Whether this is a hard problem to solve, it seems like it’s closer to the root cause. At present, when I do a ride, however it changes my prediction, as a user, I have no idea. Even something as simple as an up or down arrow on that ride indicating whether it was beneficial or not. Better would be guidance, like ‘HR response good’ or ‘power not to expectations’ etc.
The bigger issue here is that when AIFTP is working “perfectly” and the outdoor bug is fixed, seeing unexpected prediction changes -even if they’re well founded- will raise questions.