This morning I completed Loma Alta (90min sweetspot 8.1) and exceeded the interval power levels by a few %. Felt really good and appropriately hard as a SS workout. I wasn’t dying and probably could have gone harder but also its a SS workout so I thought hard but sustainable was right. So I marked it “hard” on the post-survey, but AI then comes up with a reccomended adaption downgrading next weeks SS workout?
The training plan had Wright Peak (90min SS 8.6) and I believe the suggestion was to do Westby (90min SS 8.2) I was looking forward to the higher level workout and believe I can complete it so I skipped the adaptation but it made me wonder why was the suggestion made? Certainly AI can see the results on todays workout and the PL increased. Should I have marked todays workout as “moderate” instead? I’m 5 months from my A event (210 mile gravel), so want to keep pushing gains on long sustained power now.
I understand doing this, but in the current iteration of AT, it does not properly see these “extra” efforts. As such you can potentially overdo the workout and give a harder rating than what you might have given if you had done the prescribed workout without any deviation, AT might think the prescribed workout was harder than necessary.
If this all happens, AT might kick down pending workouts thinking you need a lower level workout to progress. All a bit of speculation by me since we don’t know the specifics of how and when AT does it’s work, but per the article below I think it is best NOT to over-perform in workouts. Or, at the very least don’t expect to be rewarded for it… it may even be a “penalty” of sorts as could be seen in your case?
All that in mind, if a person plans to do more work than what is on the schedule, I suggest using the Alternates to pick an appropriately harder (or otherwise different) workout that suits your plan and expectation.
Because not every single workout is supposed to be at your limit and the AI could have been looking to give you something that was more moderate than hard during that day. Sometimes it’s about a minimal effective dose.
Since completing Loma Alta, your Sweet Spot Progression Level is 8.1.
Adaptive Training was suggesting that you knock out a level 8.2 (Westby) workout before you get to an 8.6 such as Wright Peak -6.
Let me know if this makes sense. Since you declined the adaptations, you could certainly attempt Wright Peak -6 next week if you’d like. If Loma Alta felt good for a Sweet Spot workout, you might be ready!
I should clarify: Planbuilder had put Wright Peak in the calendar for next week. Yes, it is labeled as “stretch”, but it’s a bit confusing for Planbuilder to lay this out and then have Adaptive Training come in and change that after passing a workout? If the original plan had been a “minimum dose” workout next week, I would have expected to see an “achievable” level workout next.
Uh, wait… The support doc you linked says the opposite?
Yes, looking forward in the calendar, there are some spots I am thinking of picking alternatives, but in going harder I tend to want to pick things with the approach of easing up to the point of failure rather than going too far and having to dial it back after failing a workout. Today, the workout looked good to begin with, but my legs told me to go a bit more so I did
To clarify on the survey response, I am under the belief that “Hard” being the middle option should be interpreted as “The correct level of hard”? While to my mind responding “Moderate” to a productive workout means “Please make it more harder next time”
Interesting… I could have sworn it was different in the past and didn’t claim to recognize the deltas. That or I have taken the other related comment further than the text intends.
Note: Changing workout intensity or going harder/easier than what your workout prescribes will not change the Workout Level for a workout.
Or this one as well, but the context is different than I was personally thinking when I shared the info above.
Significantly higher/lower power compared to the prescribed target power of a workout may move the workout outside its originally intended power zone. Adaptive Training does not currently recognize when this occurs, but we plan to address this with future product updates.
Reading this all again makes me a bit confused if I’m honest.
Massive can of worms here that has been covered at various times over the years. I’d have to find TR’s official wording, but they typically say to keep it simple and rate it how it feels (regardless of expectation).
That answer is insufficient for some, and lead to much discussion not to mention my own creation of an expanded list based upon info shared from TR:
From what I understand, Plan Builder may put a workout in when the plan was “built”, but as you get closer to pending workouts, AT will fine-tune the original plan.
Right now I am staring at a 4.5 VO2 workout on Monday but my current Vo2 PL sits at 2.5. I’m guessing that’s going to change downward here in a couple of days.
really appreciate that chart. In my mind I put easy as, nothing could be easier, and all out meaning I failed at some point to complete everything.
I find it hard at times to rate things between easy and moderate, and the context is something I’m always questioning. Like. Is a recovery ride that is up to expectation easy? or is it moderate? Is an endurance ride that is up to expectation, moderate, or by that chart easy? The implication of how the system is going to interpret my answers does weigh on my mind. Does it have an expectation of how easy or hard it should be? Is an achievable ride meant to be moderate? Is a breakthrough endurance ride marked as moderate going to trigger a big uplift, or just…status quo? The combinations of ride types and intensities makes it challenging.
Speaking of expectations. It might be nice to have the AI give what it expected me to say (the status quo) when I am rating it. Or…better yet give an indication of what the uplift might be to levels as I mouse over the answers. :D. I guess I could always go back and change my answer after I see the outcome.
I recall the working harder thing being discussed shortly after adaptative training was released, blanking on the specifics but my understanding was if you say bumped it up 5% or knocked it down 5%, or just pedaled less/more in resistance then it would factor that in. That % was not fixed though, I’m just using random numbers. If you went too far up or down since that would now be outside of the target for a workout of a certain workout level then it could only do so much when factoring that in so like a 5% and a 10% may be treated the same as far as the algorithm is concerned. Going well below say 10% though could trigger a fail so it does seem to acknowledge that in that direction at least.
WL never changed (no one seems to be contesting) if you did a level 8 at 10% over it still counted as an 8 even if it maybe was equal to an 8.5, but there was some sort of window that it used in conjunction with your post workout survey that factored into the future adaptations. If you did 5% over it may be able to factor that in and if your next planned work out was an 8.2 it would give you an 8.3 instead, but going 10% over might also just give you an 8.3 since you’ve maxed out what it wants to consider for its adaptations.
Wish I could remember because I recall it was discussed in detail when it was first released for public beta as much of the wording was even more confusing than the current description linked above.
Repeating what I said above, TR really stresses to ignore expectation. It should not factor into your survey selection at all from what they say.
To that end, I try to wait on my rating consideration until I hit the cool down at the end of the workout and often until the survey pops up. Take a quick gut check on how good or wrecked I feel and pick from there.
That’s the ideal at least and I do that most days. But I do slip at times and start considering ratings during some workouts despite knowing better. I just try to push off those thoughts and aim to finish the workout first, rate last.
Ditto, I also fall into these thoughts at time so I get it. But per above, it’s best to try and ignore those thoughts and pick the rating based purely on how you feel once the timer hits zero.
And this is one of the catch-22’s that TR stresses related to expectation. From all they have studied on their end, the “Keep It Simple… Feel is all that matters” approach is superior. Those pushing beyond that in various ways seem to have more undesired results according to TR.
Don’t overthink it. Consider the overall difficulty of the ride and not how you expected it to feel.
When answering the survey, consider the entire ride, and don’t worry about how you expected it to feel based on workout type.
How to Rate Your Effort
Easy
This ride felt easy and non-taxing, requiring little effort or focus. You could repeat the ride and pass it without issue.
Moderate
This ride was somewhat comfortable but required some focus to complete. You felt a little challenged but had confidence that you could finish. If the ride had an additional set of intervals, you could complete it.
Hard
This ride required effort and focus and was challenging to complete. This will feel tough and you’ll look forward to this ride ending. If there was an additional interval, you could have done it with some focus.
Very Hard
This ride was very difficult to complete. This ride tested you. If there would have been one more interval, you wouldn’t have been able to do it.
All Out
This ride was extremely difficult. It pushed me well beyond my abilities and took a massive amount of energy and focus to complete. You’ll feel like you barely made it to the end of this ride, and that you had to pull out every mental trick in the book to finish.
This may have what you are looking for, since this was the far reaching & long running discussion about surveys. Much of the items I hit relate to the survey info as well as my own experience over-performing and not getting expected results back then (expected PL to be higher than prescribed, but wasn’t).
I think it must in the dumbest sense that it is just a math equation somewhere there is an expectation predetermined as to what to do with the outcome of that equation. So if a 2 a(achievable) is an x (easy) then give him a 3 p (productive) next, if a 3 p is a Z(hard) give him a 2 next. Obviously there is nuance like how it has to factor in working slightly harder/lower intensities, and that nuance is why they will never tell us since that is the secret sauce and we’ll all be anxious every time we click the button on the survey.
Yeah. I do think that knowing they expect me to just say what I feel is important, and giving me a guide of what those feelings might ‘sound like’, that’s excellent. I’m glad I saw this and the descriptions. I won’t overthink it, but I am currently driving the AI completely nuts to trying to shape it, while still trying to honor the intention of the program. It’s got an extremely casual approach currently (didn’t use TR in 2023). Most of my workouts when I’m done I feel like I didn’t actually do anything. /shrug. I was sort of trending toward interpreting intention, now I know just to rate it on feel. thanks @mcneese.chad!
