Got back on the bike after a long hiatus. Did an ftp test and built a one month general fitness plan to get things going. after doing a nice 2.1 endurance ride yesterday, Trainerroad suggests a 7.6 vo2 max workout today and a 9.4 anaerobic workout at the start of next week. All my levels except endurance are 1.0 and i haven’t been as unfit as i am today for many years. I have a hard time believing I should trust the AI on this. Could anyone clarify or remove my doubts? My legs are terrified!
2 Likes
I’ve had a few of these come up lately, and been equally scared, but every one that’s been scheduled has been hard but doable. I’d brave it and see what happens!
2 Likes
I gave it a shot, and to my surprise i was able to finish it, 8/10 difficulty but it was definitely possible. That said, I probably won’t be doing much walking tomorrow - but that’s another matter 
6 Likes
I’ve not seen this kind of thing in particular in a long time but it does sound familiar from the past. I have seen a few times this season though where it seems like adaptive training is deliberately trying to find or get close to my point of failure with aggressive PL ramp rates in VO2/anaerobic workouts.
I’m not sure how much that is a deliberate strategy in the TR product design and how much it’s something the AI is learning to do on its own, but I suspect it’s serving a similar calibration function to a ramp test. All these workouts tend to be shorter, harder intervals so, even if you blow up and fail to complete the workout 75-90% of the way through, you’re probably still getting a decent training benefit and the system is getting some really helpful data about what your current limits look like.
I guess it’s similar to when traditional human coaches give workout instructions along the lines of, ‘repeat the set/interval as many times as you can and stop when you can no longer keep the power above X%’. Helpful for confirming an educated guess about how much an athlete might be capable of when you don’t have a lot of recent data to reference.
1 Like