This is not the first time I have seen what appears to be incorrect mWR information for a session but this appears to be the most off base. As you can see in the attached screenshot the graph shows my lowest mWR at ~42.85 but the summary and the AI generated recap indicated a mWR of -4%. This would also not mesh with my RPE of 3 I would not think.
Am I missing something or is this wildly incorrect?
Thanks for noting that. Workout Reserve is always looking at today’s performance in terms of the running maximal mean power relative to the historical (6 weeks) maximal mean power. Here is the original blog post that explains the feature. Here is one example of a pro cyclist winning a sprint stage and another placing second at the Giro on a mountain stage. Finally just so the collection is complete, here is how you can get the feature in real time on a Garmin.
I may be missing something but when I eyeball your session I can’t see the WR (green line) hitting minus 4%. This could be a bug. I’ll cc @Andrea here the inventor and hope he can look into this for us.
Thanks @Prof. The fact that the plotted mWR does not drop below 42.80% is my concern. If my future sessions are being driven by the incorrect -4% that is an issue as I never “dug deep” during today’s workout. The only reason my mWR dropped below mid-50% is because I was at ~180% on the last 30 second interval as opposed to the ~110%-130% of the previous 69 30 second intervals.
Dear @unospeeder and @Prof, the discrepancies between table and plot values might be due to the “smoothing” applied to the graph.
Is that the case? I can still see a minor difference between the -4% in the table and the 0% of the graph at the end of the last interval. But you can clearly see that the mWR goes below 40% and reaches very low values eventually.
Appreciate the reply @Andrea, it appears that it is related to the smoothing. Easy to miss as the ability to remove the 10 second smoothing is below the fold on anything but a very large display/resolution.