I get the same with average power graph, at first I thought you were posting the NP chart. I think its due to catching those random spikes. The NP one looks more normal.
Yeah, my NP graph looks pretty reasonable, and seems to be the same as yours.
My power records are kind of messed up. I was using an oval chain ring with a 4iii left crank arm. Turns out that combination can inflate your watts by 5-10% when you’re riding outdoors.
It was great to find that out, because I thought I was incredibly un-aerodynamic or that my old TT frame was soaking up a lot of power.
How could I go about fixing this mess? I’ve thought about deleting rides that are inflating my Power Record on a one by one basis until I get rides that I know are more accurate.
Yeah interesting - looks like that “sweet spot with spikes” profile gets really messy in the NP chart…
I’m very interested in this too. I’ve about 2 months of rides to clean up, where power data is … too high. I’ve since gotten rid of that power meter. TR offers a way to exclude individual rides from power records, so do that, and then ride for six weeks to “reset the AI?”
In my opinion, this feature is more helpful than AI FTP detection and workout selection to understand the two latter features.
This should have been released first, and then explaining the the huge shift to AI FTP would have made more sense. Explaining how the AI just looks at watts and doesn’t consider “FTP” would have been as simple as “look at the power match and see."
When I was beta testing, the only way I was able to make sense of how the FTP detection made sense was to use the season match and see how I was progressing through the workouts. That’s what kept me excited about the new developments.
All this to say, I love this feature as it clarifies the other major features that have been released. Thanks TrainerRoad Team
What is that pin icon for? Seems to do nothing on my iPad.
@nick I had thought about a feature request… how about having a +1% and -1% intensity button for the workout (just like in the workout player when I adjust the intensity)?
So that I could manipulate the workout curve and live see the changes in comparison to my last six week curve and maybe also see how changing the intensity of the workout would change the estimated RPE.
Edit: My example would be the workout from the screenshot Brasted. It’s 30/15s on/off at 120% whereas I mostly raise the intensity so that the 30s are at 130% and would like to see that (unfortunately there are no 30/15s at 130% in the library)
It just indicates that you have the data field pinned to the selected duration instead of it following your cursor. In the mobile app, it disappears when your swipe with your finger and it reappears when you lift your finger. In a web browser, it only appears when you click on a point in the graph and you unpin by clicking again.
Based on the wording, I believe I found a bug. I just finished a ride that set some yearly power PRs, but it seems like the “6 weeks before” setting is actually including this ride. My interpretation of that, is that the ride should be excluded so that I can easily see where (and by how much) I exceeded the previous best(s). Love the feature though!
When you look at the workout beforehand you can see the difference, once you’ve completed the workout its part of your 6 week PBs.
I’d say regardless of FTP. The 6 week power curve makes evaluating workouts much easier than evaluating percentages of FTP.
Average Power is still not being calculated correctly. ?
I am doing over/unders today with the Over being 1min at 338w - when I go to ‘this workout’ blue line at the 57sec mark… it highlights the right section, my 1min over planned, but says 330w
See issue on my 30/20’s coming up Gendarme+3 ,(using Spring tab)
Don’t see issue on Heseman endurance workout.
This is just one example, usually more than this but depends on adjacent power or magnitude of the power. Oftens most on <4min range. Seems it is calculating often AvgPower very similar to NP.
I still have that bug. I can’t believe this hasn’t been fixed yet, it doesn’t seem complicated at all to fix. I get having complicated problems with machine learning/AI, but this can be calculated in excel. Honestly it’s making me lose some faith in TR. How has this been out for a month now and hasn’t been fixed?
I’m not seeing an issue with Disaster..
Are you still experiencing this? Is it relevant with other workouts?
It only appears to show up with power not NP.
Hi @eddie, something I just noticed, not sure if it’s been picked up yet, but I found this odd.
Today’s workout, completed on a gym bike not compatible with TR, its own app crashed, so no data, so all I could do in TR was convert it to outside & mark as complete. Anyway…
So this workout is 4×4’ VO2s (329, 335, 341, 329 for me), with a 30" ramp-up that’s officially part of the recovery beforehand. Now looking at the “power” chart, (not NP)… the maximum east of 1’ is 341 (expected because that’s the hottest 4’ interval target), but it rolls off after 3’30, as shown. Shouldn’t the blue line continue at 341 all the way out to the interval duration of 4’, then roll off to 4’30" (including the lead-in), then again to 8’ (including the rest), then pick back up to 12’ by including the interval before?
I would expect NP to roll off at 3’30" because NP takes a 30" average. But interestingly, the NP chart remains at 341 until 3’45".
Antelope which I have scheduled tomorrow does a similar thing: AP chart rolling off at 9’30" when the hottest interval is 10’.
Screenshot showing what happens at 3’31, & that the interval in question is indeed 341w:
Interesting.
I’ve just reported this to the team. Thanks for the heads-up!