless you are pro-level (or perhaps focus on ultra-endurance events). Recreational athletes probably being best served by focusing on improving FTP instead.
That also incorporates personal experience of me having a lower FTP and much worse 40k TT performance than a few friends… but kicking their asses on 4-8h rides.
This. Especially for those with less (competitive) experience.
You are apparently misreading the graph. The correlation between (model-derived) FTP and power is >0.9 for all durations from 2.5 min out to not quite 6 h.
ETA: That might be my fault. The Y axis of the last graph I posted is the correlation coefficient, not the CV as indicated (forgot to update it from the prior chart).
You’re missing the point. Just because you can’t detect differences in ability between individuals at shorter durations doesn’t mean that other factors are necessarily important at longer durations.
As for the studies you mention, #sportsscienceatitsfinest
Sorry, but I’m not watching a video…
As for fatigue, as I tell my students, it is always multi-factorial.
1000 kJ isn’t a lot really. I usually don’t start intervals until I’m at minimum 600-800 kJ in, often 1000-1200. The other day I did an interval block starting around 2000 kJ. I’ve also snooped some pros on strava, and saw them doing intervals deep (3000+ kJ) into rides. I’ve seen them do variations where they also spread out 10-20’ intervals evenly on a 3-5 hour ride. I’m not sure if there are any additional benefits over doing them all at once, or at the beginning of a ride vs at the end. The only thing I would question for us amateurs is quality. If you start intervals at 2000 kJ and you fail to hit your target power or time goals, it probably would’ve been better to do them earlier in the ride. But if you are nailing the intervals without issue, then who’s to say it’s a bad idea? You’ll certainly be doing those whether you want to or not come race day.