Can we get a wider margin for paces before a run is automatically reclassified/penalized?
I’ve been doing several endurance sessions lately. Because I wanted to push the volume, I ran some of them slightly faster (by 5–10 sec/km). The algorithm reclassified these as “Tempo” runs and subsequently pulled back my suggested load, claiming I did too much high-intensity work.
To fix this, I’ve manually set my endurance pace 20 seconds faster to create a personal “buffer,” but it would be great if the algo was a bit more lenient. Running 5 seconds into the tempo zone is a “fast endurance” day, not a “threshold” day. Could the algorithm distinguish between clipping the edge of a zone and running in the middle of it?
1 Like
It’s true the margins/tolerances for the zones are exact right now, for example you’re either below or above the Tempo limit (no pun intended)
I’m planning a per workout compliance feature for the near future. There we can definitely include some sensible tolerances
3 Likes
The analogy the Podcast School of Exercise Physiology has lead me to understand is the training zones are not switches but dimmers.
I could imagine that simple algorithmic tolerances (eg: +/- 5% for endurance, +/- 2% for anaerobic), is there research on how much training zones overlap, or would this need to be done using ML?
yes definitely not an exact science, problem right now is that we have to classify it in either zone for compliance. Once we add per activity compliance that will help