Just noticed the new feature shaming me for stretching for 15 seconds during the recovery valleys. Great, something new to be obsesessed over.
16 Likes
I got the same thing today. Took a restroom break during a recovery block 60 minutes into the 90 minute over/under workout, which ended up being 3 minutes after my wife stopped me to ask a question on my way back to the bike. TR thought I would rate it as a hard workout beforehand, I did mark it as hard but said I didnât struggle, and TR dropped my AI FTP prediction by 4 watts. I assume it will normalize out over the next several workouts, but thereâs a tiny part of me that will blame my wife if I donât get those 4 watts back.
19 Likes
Yea, this going to drive me crazy. Stop for a bathroom break? TRFTP down 4 watts. Ride an extra 30 mins at the end of a workout? TRFTP up 2 watts.
I wish there was some way to hide this. I donât need to know minute-by-minute what my predicted TRFTP is.
10 Likes
You can turn it off! Go to your calendar (maybe on the web only?), find your future prediction, click on it, find the edit/pen in the lower left corner, and deselect AiFTP prediction. Then you wonât have to deal with this!
you can do it on the app too!
6 Likes
What happens if you rewind the workout to just before you took the break?
1 Like
Why are so many people surprised to see that modifying a workout changes the outcome of the workout? I would never think I could just take time off and it would have zero impact, no matter how short. If youâre cooking and you follow a recipe but modify one of the ingredients, you donât say you followed the recipe 100%. I think we need to get past feeling like getting a 99 on a test is a disappointment.
27 Likes
What happens if you extend your recovery by more than the time you stopped?
This is equivalent to missing 5 grains of sugar in your sugar cookie recipe. I think my HR is often higher walking to the bathroom and trying not to slip and fall in road cleats than during a rest interval at 40% ftp anyways. Maybe if I run there and back my score will go up?
Or should we all become triathletes and just pee on the bike while riding.
8 Likes
Firstly⌠the PL scoring at the end of workouts is a completely separate system than the AI FTP and AI workout selection.
The AI just looks at raw watts - the levels are a visual aid.
You will become and less bothered about these little workout scoring adjustments when you realised how unimportant they are now.
SecondlyâŚ. the predicted AIFTP shifts around a lot even if you perform as expected on the workouts. It becomes more stable the closer you get to the prediction day but the normal variation on it is too great to pick up the fact you stopped 15s to stretch.
9 Likes
I got told yesterday that my 1:29 bathroom break during a 12 minute recovery valley lowered my score. But looking at the summary afterwards I see that I got the exact same score that the workout has. I suspect I got a slight bump for extending the cool down, which cancelled out the bathroom break, but when Iâm only told about the latter it can get a bit confusing.
Apologies if you already know thisâŚ. but after the workout is completed the workout score on your calender and workout information tab gets adjusted to show the actual work done not the level of the original workout. So these levels will always match anyway.
Also possible that the extra work has balanced it out to go back to the original score too though.
2 Likes
Yeah, I was talking about the workout summary where it says what WL score you got compared to what the workout was supposed to give.
For my example it says 3.9/3.9 for threshold and 4.5/4.5 for sweetspot. But I did get 0.1 extra for endurance and only 0.8/2.2 for VO2Max. (It was an over under workout with the overs in VO2Max territory.)
I suspect I got extra to most zones because of the extended cool down, which might have been negated by the break, but also might have been because the AI seemingly doesnât like me doing the overs in ERG mode, which resulted in average power being a couple of watts lower then it should. Who really knows?
The part that feels a bit unnecessary to me is that of the many reasons that your WLs might be a bit different, only one is communicated.
1 Like
We should be happy we got a 99 on the test
5 Likes
Thanks - I turned it off this morning. I would have thought it would been in settings, but at least itâs available.
Thanks, as someone who is currently only scheduling a few workouts in advance, this was very annoying!
1 Like
When you say âshifts around a lotâ, how much is a lot?
Iâve not started a plan yet with the new AIFTP prediction in play, but when I do, a prediction that isnât stable would lose all of itâs value to me - ie, a motivational aid to encourage me to complete the prescribed plan to gain the prescribed ârewardâ.
The more I see of this new prediction feature, the more Iâm inclined to turn it off and revert back to manually triggering a new FTP detection at the end of each recovery week.
1 Like
Plus minus 5w maybe 
Wow, Iâm lucky to see 5W increase across a whole 4 week block. If it can shift by 100% during that time, then the prediction would become useless. Iâm curious to see how this plays out for me in my next block.
I have gotten used to increasing 2-3 watts per block. During the whole winter training max 20. Now my prediction is a 20 watt increase this block, after already having had most of my âusualâ increase. Som, you might be in for a surprise.
1 Like
If it was during the work interval sure, but 15 seconds during the recovery is nothing.
Thatâs like saying that instead of using 10mL of oil you use 10.5mL. That doesnât change the recipe in the slightest.
More like getting a 99 and TR rounding it down to 90.
3 Likes