So ever since the launch of progression levels in 2021 there’s been talk about someday in the future having TR assign workout levels to “unstructured” workouts (anything not associated with a TR workout). This would then allow an athlete’s progression levels to reflect what they’ve accomplished outside.
Is this still in the works? Or has TR decided this isn’t a viable/worthwhile feature?
Obviously given the past complete misses on indicated launch dates, it’s wise not to make any date commitments, but it would at least be nice to know if it’s even being worked on anymore.
5 Likes
It seems to me that something like it is going on now. They’ve taken the Productive, and Achievable labels off the calendar and lately when I’ve done a workout outside I get adaptations later that drop the levels of workouts I have coming up on my calendar. I assume that’s taking into account the fact that I did the workout outside and didn’t perfectly nail it so I didn’t actually achieve the Level that was prescribed. When the weather got a little worse I did a workout inside, so it was perfectly executed, and I later got some adaptations that bumped up the Levels of future workouts. Maybe I’m reading to much into it, but things have definitely been adapting differently these last couple months than they were in previous years.
2 Likes
I got a bunch of big positive adaptations after doing two weeks of nothing but outdoor rides associated with workouts as “Solo Rides”. I brought that up here and the official TR response was that it didn’t happen, because TR doesn’t take into account outdoor rides. But it did happen, so IDK what exactly is going on.
3 Likes
I think both of those examples are just AT adapting the intensity of your workouts based on how much recent training stress you’ve had using RLGL. It’s done that since the launch of RLGL, which was said to be a stepping stone towards WLv2, which is why I am wondering if they will stop there or are still working on fully incorporating unstructured workouts/activities.
2 Likes
In my mind RLGL is complementary. Having a bit of experience in the ML space, I have an educated guess as to why we haven’t seen WLv2: workouts target one specific zone with pinpoint accuracy. Outdoor rides typically contain a broad mix. So you won’t push yourself to the limit in one particular zone. Translated to PLs, you have one workout that is an achievable effort if you view the various zones separately. Put another way, it is quite likely that none of your PL scores are higher than your current PLs.
In ML/statistical terms this leads to a weak signal with a lot of noise. That’s hard to train for.
In contrast, RLGL has an easier problem as all you are interested in is increasing the chances that you are able complete the next important workouts and balance fatigue with your ability to recover.
PS Just to be clear, I still want WLv2! 
2 Likes
I think they are moving away from Progression Levels because it was impossible to give precise PL’s to random rides and the PL’s they would have given were so much lower than what the rider expected it would have been discouraging and no one would have believed it. Instead they are analyzing everything and suggesting the best ride to do next regardless of what the Level is.
4 Likes
And the PL levels for many custom workouts are way out of wack.
1 Like
What I expect TR to do is to focus on dynamic workout generation. The names were cute, but nowadays it is hard to find “the next hardest/easiest version” of a given workout with the same precise structure. So I would expect TR to simply do away with PLs and focus on that. You can then look at an easier/harder slider and adjust the workout depending on how you feel and how much time you have.
3 Likes