I really like the adaptive trainer feature. I am not sure if it has adapted to me enough yet, but that’s a different topic.
I think it would be really cool if TrainerRoad combined its adaptive training with data from wearables. I am not sure if Whoops technology is there yet; sometimes it gives me green days when I know I slept terribly, and I try training, and times when I am reading, and I feel great.
But if TrainerRoad took data like HRV, resting heart rate and added it to the data they already have, it could lead to really cool results. I would imagine at first you could just train an AI on the data without implementing it, if it gets good enough at predicting performance, you could try a Beta.
Data from different wearables would obviously need to be treated differently. I think someone will eventually do this, and TrainerRoad may be best positioned to do this. From what I understand, they may have the largest dataset of all platforms and the highest quality. Zwift may have more data, but it probably has more variables.
So far, I mostly use my Whoop for fun and to track my sleep consistency. I think one day wearables will get so good they can tell us better than a coach can when to train and when to rest.
IDK. I got a Garmin Venu3 six months ago to track fitness and it has a lot of data but I don’t see anything usable or actionable from any of it beyond what I already know without it. Maybe AI could come up with something.
I agree. A new platform (mats) integrates for its AI coaching Whoop and other devices. But there is a short questionnaire you have to answer every morning and the results outrule the Whoop Data in case of doubt.
Does Trainerroad include some sort of questionnaire in the meantime as well?
I see some HR weirdness at times but cycling it’s been pretty good. I make sure the band is snug. Bouncing around on the wrist can make the readings erratic. I’ve recently done a number of rides dual recording with my Venu3 broadcasting HR to my Garmin Edge and my Garmin chest strap recording on different Edge. They’re just about always within 1 bpm and I get the same avg and max for an activity. There have been other times when it’s been WTF? I did a treadmill cardiac stress test a couple of weeks ago hooked up to a EKG. HR went to 146 but afterwards when I looked at the data from the Venu3 it showed only 99. My hands were holding a stationary bar so no bouncing. I hadn’t started recording an activity with the Venu3 and it was just doing its background recording so maybe that has something to do with it.
FWIW, I still find my Whoop terrible with high intensity HR’s. They were supposed to have improved it with the new one, and maybe they have, but I find it useless for HR above threshold. It therefore leads to inappropriate conclusions down stream like their strain calculation, VO2max estimation, etc.
I’ve tried tightening the band, loosening the band, wearing it in different places, and it’s always off at intensity. But maybe that’s just me. Bottom line, I’d be very cautious about integrating wearable data in a meaningful way with TR.
I have had similar problems with both my Whoop and Garmin watches. I find wrist based hr is terrible, especially for high heart rate work. I am not even talking quick spikes, even long, high heart rate sessions will show 20-40bpm lower. I have found that for running, it works ok, but for anything where you are doing something with your hands, I wear my chest strap. If I am just sitting around, my Garmin and Whoop seem to both be the same as my heart rate strap. I should try sleeping with my heart rate strap to compare. I also have a bicep strap for my Whoop, and that seems to be much better. Like I said in my original post I don’t think wearables are good enough yet, but you could try and train an AI on the data and see if the predicted performace lines up with performance. It’s not like it needs to be perfect just better then a human.
I have the new one, and it’s still terrible on my wrist for high-intensity stuff. I was getting annoyed because it was not showing my V02 max work, and so I wasn’t getting the age I wanted(not that it really matters). I bought a bicep strap, and now it’s basically the same as a chest strap. I am disappointed overall, though it’s an expensive device, especially the new one, and it’s not that great, when I wear the bicep strap rolls over on itself, so the sensor is pointed up.
Often, day-to-day metrics don’t really tell you anything. If my HRV is bad one day, I just ignore it unless I know I slept less than 5 hrs. The only time I really respect red days on my Whoop is if I know I did a big workout the day before, I know I slept terribly, or I had multiple days of low HRV.
I tracked everything on my Garmin Fenix for 3 years. Interesting data but at the end of the day, nothing really usable. HRV did show degrading before I got COVID. If I do not sleep well, I cannot do anything about it. I am not avoiding lengthy sleep. I stopped tracking habit about 3 weeks ago.
I’ve had issues with both wrist based and biceps based HR straps - they just don’t work consistently enough to provide good data during a workout. Chest straps are much, much better. For me personally, Heart Rate is always dead on and a very reliable metric during exercise as long as I’m using the chest strap.
Personally - I think the sleep duration data from the multiple devices I’ve tried are all hot garbage. HRV and heart rate are better, but there are times when I’ll literally be lying in bed awake for 75% of the night and it thinks I slept great. You’d need to base the sleep portion on a survey.
