Just wondered what other regular TrainerRoad users on here think about the new platform that’s about to be released, using machine learning to adapt users’ individual plans? The theoretical potential sounds great and as someone that’s had to constantly dial back some of the sessions to be able to complete them, I’m quite interested to see how it will work.
2 Likes
Still listening (¾ through after my 90min long run), but the latest TR podcast is quite illuminating. Sounds really intriguing. Although not sure how it’d work alongside being coached!
The main issue as I’m listening is for triathletes, the continued lack of pulling in other training stress is really becoming a drawback in what they’re doing, as it otherwise sounds excellent.
1 Like
Yep.
We’ll see how it goes when I get on the beta, it may be less impactful than at first I assume but it will depend I suspect on how consistent you are with your other sports.
For example, I’m doing a “max strength” block at the moment and will soon be hitting my limit on squats, unsurprisingly this is starting to affect my bike sessions. When I get on AT it will make some judgement on my ability to do VO2max intervals based on the fact I canned one of the workouts. Later I’ll fail workouts because of an overly hard run at some point.
Optimistically I’m thinking it will learn quickly to adjust back when I drop the weight training stress but we’ll see.
I reckon they’ll have run onboard and AT this time next year.
1 Like
I was hoping that, but it really didn’t sound like it. They talked about pulling in step data as being on the roadmap (“it’ll be really interesting to see how a higher than normal or lower than normal step count impacts compliance / performance”), but that was it. Which I found very surprising. Not least because TR was born from Nate being an Ironman athlete at the time.
Cyclists are definitely their key market, but they’ve skirted truly supporting triathlon in any development since I’ve been with them (about 6 or 7 years). The swim/run sessions in training plans that they added are really only a token gesture seeing as they don’t pull in the data/compliance. I only really stick with them due to my grandfathered pricing (a big factor), the fact I like their PR presentation and season cutting, and the fact that the workout library is excellent, so if I just want to do a random bonus workout, they have me sorted. The new Train Now feature will be good for that.
It’s also good that they now automatically pull in Training Peaks custom workouts my coach sets, so if I want to watch TV I’ll still use the TR minimised layout feature. But I actually now prefer zwift for workouts surprisingly. My coach sets cadence work within power targets, and zwift incorporates that into the segment targets, with popups if my spin slows too much, or gets too high etc. Given TR shows cadence, and Chad’s on screen text gives cadence instructions, it’s always surprised me they don’t show that more formally
1 Like
Yeah I wouldn’t count on my prediction 
I lost my patience with them on importing swim/run a few months ago on the feature request thread that’s been running for three years or something daft.
Which is why I’m using TrainingPeaks for run records now. I did consider sorting my own bike workouts on Zwift or Rouvy but I saw good progress on the bike last year and I still like their app, plans and workouts.