For every level in TR there are two for Zwift. You could break it down that way and if you are stuck between levels pick the higher (ie. Easy starts at 1, moderate at 3, hard at 5, very hard at 7 and all out at 9). If you have a hard time or feel between levels then pick the next number in between.
Or you could do what I do and dismiss the Zwift survey and just do it later in TrainerRoad
Same 
Sounds very interesting. Once it is live will we get a comprehensive breakdown of how this works and how to use it. Right now these are both a little fuzzy for me. For example I want to increase my FTP to 300. Is that a condition I can define and AI will do the rest? Also can I define timeframes when I want to hit goals? Will AI tell me if my goals are realistic? WIll this also use a real language chat session similar to other AI clients?
these guideline should be linked to the score at the end of a workout. Maybe there are but I have not found them. Seems like a simple feature for the product backlog.
In the past, Jonathan has said that he thinks documenting it over complicates it and he prefers we just answer based on how we feel. I donāt agree, because my perception of āhardā can be different than yours, and it has implications on future workouts the AI assigns, but thatās the answer Jonathan has shared multiple times.
when will we see this on the app?
Been there, done that.
I doubt we will ever see it return despite this coming up repeatedly since introduction. I have no data (like they probably do), but it seems that at least a decent sized minority can and does benefit from a guideline. I understand it may be an issue for others, but I do NOT see why this has to be an all or nothing proposition.
Information like this isnāt inherently good/bad, as it will depend on numerous factors. Based on the repeat sharing and admissions from many that they like it, I still believe this info can serve a useful purpose for quite a few people.
Donāt hold your breath.
Based on the mention of the newer rating within Zwift, I agree with the methods stated above, so I added them to the sheet:
I still refer to this regularly. I often find myself in between ratings. I try not to overthink it but probably do.
Could I repeat the ENTIRE workout? No, but it was easy enough it doesnāt deserve a Moderate rating.
Or if I rate a workout as Hard (I could do one more interval), and a part of me says āYou wanna bet?ā
Overall, I have had pretty good luck getting workouts that feel right from TR AI. Interval days are feeling Hard or Very Hard, and endurance days are feeling easy or moderate (Low volume, Masters).
Isnāt it more likely that athletes will err on the side of doing too much rather than too littleāwith human or AI coaches? Thereās always that temptation to under report effort because we have in our heads this idea of constant up and to the right progression.
Is this a real answer? If so, seems a bit snarky.
Lor
MĆŖme problĆØme de volume pour moi, il y a un truc qui ne fonctionne pas correctement, et pas de solution apportĆ© alors que je payeā¦
It is as real as my guess, based on many years of dealing with TR on this topic (since introduction) and hearing their answer repeated countless time since then. They are not inclined to change from what I know.
No snark intended. Just as about as brief as I can answer the question.
If I have one quibble, itās that āall outā seems like a superlative. You canāt go higher. Rating an effort all out and selecting 9 seems weird. Maybe it should be 7-8-9. I might even argue a zwift 1 doesnāt exist. Haha, I think Iām now overthinking the chart. Lol.
I will leave that to TR to handle, since it was their wording (not mine).
Perhaps. I plan to leave it as is since I think the consequences of slight differences from what I added are likely marginal at best. Besides, itās an unofficial guide that people are free to ignore and/or morph to their own preferences 
okayā¦didnāt realize this has some history. thanks for the update.
Lori
Over 4 years and running. I spent more than my fair share of time & effort in the discussions, making the list above (along with underlying support info available in my shared TR doc) and lobbying TR to do more than they have chosen along the way. Dead horse & brick wall + head sort of process at this point. So despite the short comment above, this goes rather deep.
It wasnāt a critique of your work Chad, hope it didnāt come across that way, Itās more that I think TR should standardize to a 1-10 system personally. Then itāll match with Zwift and Intervals.icu. Just my opinion. Thanks for the effort!
100% , you can add Garmin and Training Peaks to that list as well. Even on a medical scale, youāre asked āHow is the pain on a 1 to 10 scale?ā , Having two different scoring systems (with the Zwift connection) is just plain confusing.
Deleted post
