Adaptive Training Closed Beta Update

I understand it was a joke, but there are thousands of posts here and trying to differentiate between legitimate comments/defects is only made more challenging when we create confusion about whether something is an actual gap or not.

1 Like

That’s my method too. Like adjusting the cruise control up and down!

1 Like

I made a legitimate suggestion trying to help you. I apologize for making a joke that somehow offended you, but let’s not pretend it created any confusion. I’ll stop responding.

Speaking as a behavioural scientist: I feel like TR needs to hire more behavioural scientists. A lot of how AT is going to work depends on users trusting the plan modifications they are getting and not trying to game the system in order to make subsequent workouts too easy or too hard. If users know what the questionnaires are going to do to their training levels, it’s going to bias how they respond to them.

Of course human coaches do this stuff all the time. The devil is getting an AI system to the point where users understand what’s happening and why, and are willing to go along for the ride. I don’t think TR is quite ‘there’ yet.

9 Likes

Bingo! 100% agree. I was responding how I thought it should be, then was changing my responses based on what I thought TR and AT wanted me to base it on. But then I realized that I was probably causing problems with the system. But now I have no idea. Which then led me to give up on AT for now and wait for it to get fixed.

Edit: I don’t want to be able to game the system. But I also don’t want to be basing my ratings on something that’s completely off. So I don’t want to be changing my answers to get the result I want. But at the same time, if my training plan gets downgraded again and again because I keep rating the VO2 max as Hard or Very Hard, that’s also a problem. I’ll never get productive VO2 max workouts; or at least I won’t be getting the most effective if they just get easier and easier. It’s a tough problem to address.

1 Like

I have stuck to a 3 point system that isnt really different then what you are indicating.

1 easy
2 moderate…answer always if I completed the workout with no breaks or adjustments.
3 hard…completed the workout but noticed that if we keep increasing levels then I more then likely wouldnt be able to do the next one.

I ignore 4 and 5. 5 would be I failed.

My own experience when I failed one…the type of response you give determines what happens. So if you say the intensity was too much…then future adjustments are made vs if you say you were tired/fatigue then no adjustments are made.

We do need guidance on what are answers are to mean. I wouldnt even mind understanding the progression rate. What I have noticed from my own use is short power build was getting my VO2 workouts to be at a level that was definitely challenging. The debate I had going forward was how do I keep going forward. Some I know I couldnt do before even starting…so I would pick an alternate at a lower level…still a higher level then I had before but the one I knew I couldnt do would keep coming back. Maybe it is just me but I know what I need based on my prior experience with TR. As it was I crashed and burned on the energy front after getting the 2nd covid vaccine shot so training stopped following the plan.

Pushing forward I would normally of done the workout that I thought was above me but would then play with the intensity to get the power to a level that made sense. So before PL I would do a workout…and say there were 5 sets…the first 3 would be 100% but the last two would adjust downward by a couple of percent. I now dont even consider touching the intensity as I assume it will just be a failure and I am better to pick an easier workout. Guidance would really help as I am making assumptions with no basis.

The PLs overall still dont feel quite right. So if we start at a level 1.0 which we know makes no sense in a lot of cases…does the workout we do that adjusts the PLs impact how successful we are going to be. So lets say it is sweet spot and I take it to a level 5.0 and ramp up from there. Do I hit a plateau that I cant do or get worn out vs Doing a level 3.0 and let it ramp up even if I do easy workouts for a couple of weeks. Yes it is a beta…but this is the part that I do think doesnt have me feeling like I can trust the system

3 Likes
  • Bingo. I mentioned this a while ago. No idea the trigger point, but dropping Workout Intensity past some point will likely trigger a Struggle survey. This is a “new” problem if we consider the countless times TR has said to “just drop the intensity” for things like VO2 workouts and the like. It has been mentioned on the podcast a ton of times, and some of the workouts even mention doing it specifically.

  • But like you, I have taken to not touching it since I have no idea how low is too low. In a sense, we are forced to predict any issues and make a workout substitution (via Alternates) prior to doing a workout, so we can do that workout with no adjustments. If we follow the old advice, it may well lead to misinterpretations if we are out of alignment with the scheduled workout for any reason.

3 Likes

Yup. They may as well replace the post-ride survey thing with a big “lever” control where we just manually position it according to how (informed by how our workout felt) we want our training plan glide path for that energy system to adjust:

Engine_Order_Telegraph

  • Too easy - give me moar! :facepunch: ㅤ [faster progression rate desired]
  • On track :+1: ㅤ [maintain progression rate]
  • Tough gig :grimacing: ㅤ [slower progression rate desired]
  • Bit of a 'mare :pleading_face: [decrease PL(s) & slower progression rate desired]

Kinda similar to @mcneese.chad’s blue text column in the other thread…

:wink:

7 Likes

And my concern is that the adaptations recommended are actually taking my training in the wrong direction compared to not using AT at all. Add to that the time/energy to try to sort this stuff out (heck, just read the posts on the AT-related threads!) my beta participation may be personally counterproductive. (And yes, it is a personal choice on which path to follow, particularly now that I feel more informed about AT limitations than I had inferred from TR messaging.)

Not sure about hiring or how to bring that perspective to the table, however, your observation is spot on.

And just edited my post after reading this:

Wow, hadn’t even thought of that but actually quite likely. Now I have to think really hard – and TR may have a big problem in that the way people are evolving to using AT surveys may send ML way off track and degrading what it’s doing vs improving what it’s doing… again, wow…

2 Likes

I’ve opted back out of AT for the same reason. I felt like it was actually pushing me backwards compared to just using a plan. I’m in a recovery week now but next week start a Short Power Build and CX specialty that I’ll be using the plans off AT. Less to think/worry about. I’ll let TR get the kinks out before I’m back.

1 Like

…that would be OK if we ever actually had any guidance on how to complete this in the first place?

I haven’t yet seen any single place with definitive answers to questions like this. AT FAQ anyone? Just another reason this whole ‘beta’ experience feels a million miles from being ready for customer use IMHO

3 Likes

I had no idea about the ratings… suddenly Adaptive Training giving stupid suggestions (IMHO) makes a ton of sense.

Unrelated, I assume that a cut short workout is useless from an adaptation point of view? I stopped a workout about two thirds in because, well, I couldn’t be arsed. Actually set some (post injury) power records, but obviously the system still thinks I’m a filthy failure and makes no changes right?

Personally I’m starting to wonder if PLs are causing athletes to focus in the wrong direction. The forum threads make me think AT/PL is training gamification and for some its all about leveling up.

5 Likes

The “first place” was covered in a many posts by TR reps in the AT debut thread, as well as the following blog post that hits these specific points:

The issue here isn’t that we didn’t get initial guidance. The issue is that the initial guidance has been contradicted by multiple TR reps via the support channels with individual beta testers when they asked for clarity on the issue.

6 Likes

What do you see if you open that workout on the web?

Did you receive any adjustments to your PL’s, or is it something like the “+0.0” solution?

That page may give you an indirect answer to your question.

very good point…

I like the idea of PL’s as they have made searching for appropriate workouts much easier when not using a plan. BUT your point is very valid as I already look at numbers and start to think “I want a 4.x or 5.x because 3.x doesnt sound that good…”

I also think you’re right in potentially taking people in the worng direction. It would be easy to look at your PLs and do training to get 7’s across the board or 8’s by next month etc, without thinking about why and when you might want to work particular focus areas to achieve your goals. “Oh, my tempo is only 2.3 so I ought to bring that up…” even though spending time working on tempo might be completely counter productive to your goals and performance.

Hmmmm…

2 Likes

Another captain obvious post… AT is in a long beta, clearly the thinking at TR on post-workout surveys is evolving and not static.

I get your point here, but in the beginning, getting the levels correct (or at least close) is essential. Otherwise every workout you do is overly easy. If you start with a bunch of Level 1.0 or even 3.0 ratings, but the last training block you did with the exact same ftp had a bunch of rides that were more like a 5.0, then it literally takes months to get back to those correct levels when you mark a ride as “Moderate” and that only moves you up .2 per ride.

1 Like

Maybe, in the sense that we now have 8 metrics to worry about (FTP + 7 PL’s) :stuck_out_tongue:

It’s possible some will game to kick them up. But from what I see, many of us just want the function that TR promised, where we do a workout & answer a simple survey, and let AT guide the way.

That of course takes good data (lacking from the stuff AT ignores at the moment), good rider feedback (clouded with all the uncertainty of current surveys), and our trust in the system to do the right thing (see points 1 & 2).

3 Likes

Yep - that table of descriptions is also completely useless. All it says is rate it “how it actually felt” and then we’re back to all the comments made in hundreds of posts above and elsewhere…

Nowhere does it say ‘Hard = these feelings’, ‘Very hard = these different feelings’ etc

1 Like