Random thought about Adaptive Training and threshold workouts

As it relates to FTP change recommendations by Adaptive Training (which has been discussed in other threads), I had a thought yesterday after completing a 4.0 threshold workout. It looks like threshold workouts continue to go up in intensity/difficulty each week in my current plan, which got me thinking: Is there a “tipping point” in which one can complete a certain threshold workout (say 5.5 or above), and Adaptive Training would recommend/discuss an icrease in FTP?

I understand that FTP is much more complicated than anything I could ever imagine to incorporate into a single post here, but am curious if anyone else has thought like I have above^^.

edit: I have no problem at all doing FTP testing periodically. Just curious about Adaptive Training algorithms.

They’ve talked about this on the podcast being a long term goal for AT. That you never have to do a ramp test again. You keep progressing through levels, you get up to accomplishing 6-8 threshold workouts without coming undone and it recommends increasing your FTP and then drops you back to level 3’s or something.

I’m not in AT, but as far as I know that’s not built in yet. Could probably do it yourself in the meantime…simply knowing that of you’re knocking out higher level workouts week after week, it’s probably time for a 5-10 watt ftp bump, then drop the level back down and see how it goes.

1 Like

I think this would take some work…

The PL rating for threshold sessions could be influenced by several factors - intensity (< or > FTP?, rest durations (or lack of), over-unders, length of intervals etc, and not all of these might mean an increase in FTP is warranted.

The length of intervals is an iteresting one, as its entirely possible you push out your ability to hold threshold intervals from perhaps 10 to 40+ mins but that doesn’t necessarily mean your FTP has actually changed. I’m sure there are other factors that influence the PL that might also not really relate to your actual FTP as well.


Agree completely. Perhaps the algorithm would take into account IF. like maybe has to be .88 or higher.

At the present, if you manually adjust your FTP in AT, it’ll reset all your progression levels to 1.0. Not a problem because you can simply go into the workout library (or your workout history) and choose a workout that you think you can complete at your new FTP. Nail it, and that workout’s level becomes your new progression level in its respective energy system.

I think TrainerRoad’s implementation of AT is pretty clever. It takes into account the limitations of today’s hardware and does its best to provide the best usable platform until hardware technology improves.

I predict that in some number of years (could be three, could be 10+), FTP will be eliminated. Instead of using one anchor point, the software will instead use many anchor points, i.e. your entire power curve.

Also, the current workouts are all static in nature. That’s why TR added hundreds of new workouts recently to fill in the progression gaps. Again, a clever way to address today’s limitations, but I’d be surprised if the workouts don’t become dynamic in the future, where AT makes microchanges to the workout. Adding watts or seconds here, subtracting watts or seconds there, etc.

Hardware limitations are an issue today. 1 second power and heart rate data is too coarse. I think a much higher sampling rate like 1/100 or 1/1000 will make a big difference in capturing certain nuances to better predict whether a rider is struggling slightly or kicking ass.

1 Like

Awesome! Thanks for your thoughts and insight. Definitely things I haven’t considered.