Nah, Nate has been hinting and talking about it for almost a year now.
2 Likes
I guess we’ll see
Here is some interesting outlook:
Quote:
“Every 24 hours, the model aims to correlate your best historical performances with your historical training inputs (pace, power, hr) to find a best-fit signature. This signature is then used to personalise relatively ‘normal’ plans that you would be familiar with ie you tell the model your training preferences and constraints. The resulting plans have some degree of adaptability both to your weekly schedule and changes in the signature.”
To me this sounds like what a combination of TR and ML could look like 
…
I mentioned in another post that I wondered what the prevailing priorities would be for users. My take as a participant in this forum is that the distribution would be pretty wide, but wonder if the group would coalesce around any central themes, or perhaps get segmented into 2-3 archetypes.
For me, it would be better outdoor workout integration, better analytics, and increased app functionality - simple things like skip ahead/back, choosing metrics, as well as website functionality in the app itself. And a more functional/flexible calendar and plan builder. Sort of table stakes stuff that improve functionality rather than jet packs.
1 Like
It’s interesting that you view these are contrary goals. I would hope that any implementation of a custom plan creation based on machine learning would take into account goal types as well as your past training history.
Frankly, if all that we get out of a potential machine learning ‘feature’ is modifications to existing plans then it isn’t really a win.
Certainly building out more variety of training plans, things focused on other event types etc. is distinct from any type of machine learning - but I think there’s a decent shot that machine learning solves the problem you are looking for.
Now, as to whether TrainerRoad delivers that level of functionality or not is still an open question
2 Likes
Well that’s precisely why it’s opposed. It’s limited resources going to that as opposed to plan creation. TR has given no indication that other training plans and philosophies will ever be on the table whereas the AI is teased constantly.
Beyond that, Kolie Moore’s latest podcast actually did an excellent job explaining why machine learning is not a panacea for training. There is just no way to capture all the inputs necessary. Your response to a particular session or group of session is a microscopic piece of the puzzle for dictating future training stimulus needs. I won’t go into all of it, but you can give the episode a listen, it’s the one that he does a reddit AMA question session.
I’m feeling similarly. That said, those base and build phases I seem to be able to stick to indoors and tend to follow a TR plan, later in the year, I have to modify. In essence I think the biggest value for me is in the winter months when I’m really using the platform, as intended - those months seem like a no-brainer to continue on with TR, also grandfathered into better pricing from 2 years ago.
I’m not familiar with the century plan, but training different energy systems is what TR is all about. Certainly most of what’s on the century must be longer rides, or SS work? Honest question.
Either way, it would seem that a lot of riders who use TR find that they need to tweak the plan in some way to compensate for a plateau, or an over-reach in training. For me, I can’t do the LV plans in the build or specialty phase for XC plans. It’s too much HIIT. So, the thinking is that if that’s the case, why am I paying for plan builder if I’m tweaking it so much? I guess I answer that with the fact that TR is a self coaching platform and they are casting a wide net especially with plan builder. If you get past that and enjoy the UI and calendar, it might still be worth it.
For me,
- I’m still using a TR plan for at least a portion of the year
- The UI is really nice, and I know it well
- I think the workouts are typically well designed
- The calendar is great
For those 4 reasons, I’m staying.
1 Like
I am right there with you on number 2 and 4.
Number 3 I don’t disagree with per se, but I think they are not well designed for what I am trying to do, which leads me to make my own workouts.
That does raise one thing I left out. The Workout Creator. I really would like to see the new WC. The existing one works ok, and is definitely part of what adds value, but a web based one would be so much nicer so I wasn’t stuck with using my PC to make workouts, and hopefully the functionality like copy and pasting becomes much better. The new WC arriving and meeting expectations would absolutely swing the balance to staying as a subscriber.
1 Like
Just honest curiosity on my part, what type of workouts do you feel the need to create? And hey, good luck to you no matter your decision. I actually had to put some thought into staying this year for sure, so I understand.
One thing that I left out is straight up convenience; I’m already here. And, I don’t want to learn another platform. It’s a tiny thing for sure but that’s how I am, make it convenient and I’ll bite! 
Started triathlon last year… first olimpic tri race and I did the bike leg in 1h11min
Joined TR during pandemic
Followed SSBMV1 → SSBMV2 → SPBMV → Oly Tri Spec(only half of it)
First olimpic tri post-pandemic (on same circuit and similar weather conditons) Chopped almost 8min of the bike leg… 1h:03min
Although it maybe newbie gains … i got faster and i’m pretty happy about it… no reason to leave it at all.
Mostly the longer sweet spot intervals (45 -90 minute) and i have to tweak the Tempo interval workouts down a bit to actually be in Tempo and not SS.
Good points. I think I’ve said it before, but I’ll say it again, I think the issue is the testing. When I was doing Trainer Road, I’d do the 20min FTP test, try really hard. Then started training and the Sweet Spot workouts would kill me. My HR was climbing and climbing over the first 5 minutes. I would be fearful of my upcoming workout. I would have rather done a 20min FTP test every night.
I am now doing 3x30min Tempo workouts while watch YouTube. HR target based off of lactate testing and feel.
3 Likes
Totally agree. Our WC works but we are not proud of it. We got about 50% of the way rebuilding it on the web and then other stuff took priority.
We’ve got the whole company on one project, but after that is released we plan on finishing the web based WC.
We may have it in the desktop app too.
One thing I want to be able to do is to easily edit a TR workout to your likening. Ie you load the workout up, adjust the rest period or add a set and click start workout.
17 Likes
This sounds promising. The sooner the better really as the Adobe based WC is a nightmare.
I’m on my third year and am starting to see some limitations, and I only use the program in winter months. Can’t imagine having to go through the same base build specialty cycle over and over and over year round like some do. My suggestions would be as follows
-
Have a plan builder that focuses around time, let’s say I want to commit 10 hours a week, the builder should be able to build a plan that maintains 10 hours a week (maybe with exception of the rest week) and scale workouts up and down to keep TSS progression the same. I ran into this trying to do TB, HV in earlier weeks was fine but later weeks was too much, but if I did HV for TB1/2 and then MV for TB3, I had a TSS drop for part 3
-
Better suggested workouts: if I open Pettit for example, give me a section along the lines of “you may also like” and pop in some similar workouts with similar TSS.
-
Revise the app: it’s completely useless. Can’t schedule a workout for later, can’t analyze completed workouts, can’t view power curve info, I have to jump to the website for pretty much every single thing
-
Better workout filtering, if I want to do over unders (my weak spot) there is no way to filter just those workouts but have to find them among a sea of other threshold stuff (I WAS WRONG THIS IS ALREADY THERE)
-
There’s no way to “take a break” mid plan, not that I’ve found yet. Let’s say I’m in week 3 of SSB and get sick and decide to take a week off, there seems to be no way to drag and drop the plan further out without doing it day by day which is extremely annoying if you have 20+ weeks of training ahead. (WRONG ON THIS TOO)
These are just a few suggestions. Im grandfathered in at $129 a year but in the 2 years I’ve been a subscriber I haven’t seen much meaningful change outside of the plan builder which is also not perfect IMO. I speak highly of TR from a results standpoint, it works, but so do many old Windows 98 computers and after 2 years of little change the platform seems kind of antiquated. Im not asking for bright colors and fluff but some of the things I mentioned should be there IMO
4 Likes
Lots of agreement in your points. For this one, you can push/pull weeks as well as copy/move (or clear) them using the “…” menu next to the Week # rather than doing each day. The push feature affects everything after it which you may or may not want so you need to be careful with that one if you have a long term plan on the calendar and don’t want to completely affect the whole thing.
1 Like
Those and more covered here.
2 Likes
Ok I take that one back. My other criticisms stand
Include search to improve filtering.
There’s no way to find over unders specifically that I’ve found. I’m talking Palisade, McAdie, Leconte, or even stuff like Avalanche Spire