Current state TR?

Uncategorized
1

Hi all. I’ve been a TR subscriber on and off for about 7 years. I had fantastic results for about the first 5 or so years but the wheels really came off last year. This was no doubt due to a mixture of reasons, including aging, stress & lack of consistency.

on top of those personal reasons though, I found the AI-related changes to be confusing and demotivating. I could no longer understand why I was being served specific workouts. The AI FTP prediction issues seemed to impact me every month (& I never do outside workouts) so that I never seemed to ‘succeed’ & it was demoralising. The whole thing seemed unbelievable sensitive to feedback surveys, not just in terms of FTP prediction but also in future workouts to be served. And then, most confusing of all for me, is that TR FTP is now different to the FTP on other systems that I trust (eg Intervals.icu) so I can’t assess or have confidence in my progress. Even though I read a lot on forum here, I couldn’t really understand what was happening (my bad). For these reasons I stopped my subscription in the summer as results on the road were way behind other years (again for the mix of reasons mentioned above).

I’m now wondering if the issues have been fixed? To be honest, I long for the old days when I could choose a training plan and can see all my future workouts in a calender at the outset, without everything changing constantly. I would like FTP on TR to align with FTP as measured elsewhere. I want to understand why future workouts have changed (and I agree it’s sometimes useful), not just be told “trust the AI”. Frankly I don’t trust the AI.

so, given my wishlist, is TR now meeting those needs or am I better off elsewhere this Winter? There’s so much right with TR I really would like to return but I can’t handle another season of confusion and second guessing.

2

+1

3

Same boat for me. I just booted up TR to create a new 2 month low volume sweet spot base program, and the AI just keeps giving me the same 3 workouts. Yeah, I get it, those 3 will work. But I’d like to see some options from the 2000 workouts. And no, I don’t want to find a substitute before every workout.

I started using Intervals.icu when it first came out and it works for my calendar and tracking. It’s my main tool to track multi-sport workouts and look at data. I use a Garmin computer and Connect, which syncs to Strava, then to Intervals (via Strava). Ideally I want my data in Garmin for safe storage and full export.

I use Biketerra for visuals and sim rides. Best part is the ‘companion’ riders that join you on your ride ap you have your own group that hangs with you. I pay monthly to access all the routes. You can just ride the free routes but the scenery is limited. The visuals used to look like 8-bit graphics whihc was just fine for me, but everything has come a long way and looks great now. MyWhoosh is fine too, (and free too) but I find the promo graphics and advertising too much and the routes are meh.

Biketerra has a few, free basic workouts, and imports workouts from intervals.icu, Trainer Day, Athletica (?) or ZWO (Zwift workout files, good for training peaks import).

So, TrainerDay is looking like a better match for training plans and general calendar. It has an AI builder too, but seems to do a much better job of using different workouts and creating plans. I am working with it but need to pay to get the planner working. TrainerDay will export your TD calendar to Intervals, and seems to do the whole plan at once, so you could activate TD make your plan, import to Intervals then cancel TD, and pull your workouts into Biketerra (or maybe do them in Intervals).

My plan is to start Trainer Day, build some plans, import to intervals, workout using Biketerra, then find a way to import into Garmin.

I will probably cancel Trainer Road and not renew.

4

I’m sure people will chime in to tell you why your concerns don’t matter or how to better use the app and learn to live with it, but the direct answer to your question is that none of the things you mentioned have changed.

5

The main issue that still seems to be looming (although never really affected me, so I don’t keep close tabs) is the impact from outdoor workouts.

Aside from that, indoor training and TR and AIFTP and all that has been phenomenal for me. While bugs can exist here and there, seems the majority of people with disappointments often stem from lack of adhering to the plan.

I don’t really experience the issues you mention. I set up a plan and it nearly gives me what I specify (except for weekend solo ride length, which is easy for me to manage). I tell it to build a specific plan with 2 intervals per week (90 minutes each) 1 endurance ride at 90-120 minutes, and a weekend long ride up to 5 hrs.

It then delivers exactly that. My future workouts may adjust based on my ability to complete certain workouts. Usually reasonable and not huge changes. For example, if a moderate workout was rated very hard, it might temper the next workout back a bit.

I also don’t really skip or shorten workouts unless I’m sick or injured, both are pretty rare. Probably 90% of my ride surveys are what it predicts.

Regarding FTP vs aiFTP I did a 20 minute test earlier this year and it was 97% of my aiFTP

For riders with good discipline and especially those that do intervals indoors, I find TR to be great.

6

Exactly this! FTFP and you will get faster, TR AI and AI FTP have updated our understanding of performance. old concepts of FTP are no longer relevant. The sooner people can trust the AI to make them faster and follow the plan the sooner they will realise that it works and enjoy the promised gains.

7

I’ve been doing most of my workouts outdoors all summer and my FTP predictions were plagued with issues. It got so bad I started taking ramp tests every month to replace the FTP predictions. However, it seems things are fixed now maybe? I did a hard group ride earlier this month and got a new FTP from intervals.icu. TR’s predicted FTP is exactly the same as interval.icu’s, which previously was never the case. Previously, my FTP would increase on intervals.icu, TR would be way off and then I take a ramp test and I’m very close to intervals.icu’s estimated FTP.

It does bear reminding that intervals.icu and TR’s FTP seem to require you to do a very hard effort for the prediction to pick up significant increases in fitness.

image
image250×396 25 KB

image
image1058×594 25.7 KB

8

I would say adhering to the ai plan but ALSO understanding your own needs and goals.

For example, I 100% do not agree with TR’s philosophy of short “long rides” preparing for 5+ hr races. For the life of me I can’t get TR to prescribe 4-5 hr weekend rides despite doing them regularly on my own just fine and have been for a couple years. I tell TR I have availability for 5 hr long rides and I still get prescribed 2-3 hr weekend rides. TR stands by this strategy which I couldn’t disagree with more.

Also being in tune when you need to dial back or adjust certain things that ai couldn’t possibly know - a cold, a nagging ache, etc.

I’ll say that the ai training is 80% great. I think if people expect to just turn off their brain and let go completely, it’s probably still better than developing their own plan, but using common sense and being involved and understand their training program also goes a long ways.

9

Sometimes a company will evolve over time and its product will become something you don’t really align with anymore. That’s a bummer but pretty normal. I’ve had to switch brands on products and services over the course of my life as products evolve and my needs change or they discontinue a product altogether and its no longer supported and I have to go hunting for an alternate solution. Just part of what markets do.

I’m sorry to hear you’re having that experience though as I’ve found the change to be a massively positive one for me although, yes, I am still flummoxed at the retooling of what FTP is and decoupling it from the industry standard usage of the term. But now I kind of don’t really care much. I know what my numbers are for different zones and workouts and the FTP number, whatever TR says it is, is completely irrelevant to me. I do sympathize though, it bugged me at first too.

10

I started using Fascat as I didnt like the AI version of TR. So far its been good, the plan builder was good and even though its AI, it doesnt constantly switch you upcoming workouts.

11

From what I’ve seen, TR is the best bang for your buck system out there if you don’t want to manage your own training. Even if you just use the training library, workout player, and calendar, it’s a decent value.

That said, I believe the best training system is the one that best motivates someone to do the training. All the debate about whether a workout is “right” or “wrong”, whether your FTP is being properly set, etc. is marginal noise in comparison to being consistent in your training approach. And from what I can see in comments about TR’s AI approach, it seems to motivate some users, but frustrates others. A system can be serving up the best plan in the world, but it doesn’t matter if that system isn’t making the athlete feel good and confident about their training approach. It seems like TR’s new AI-driven approach works best with people who are happy “trusting the system” and don’t get wrapped up in the details (ie - my ftp estimate went down after nailing my workout, my workout changed for no apparent reason, etc.). The approach obviously isn’t for everyone, but you can also use TR as a dumb system and just pick and execute workouts.

I personally think the “pre-AI” version of adaptive training was a nice balance between system driven and self-driven plan management, but I don’t have any big issues with the AI stuff either (but honestly don’t use much of it). And I expect it will get better over time. We are really at the dawn of AI-based training and I think TR is doing a nice job.

12

Another interesting voice in the ongoing “What the heck is going on with TR?!” discussion.

Although I’m a creature of habit, I just couldn’t put up with the ongoing enshittification of TR anymore. The outdoor workouts bug, with no credible fix in sight, was the final straw that made me decide not to renew my subscription.

13

I’ve sort of given up in the last couple of weeks when I realised that I wasn’t being honest with myself, and TR, in the survey because I found it so demotivating watching my AIFTP drop precipitously when I rated a workout as anything other than easy or moderate. This then led a death feedback loop of me failing workouts.
I’m aware this is my fault but the predicted AIFTP being far too high at the beginning is so demotivating to me

14

I’m actually glad to read your comments. I feel the same way. I loved TR until the AI came along. I was in the first Beta control group and asked to be removed. The ride recommendations were way out of my abilities. They declined my first requests until I insisted. I emailed TR to discuss and they actually advised me to “try the rides and quit if I needed to” or “just drop the percentage on the ride”. I like a challenge but I do not like to quit solely because something has been forced and it’s not within my abilities. The whole AI workouts and prediction of FTP became very demoralizing for me too.

15

You are better off elsewhere or actually you are best off on Zwift :grinning_face:

16

Do you think Fascat is actually AI or just some simple IF THEN logic.

17

My understanding it doesn’t give 5 hours is because that alone is more TSS then entire rest of week. Basically what its saying is if you can do 5 hours at endurance you are better to do another hard ride instead because you don’t make any gains on Endurance. You progress is only 100% on hard intervals and if you can nail those improvements will come and endurance at the same time. If you absolutely enjoy 5 hour ride. Do 1st hour FTP or Vo2 and then so other 4 hours at zone 1 just for sake riding. That 1 hour will move the needle.

18

I think there are plenty of studies (and TR) that would disagree with this. Why would they even prescribe any endurance if there were no gains from it.

19

:face_with_raised_eyebrow::thinking::face_with_monocle:

Are you saying that endurance gives you no gains? And that the only way you make progress fitness wise is through hard intervals?

20

The argument is that if you have limited training hours, the most effective way to get the most out of those hours is with a strong bias toward intensity over endurance. If you are training 7-8 hours a week, doing a 5 hour endurance ride probably isn’t the best use of your time from an optimization standpoint. The hard part is knowing what the right mix is as you add hours, every one of us is a different animal. And at some point, we should lean into the mix of training we enjoy because that’s more likely to keep us training consistently.