When I select âindoor workouts onlyâ, no data shows up. Is this a bug?
Looks like something might be off there. Iâve forwarded to the team.
I suspect it only counts indoor workouts completed using the TrainerRoad app. I did the workout in Zwift instead. I also donât see this TrainerRoad Zwift workout in the activity list when filtering by âTrainerRoad Workouts.â
Hey Sean. Actually Iâve been running a TR plan since December 1 (when I tested at 228 FTP) not just the last two weeks but it keeps getting messed up so I have to keep restarting. So actually since Iâve started the 2026 training season Iâve lost 6 watts. After many email conversations (with Jonathan and others) Iâve decided to throw caution to the wind and let the AI do itâs thing.
Weâve found a bug with the âindoor workouts onlyâ toggle. Weâre going to hide/disable that temporarily while we resolve the issue.
3 Likes
We found out whatâs happening here and have a bug fix going in now. 
The Lunch Run is being included in PRs because it was updated from Running to Cycling then back to Running. The bug is that it wasnât being excluded by the PR system after being updated back to Running.
Thatâs awesome Nick. I think I remember doing that. With Strava popping up the details for multiple activities I accidentally thought my ride was recorded as a run. 
. That must have been a fun one to track down.
This tool is super helpful, thanks! 
If you somehow would be able to incorporate repeatability in those graphs you would have an absolute top winner.
Letâs say I have a 4x10min workoutâŚnow the real useful comparison would not be to my single max 10min power but moreover Iâd like to see my âhighest 40min combinedâ so to speak. So maybe the best 40min combined interval power of all workouts from 8x5min to 1x40min. Is there an elegant mathematical solution for this?
The âIndoor workouts onlyâ bug is now fixed and this option is re-enabled for you to use! 
4 Likes
So NP covers this pretty well already! 
You just would look at at your NP that covers those intervals PLUS REST, which is SO key. We can see an extra 30 seconds of rest in between VO2 max intervals make a HUGE difference in compliance.
Even though these repeats were at 324, my NP for the 43 mins was 276. This is a good way to quantify repeatability. AP doesnât work well for that because a 0 pulls it down so much, but NP makes up for that.
Thank you for this feature. I thought I would not need it, but it had already helped me. The feature greatly helped me realize why the new FTP Detection downgraded the FTP value so much. Looking at the power curve over the last 6 weeks, it is âlooks obviousâ that the old FTP value was way to high.
Could the labels at the bottom be clickable so it can snap to the time clicked on. Appreciate space is a premium especially on mobile devices and adding click events might not work so well, but thought iâd ask.
Can you exclude Failed or âmax effortâ workouts from power record displaying?
I did an FTP O/U today. It showed in power record that I already did same power past 6 weeks.
Overall it felt not hard, but closer to very hard. I started looking carefully and realized that âpast 6 weeksâ was based on a failed workout 
So the fact that it was not failed but very hard now - is a kind of progress
Thanks for this additional feature - looks great.
This is really cool!!
However, I would like to change the default settings (currently, it shows the last 6 weeks, and it defaults to NP power, which is not what I want to see). Will you please allow the graphs to load with your default chosen setting?
You want max efforts to not be included in your max power records??? I think it would be safe to assume most peopleâs max power is achieved by going very close to max effort.
Am wondering if there is a fix planned for this? Asking as itâs still an issue Iâm seeing, and I havenât seen a TR acknowledgement of the issue unlike the other posts in this thread. I see a few users have noted it.
As an aside, I think itâs also indirectly an issue for NP, where it seems like the rolling 30 seconds includes the previous recovery, so understates actual average power. Across a whole workout this is not an issue, but for shorter intervals it is.
I guess if itâs consistent between power records and workout estimates (?) itâs ok for comparison, just that the numbers are not actually reflecting the correct average power
TR team says in their podcasts, that hard workouts which are scheduled by AI should feel âhardâ. Not âvery hardâ or âmax effortâ.
So if you train according to the plan, in reality Power record should display data based on those âhardâ workouts. Not the ones, you failed to complete.
What I expect looking at Poer records - is to see âhardâ workout power done previously in TR to understand that I will be able to complete it at prescribed power. I donât want to see data from powrly calibrated outdoor rides, etc.
Just compare apples to applesâŚ
If you want to look at your power more closely including failed workouts, you can connect intervals.icu and do it there. You can also exclude some workouts for analysis (which you canât do now in TR Power records)
You either did the power or not. It is an objective record of your highest power. As in, the best indicator of performance is performance. To me, that is apples to apples.
I have read many threads here and people have wildly different thoughts on how they score their workouts based on âfeel.â If you want to bring subjectivity of how that power felt, well, then this tool becomes of much less value. You already have a prediction from TR on how todayâs workout should feel. This new tool provides additional context. And you can always do what you did, and look into when you set those power records, and that would give you the perspective you are looking for. But I wouldnât want it to automatically exclude those efforts.
Thatâs a good idea. @eddie Feature request: It would be nice to have a filter button âHardâ which, when activated, limits the power curve to all rides with a maximum RPE of âHardâ.