How heavily does the AI system weigh HR data?

I had a thought today as I was out riding. And it made me wonder how much the AI relies or utilizes HR data to make changes.

So I was outside for the first time in probably 3 months. We’re just starting to warm up here. And it’s still a little chilly. But I was halfway through my ride and noticed my HR was lower than what’ it’s been while doing the same power indoors. So when I finished I took a deeper dive. This ride was around 90 min, avg power 191W, avg HR 111 BPM. Rated easy. Yesterday I did a little over an hour inside, avg power 199W, avg HR 129 BPM. Again rated easy. Then, I went back through other indoor Z2 rides and found my indoor HR is anywhere from 15 BPM up to 25 BPM higher for the same power and RPE. Meaning an easy ride outdoors I’m chilling around 115 BPM, maybe 120 BPM. Inside it’s fairly normal to be above 130BPM or even 140 BPM for the same power. Definitely due to heat. But it got me wondering if it’s messing with the AI system. Because I’ve seen a couple intensity days get changed to easier workouts after a seemingly chill endurance ride inside. I rate them easy because that’s what they are, but my HR is 20 BPM higher.

So is the AI seeing this HR and thinking I’m going harder than usual or think I’m fatigued? Anybody not using HR data with their training?

Hey @CaptainThunderpants :slight_smile:

Please post pictures of your Calendar and if your profile is public, a link to it. It makes it easier for everyone to see the big picture and what may be causing the Yellow Days and give feedback.

From what I understand, HR is more of a secondary signal. Power, workout completion, and how you rate the session matter way more. Indoor heat pushing HR up can confuse things a bit, but if you’re consistently rating rides as easy, the system should eventually learn that pattern.