This has been my thought process as well. If Trainingpeaks had a decent indoor player (had a lot of issues with TP virtual, as my wifi connection is not stable where I train), I‘d make the switch for sure…That (and being grandfathered at a lower price) is all, that keeps me at TR at the moment…
TPV is included with TP premium and is excellent on my Apple tv
Yes, I know.
The problem I encountered is, that it doesn‘t have an offline mode. So when the connection cuts out for a moment, you will have to manually upload the ride afterwards and delete the duplicate.
Also I was never really drawn to virtual riding, so it‘s just a bit cluttered for my needs.
Also tried Trainer Day, but it crashed on me several times.
Do you know if they’ve fixed the issues with the interface on Apple TV? It IS super configurable, but in the early days, it was almost impossible to configure because the cursor would jump all around and made it almost impossible to configure (using either the ATV remote or the iPhone remote interface). It was so bad, I quit using the app.
I think at the time they weren’t even actual polarized plans, maybe were hidden under extended base or something of that sort. I’ve been away a while and may not remember semantics but most people know exactly what I’m talking about. And I think they only ever made progress on it after that Dylan Johnson video criticizing their training methodology.
Any feature requests? Seems like TR is leading the way with Ai detection . Many are following.
I’m not sure what else I need from the platform. Love to hear thoughts on what specifically you would like. There are feature requests post on the forum.
You can actually do that if you tell TR that you have more time. I’m doing a 2-hour sweet spot workout this Saturday (PL 9.8, 1:29 hours at Sweet Spot). It requires some manual tweaking in that you need to change the workout time.
I haven’t figured out how make it give me long sweet spot intervals during my long rides. I set my workout time to 2 hours (the maximum allowed) on Sundays, but it still gives me all of my sweet spot during my shorter 1 hour workouts during the week, and on Sundays tends to prescribe short over-unders or very short threshold intervals combined with endurance. I wish there was an easy way to just reverse this so I could do long sweet spot intervals during the longer workouts. I haven’t taken the time to play with all of the options in plan builder so I might still be missing something.
There’s a Companion App out now that makes selections super easy
No, they have always been labeled as polarized plans, not extended base or anything else.
Polarized plans have been released as Early Access feature in March 2021, the earliest Dylan Johnson video on that topic was December 2020. However, that is different from the video I think you are referencing this video, though, from February 2021.
While I don’t have proof, I reckon the polarized plans had been in the works already. Our very own @mcneese.chad rolled his own polarized plans as early as 2018, so the idea definitely wasn’t new nor was it “impossible” to do polarized training with TR.
My 2c for what it’s worth:
- The move from selecting one plan from a menu to individualised AI plans has hurt the sense of community here. Previously you could post ‘I am on Mary Austin in SSBII’ and you would find some understanding of the struggle point you are at. Now with tons of different workouts, there’s no shared sense of working through a plan, and there’s little shared experience.
- In a similar vein, the standard workouts in all plans used to have in workout instruction. That meant that you were also being reminded (and distracted!) to think about your breathing, your pedal stroke technique etc. Thete was a progression in these instructions as you progressed through your plan. This helped the process feel more like a complete coaching approach rather than just a ‘workout server’. Now it’s a bit random if you get workouts with instructions. The overall feeling is more individualised but less comprehensive.
- Associated with this, TR has never really replaced Chad as the ‘coaching guru’ giving confidence that the science is right behind the scenes. It was Chads voice I heard in my head when reading the in workout instructions. This matters as now I’m not so convinced that the science and instruction is reliable. Even if I guess it is, it’s less reassuring.
It’s tough because I think the AI driven approach is probably for the best. Rereading the above I would suggest to I) ensure there are in-workout instructions in every plan, independent of the actual workouts and iI) develop a new head coach to design and represent (and defend) TR’s overall approach.
It wasn’t impossible but it was pointless to pay for training plans and then have to structure my own training plans. YMMV of course
I’m with you 100% here. After wrapping up one of my best seasons in a long time, I was very curious and eager to give TR a try after not using it any serious way in 3 years or so. After the re-onboarding period, I ran into some issues and ultimately ended up giving Fascat’s CoachCat app a try for this off-season.
So in the interest of clear product feedback this was my experience:
-
AI FTP - my detected AI FTP was 289 the day after my season’s A race, which put my IF at .98 for a hair over 4 hours (AP 231, NP 284) in my gravel race. This was not an NP buster event; I broke 600 watts for maybe 10 seconds, then I was barely over 400 during the rest of the race, since I spent the month before heavily focused on consistent pacing. It seemed low, so posted on the forum in an AIFTP thread and was told the team checked and this was correct. 289 was where I was stuck when I was using TR before, so it sorta made sense when one of the TR support mentioned there were issues with historical data when people were coming back after long breaks.
Also, part of the thing with the combination of AI FTP and Progression Levels is that it provides the flexibility to ballpark an FTP then finetune it through post-workout RPE reporting. But it seems likely that could skew Red Light/Green results one way or another if that’s still pegged to the detected FTP but the workout levels effectively make it higher or lower. Pure conjecture here, but I’ve been thinking about how the systems might interact.
-
Human Bodies are not deterministic - This is a complicated one to tease out, but it seems like there’s a baseline perception that based on A (your AI FTP), if you use a formula where X=AI workouts and Y=User Compliance, you will end up with Outcome Z. But none of the models are this precise, and the human body and training readiness can vary wildly from day to day based on a million factors. Seeing the succession of numbers (FTP, TSS, Progression Levels, etc.) creates a sense of things being much more controllable and less fuzzy than they really are and sets up users to believe that if they follow the signposts correctly they’ll continuously improve.
Then when they don’t for any of the usual reasons (Life and non-exercise commitments being the big ones), they assume the failure is their own fault because of a lack of compliance. But the two best things I did for my training were large increases in volume and leaving a tremendously stressful job. For a few years I worked very long hours doing a high-profile job and while I was mostly just sitting on my ass all day working from home, the physical stress of an insanely toxic environment was perpetually exhausting, and my body didn’t care that the stress wasn’t from exercise. I was just perpetually fatigued and couldn’t improve.
It’s super weird: the system just feels like it’s built in such a way that blame is on the table. I’ve used pretty much every training platform on the market, and TR is the only one that made me feel bad about myself.
In my mind, I’m not paying for training plans, but for a self-coaching tool that generates training plans and easily allows me to adapt them to my needs. It has always had limitations, but was flexible enough (for me) to accommodate my needs.
What we discussed here doesn’t qualify as structuring my own plans, it is adapting them.
Nowadays, the format has shifted — it’s no longer a podcast-style discussion, but more of a pre-recorded presentation on a specific topic. The pros often talk about how they train, what they eat, what challenges they’ve faced, or what they managed to avoid in top-level competitions. Of course, hearing about their 1-2-3 hour record-breaking watt numbers is impressive, but is that still genuinely interesting for amateur sports fans?
Much of this information is already well-known, especially if you regularly watch professional cycling. And if you don’t, then hearing for the second or third time how Mat held 400 watts for several hours isn’t exactly captivating.
As a European listener, I also find the strong US focus a bit frustrating — both in race commentary and in the choice of podcast presenters. TrainerRoad knows which EU cyclists use the platform; why not involve some of them occasionally? There are countless interesting topics to explore. Jonathan does a great job, but unfortunately, he often ends up being more of a host than an active participant in deeper, more diverse discussions.
Honestly it’s not all bad, like I said TR did help me make a lot of initial gains but I think more importantly helped teach me what worked for me and what didn’t. As a result I’m able to basically do the same thing on my own and there’s definitely value in what I learned. Could another training software have done that? Maybe, but I can’t say for sure and still do recommend TR for people who are completely new to training. Someone who has a few seasons of training under their belt can probably do without it though unless they just do the workouts as told and don’t put any thought into it.
Sports fans love stats. That’s why they keep them. Ever listen to an NFL broadcast… “that’s the most rushing yards by a left handed fullback wearing 44 on a Thursday!” ![]()
Seriously though… people don’t like hearing professional stats? Yeah I can’t apply it to me but it does make appreciate what it takes to be on that level.
I’ve been using TR for nearly 8 years, and learned much along the way. The majority of the time I roll my own training blocks, while still appreciating AIFTP, the Calendar, RLGL, Workout Levels, the Workout Library, the Workout Player etc. Occasionally, I let TR run things for me, since this provides more novel workouts and sequencing vs the simple progression block approach I tend to plan myself.
The product, although similar in many ways to 8 years ago, has really come on leaps and bounds IMO, and it’s easy to lose sight of the many excellent improvements - small & large - that have been made in the time I’ve been using it.
Much of the stuff underpinning Adaptive Training is still a great benefit to me as someone predominantly scheduling and managing my plans myself rather than having TR/AT do it for me. My main concern for the future is if moves to automate things more & more (helpful to more regular users), or perhaps to hide some of the inner workings, might, as an inadvertent side effect, remove useful functionality from me as someone doing a lot of manual plan management. eg. if Workout Levels were to be deprecated or moved to be behind the scenes only, for instance.
For now though, TR is still a really great tool for me - even though I largely plan my own stuff.
You are not wrong, but by the same token you wouldn’t need a coach either. Or that I don’t need Excel, because I can add, subtract and multiply. Tools like TR it simplify things, take repetitive busy work off my shoulders, it probably does those things better than I am and saves me time.
But those factors need not be decisive for a hobby.
Are you using MyWhoosh as a workout player (or trainer controller) to do your intervals on? I’m using TrainerDay and paid for it. It’s ok as long as it doesn’t crash, certainly not as slick as the TR player. Also using intervals.icu as a calendar and should probably pay them something.
I actually prefer the TrainerDay workout player. It’s far more customizable. Also, I prefer the way the Erg mode feels. It’s never crashed on me either (and neither has the TrainerRoad one).