Sorry, I was confused and confused you. I will figure it out. Thanks for your help. I will let you know
1 Like
I have not forgotten about this, for some reason this is taking longer than I would hope to figure out what is wrong.
We talked in email, but I will respond here. So we have fixed the issue. The reason was because some values are NULL in your power data, but still our code did not handle this well. Now the day that you provided that was 400+ TSS is 216, but intervalsICU and TP both show 227/228. This seems like some kind of rounding/integer difference.
You can press refresh strava and it will update your activities. But we are still searching for this small difference in TSS. We are sure our calculations are good. And verified with chatGPT provided code which also gives 216, but not sure why theirs are different.
Alex, I appreciate you looking into this for me and getting it corrected. A few percent difference is a non issue and I’m sure very common between platforms. It looks as though some of my past outdoor rides are still incorrect but I’m more concerned with future accuracy.
Cool I would love it if you would send me a TCX of this older one with a problem so we can make sure we cover all cases.
Looking a bit harder into it there are quite a few rides(outdoor) there have varying degrees of significant error so not sure how deep you want to dig into this. Most of my rides have power and a few only heartrate due to lack of power meter on a particular bike. TR calculated TSS via heartrate in those cases. I’ll send a few of them that have the largest difference.
Yes, just with power, we might use strava’s estimated power but we don’t use HR for TSS calculation at this time so those would be way off.
Thanks for sending me those workouts. I just confirmed we don’t recalculate older than 1 month ago. So we should be good to go, going forward.