AT: Easy vs Moderate vs Hard vs ... [Pass / Success Survey Responses]

@mcneese.chad , as usual great observations and nice work jumping in the the chart. Thank you!

I was one of the posters a while back asking for FAQs, users’ guide, etc from TR. A beta would be a very good time to leverage participants in providing input and feedback so it’s ready to go when AT goes into production.

And yes, the survey ratings seem to be a key factor in helping people get faster quicker and more consistently so should have appropriate attention.

2 Likes

I think the problem is they do not have a “best practices” yet. I think they have way more issues to figure out than they anticipated and one of them is communication around the surveys.

2 Likes

I agree.

I, and I think plenty of other beta users, including yourself, have been responding to the survey in accordance with the simple, specific question being asked, which is:

“How did this effort feel?”

The question isn’t “how did it feel compared to what you expected” or any other variation that’s possible or been discussed, it’s: “How did this effort feel?”. There’s no wiggle room there - it’s a simple question and I think many of us have been answering (correctly :wink: ) in accordance with the straightforward question being asked.

It’s enlightening that a few people have posted up some of the feedback they’ve received from TR support staff, because if those support staff are correct (and perhaps they’re not on script and have made an error?), it’s implying that the question we’re currently being asked in the survey is perhaps nonsense and needs to be re-worded. And, it feels a little like the rug has been pulled from under us.

The question wording is crucially important so that we know what question to answer - doh! - and not have to attempt to game our responses in order to engineer an outcome that we think is correct. Pretty fundamental stuff…

Responding to the survey should be straightforward without requiring any over-thinking, second-guessing or gaming. The wording needs to be aligned with how TR interpret our responses and it needs to be clear, otherwise our responses are just feeding back garbage.

Hopefully this issue will be clarified, and if TR really is asking us something different to the (current) simple “How did this effort feel?” then they need to (a) inform us of that and (b) they need to change the wording of the question so that we know what we’re being asked. This is - or should be - basic stuff.

13 Likes

Great post!!! :+1:

I added the actual survey question into my Google sheet, because I think stepping back to that reality is important!

image

1 Like

I don’t think I know my cycle-self enough to know how I expected it to feel. I always answer in respect to RPE. I hate RPE.
Imagine asking someone new to exercise “How did you expect it to feel?” Huh???

2 Likes

Hence the reason I made a bit of a joke,

  • image

and another one,

  • image
31 Likes

:joy::rofl::+1:

Looking for some clarification from the team on if/how we’d like to best provide an update here, stay tuned!

21 Likes

There are clearly problems here.

AT is a system that says “If this [workout + survey response}, then that [particular kind of adaptation.” But if we game the system to get the result that we want [eg, we mark a 2-hour threshold workout as “easy”] because we want a faster progression, then we are doing what AT was meant to be doing. What will AT’s response to this kind of behaviour be? Can it react sensibly?

What we don’t know is whether AT’s adaptations are “linear”. For example, if we mark an Achievable workout as very hard, then we are likely to see a reduction in out Progression Level for that energy system. We’ve heard examples of this. But if we mark a Stretch workout as hard/very hard, would we see a similar reduction in PLs? Our examples do not indicate an example to this question.

FWIW, I follow @mcneese.chad’s scale. Comparing the workout to my expectation is contrary to the question asked [thanks, @AldridgePrior for this clarification] and could only really be meaningful if the question was posed as: “This workout was meant to be easy [or moderate or hard] for you; was it easier/harder than or the same as that?”

4 Likes

Yeah this is key. Assuming the recent responses from TR Support are accurate, and Moderate actually means as expected, it sounds like this is the question AT thinks it’s asking. If we all answer the question that’s actually displayed, AT’s not going to function properly for us even though we’re Technically Correct™.

Also I love your wording. I can SEE IT NOW

12 Likes

You took the words out of my mouth, this is exactly how I feel about this. And why I’ve been frustrated with the lack of and/or the proposed adaptations based on my survey responses. I have yet to give anything besides Z2 and recovery rides a Moderate, and have not given anything an Easy. Most of my rides have been Hard or Very Hard, because that’s how they felt to me. It would never even cross my mind that I should rate a VO2 max workout as Moderate. So now I’m getting downgraded adaptations and I think I’m going crazy because I’m hitting my workouts but still getting adaptations to decrease the workout levels.

So I either trust AT and pick a lower workout (sometimes by a full point or more), which would still probably get a Hard or Very Hard rating if it’s VO2, and I create a spiral of never getting harder adaptations because I never rate Moderate. Or I continue to decline adaptations and progress based on workout levels, and keep rating Hard or Very Hard. But then I’m in the situation where I never accept adaptations, and AT isn’t doing anything for me and I’m just picking my own workouts.

Either way it’s frustrating. And it seems that there hasn’t been a specific answer.

3 Likes

Ivy’s working on it :slight_smile:

1 Like

According to this blog post (How to Use Adaptive Training - TrainerRoad Blog), it should be RPE based and not how it was expected to go.

“You will see different surveys based on how your effort in that particular workout went. The most common survey you will see is the one pictured above, which asks you to rate how the effort felt on a scale from one to five. When answering the survey, consider the entire workout, and don’t worry about how you expected it to feel based on workout type.”

It literally says to base the survey on how it felt, NOT how we expected it to feel. To me, the TR support answer is the complete opposite of this blog post.

7 Likes

Actually in the situation that I posted the Customer Service response, the situation was close to the second. I had performed a stretch workout successfully, but rated it as very hard. it slightly downgraded my adaptations for the following weeks.

When I take a step back from being frustrated about my training potentially being less effective than I had anticipated based on the TR promotion of all the wonder of AT, and realize that beta is still fraught with issues, then this discussion should do exactly what a software company would want to get out of beta. IMHO that would be that a gap/issue has been clearly identified so the company can determine what should happen, resolve the issues, and deploy a solution.

So let’s see how TR does acknowledging the issue and getting clearer direction to those of us in the beta so we can both a) experience the benefits being touted and b) find different things that TR could/should address in their continued progress toward general release.

8 Likes

Great picture! Exactly.

I know this suggestion is essentially “wrong” and not what TR wants, but I feel the underlying effect of what I wrote is what it seems they are actually doing with our survey responses. Look at the middle column with blue text, as my guess to the actual results when we pick between the current options.

I am sure this is not entirely accurate, but it seems more aligned with what we are learning from TR support. Despite the apparent aim to make this survey individual and adaptive, it looks like they are just applying simple rules that set a defined outcome from each selection. And IMO, the apparent outcome doesn’t match with the broader interpretation of the loose & vague terms TR applied in the “feeling” survey.

I get what they were trying to do, but I think there is a very big miss with the current implementation.

3 Likes

Yeah, I never would have guessed that I was supposed to be comparing to how the workout was “supposed” to be, which is also kind of a hard thing to know. My own RPE is much more straightforward and easier to judge.

5 Likes

This has been a big reason I have actually recently stopped using AT. I’ve been so frustrated with trying to figure it out that I feel my actual training has suffered. Going back to just using the set plans I feel a lot better so far. Maybe when they fix all the problems for the main release I’ll come back.

1 Like

Even if I rely on my own RPE, that’s a very imprecise measure. Consider a 2 min VO2Max, 2 min recover repeated 10 times. 2 min at VO2Max certainly qualifies as a hard effort, but with plenty of recovery and only 10 repetitions, I’m still feeling pretty fresh after the workout. So should the workout be easy/moderate because it was easy to finish and I feel fine afterwards, or should it be hard because the 20 min of VO2Max were right at the edge of what I could do for 2 min at a time?

How about a ramp test? By definition, I’m going til I can’t go any more, but 5 minutes after I finish the ramp, I’m feeling fresh and can take an easy/moderate ride with no difficulty. How do I rate that workout?