I think you are misunderstanding what has been said.
ㅤ
Here:
ㅤ
Bolded bit:
I nearly posted something about this on seeing the TR post, just to mention that I thought it represented the first instance of AT adapting “off-plan” workouts.
10 Likes
A central issue here is that energy systems are not isolated and can improve even if we don’t work on them. For example, it’s well known that long endurance rides at the bottom of Z2 lead to an increase of your ftp. If TR doesn’t recognize it in your Threshold progression level you will probably waste your time indoors after riding outside next time you do a “productive” threshold workout as prescribed by trainerroad just because your PL is underestimated. You might argue try a higher PL workout but if it’s up to you deciding it what’s the point of PLs?
Oops. I misspoke then. They’re talking about TR workouts not in your plan. I thought it said custom workouts.
1 Like
Custom workouts will be impacted by RLGL now as well! If you get a red or yellow day that falls on the day of a custom workout, you’ll get suggested adaptations that you can accept or decline.
7 Likes
LOL. Thanks for the clarification. And yay!
2 Likes
I don’t know, but that is my understanding. I haven’t had much time to ride outdoors lately, so maybe I am wrong.
Edit: I was wrong, and I stand corrected. 
I’m not sure you can separate the two, seeing as TR is a training platform.
Even if we ignore adaptations, I assume RLGL does use AT under the hood by trying to link workout outcomes and larger patterns in your training to fatigue. I spitballed how TR might do it, and analyzing workouts and larger patterns in your training seem to be an important differentiator.
1 Like
Now that RLGL is available to everyone, what are the remaining building blocks to be put in place for WLV2 prior to its release? It seems it could be just the changes to PLs which an unstructured ride will generate? RLGL uses WLV2, and Nate mentioned in the last podcast that there are a few things on the horizon…
2 Likes
If I had to guess, then the bits that are crucial to RLGL are working ok, but the other user-facing features do not yet work to TR’s satisfaction. The fact that it underpins RLGL, though, indicates WLv2 is maturing.
Exactly my point. Other than changes to PLs as a result of unstructured rides can anyone think of any other significant building blocks needed? I’m just sitting here full of snot wishing I was ready to restart my training and my mind is wandering.
1 Like
From my perspective, RLGL is basically a stress measuring stick, somewhat similar to stress balance when looking at CTL. Cool feature, but I’m not sure it’s really addressing the hard part of v2 (based on what’s been exposed so far). Not all stress is created equal and more specifically not all stress touches on the same training adaptations (endurance vs. threshold vs. etc…). I suspect the really hard lift with V2 is looking at completely unstructured workouts and trying to pick out things to “give credit for”. As we know, going on a hard group ride or race where your power is all over the place and racking up X minutes of vo2max time in zone (sometimes 5 seconds at a time) does not create the same adaptations as doing that work in structured intervals. I expect one of their hesitations releasing the feature is that everyone is going to be complaining when their 300tss saturday smash fest doesn’t move the training needle that much. Honestly, I would be happy in the initial release if they would only recognize fairly obvious intervals (like intervals.icu) and give credit for that work, but don’t move progression levels for anything else. That seems much less daunting than trying to pull meaning out of truly unstructured riding (which could come later).
2 Likes
Are you not arguing against yourself here? RLGL assesses the totality of your work, and suggests adaptations around that to account for all of your stress, either in structured workouts or free rides. It changes your future workouts, perhaps albeit only in the relatively short term. I agree with your point that people are likely to be irked when their PLs don’t change hugely after en epic ride, but if that is the truth of the matter then they’ll have to suck it up.
I don’t think so. RLGL seems to do a nice job of measuring overall stress. But determining a training adaptation “value” to that stress and putting it into buckets appropriately for different energy systems is much harder. When RLGL recognizes that I’m overly fatigued and recommends not riding or switching to a Z2 workout, I struggle to see how that has anything to do with picking apart an unstructured workout and estimating what training adaptations came out of it. Maybe I’m missing something that RLGL is doing, I just see it recognizing stress limits and telling me to keep pushing forward or not from an overall training perspective (not telling me how my threshold, vo2max, etc. progression took a step forward or back).
When AIFTP detection came out, it felt like a significant step toward v2 (more significant than RLGL in my mind). Based on unstructured rides, AIFTP is able to estimate where I’ve progressed to in Z4. To make v2 work across the board, I believe they need their AI engine to recognize progressions in all the other zones based on unstructured rides. This is all speculation based on a bias I have on how I think the architecture/engine should work. There is a good chance they have have a different and/or better approach vs. how I’m thinking about it.
3 Likes
Hard to fault TR for gravitating towards the Coggan/Allen view of the world based on the timing and how power was starting to become mainstream at the time. But yeah, a 3 zone view would simplify some things if you were starting from ground zero. Always harder to add big new things when the train is moving.
1 Like
If 300 TSS doesn’t move the needle in your training, isn’t that a failure of your model?
No, not all work is equally productive and at some point the stress can make it counterproductive. You can have a 300tss ride that adds a bunch of stress while providing minimal adaptations (ie - a bunch of random work across all your zones) or you can have a 120tss interval session that targets some very specific adaptations.
For me, that’s the difficultly of determining adaptations from unstructured rides. I’m not saying that a 300tss ride isn’t driving any adaptations, but it may not accomplish what someone is looking for and progressions do not necessarily align with how hard/long a ride is.
For what it’s worth, I’m not bagging on the smash-fest 300+ tss group rides, they can be beneficial. I do a 300+ tss ride pretty much every saturday, but I’ll switch from group rides (with efforts all over the place that I have limited control of) to solo efforts (zone and duration discipline) when I need to train seriously and drive specific adaptations. I would expect significantly more benefit/adaptation from a 300tss effort structured for specific goals compared to just going out and riding hard with a group and riding at the whim of the group dynamics.
5 Likes
Yeah, I think I’m mostly with you. If I understood what Coggan was saying on this forum, 300 is more than 120, which is more, so it’s better, assuming your body can handle the stress.
TR has the first problem here, are there 7 zones? Then they have the 2nd one, which you’re getting at, which is IF they are 7, and it’s even possible to target them so precisely, how would you measure an “unstructured” ride?
I hope they crack it, and I hope then we eventually get to see data that shows that RLGL and WLV2 make their users faster.
I don’t really think it’d actually simplify things. You’d still need to distinguish between threshold and VO2max workouts, for example. If you add both in the same bucket for time-in-zone measurements, you’ll end up at something that’s IMHO misleading. 40 minutes at threshold ≠ 40 minutes at VO2max.
Ditto for putting recovery rides and endurance rides in the same bucket. They are not the same thing.
2 Likes
This is just a model, which takes into account that 1 hour at recovery intensity puts much less strain on your body than 1 hour at the upper end of endurance. Ditto for the distinction between threshold and VO2max. IMHO the 7-zone model captures the different workout types much better.
1 Like
4 zones
Recovery
Endurance
Spicy Endurance
Hammering
7 Likes