I’m reaching out again because I’m starting to have doubts about the TrainerRoad program is really working for me.
Last season I was training 8–12 hours per week and had an FTP of 314W.
After taking a one-month off-season break, I restarted with TrainerRoad in January. The AI gave me an FTP of 288W.
I then created a plan with TrainerRoad and followed it. By the end of January, my FTP had dropped to 274W. Okay, maybe as a comeback it’s possible that the initial FTP was a bit too high.
When the new AI was released in February, I received an FTP of 273W. Unfortunately, I got sick for a week and then restarted last week at 273W. After three workouts yesterday, the AI predicted an FTP of 293W within the next seven workouts. Today I opened the app and suddenly it showed a predicted FTP of 273W — so +0W.
I’m currently following the “Increase FTP” plan, and somehow I’m not getting any endurance rides except during recovery weeks. I don’t have any time constraints and could easily train 8–10 hours per week, but TrainerRoad is only giving me three hard sessions each week (1x VO2max, 2x threshold, over/unders, or SST). Of course, in winter that might not be too bad when you’re not spending hours on the trainer, but I’m worried about ruining my aerobic base by constantly doing only high-intensity sessions.
I’ve been telling myself to “trust the process,” but I’ve reached a point where I’m not sure if this is the right training for me.
Has anyone had similar experiences? Am I doing something wrong, or why isn’t it working?
Hard to tell whether this is just a case of the system meeting you where you are currently at or if you have made an incorrect selection when setting up your plan….
I think your FTP is probably in line with a month off and then doing 2/3 weeks of base with you getting sick. I’ve had to miss some weeks for works and my aiFTP only increases a couple watts. Really shows how consistency is important for gains.
As for the volume thing and no endurance rides. I would reccomend going into your plan and hitting check volume. If that doesn’t add anything, just start adding some extra endurance volume on your own. A lot of times the system likes to see that you can handle the work before it ramps you up.
A lot of newer users come over to TR and are like I trained 6-12 hours a week last year, but that was volume not necessarily all the intervals TR does. I often think the system wants to prevent people from burning themselves out and can progress a little slower than normal on volume.
That said if you look at it from the AIs view, you’ve just started training again and you’ve already had to take a week off. It’s probably going to fair a bit conservative since it doesn’t seem like you can handle that volume even if you were sick. I would certainly add that annotation on your calendar if you haven’t for time off.
I take the option that Trainerroad advised me when i created the plan this was 3 Workouts a Week 2x1hr and 1x1:15hr.
Now i tried the Check volume and it added 15min on the 1:15hr ride.
I don’t know is it allowed to post my calendar here?
The strange thing is the whole week the predicted FTP was 293w, yesterday and today is rest day and when i looked today in the app it was at 273w. I don’t understand why.
The intensity of the workouts is quite hard, i don’t think its wrong from that part, but in case of classic training the endurance part isn’t aviable. In my thoughts i think 2 hard workouts and a longer endurance ride with about 2-3hrs would be better for the development.
Now i had around 1 month left at indoor cycling, when i get outside I don’t wanna ride every day hard, i want to enjoy cycling also on longer coffee rides. Maybe this help to understand my thoughts.
Sure that would be nice, i can‘t post it because i have no computer by the hand at the moment, and on the mobile device i can‘t make a overall screenshot.
After taking a one-month off-season break, I restarted with TrainerRoad in January. The AI gave me an FTP of 288W.
This value is from the old AI FTP Detection, so we’ll have to look at things based on the new AI FTP Detection (recalibrated FTP) of 274W.
Based on the screenshots you were able to show, going from 274W to 273W with being sick at the end of the 28 day window between predictions makes sense.
Now, from the current prediction of 273W to 280W (as of today) is also making sense It is still a ways to go until your new AI FTP Detection on March 18th… so, between now and then it may fluctuate like the weather. As it gets closer the number should become more accurate.
That said, I will look into why maybe the value was showing 273 vs the 280 today to make sure all is well!
According to the Podcast the calculation and calendar are updated at Noon each day so there is that little mystery in the equation.
But it is also updated after/when you do a workout or change your planned future work. So if it updates at 11:12am because you rated your morning workout Moderate… what gets updated at Noon?
If you do your planned workout at 6am with no other workouts for the day… why would it ever change after you rate the workout? Is it worried or expecting you to throw in a random workout? or is it looking in history to see you do gym work on that day 11.5% of the time in the evening and might by chance do it? So at midnight it gets a confirmation….”ooh , nice day is finished with no random workout today… update FTP because you are fresh to do good work over the coming weeks”
I have the same issues follow the plan, maybe at a 1-2% to intensity here and there… respond as expected to workouts. Then a couple days go by…. prediction drops. ??!?! Not sure without inputs… why output would change.
TR used to never use HR data, even talked much about how it is fairly worthless…. NOW they keep talking about on the podcast, but not really explaining its use fully. Makes me want to not wear one almost at this point, to reduce error and AI.confusion. (especially with no checkboxes for heat training, fan off, hungover, hot/cold, caffeine etc)
Same results for me, worked my ass off durring the winter/spring, says mine is 210! And i just raced a 1h long crit this weekend with a NP of 303! Cmon’ TR
Huh I don’t recall them saying noon specifically at any point, guess will need to find that for the full context.
If you didn’t have power they were using HR before as far as I recall, maybe for fatigue detection? Not sure what specifically but they would still suggest people use it if they didn’t have power, it wasn’t completely ignored from what I recall. . If you had power it wasn’t using it for anything as far as figuring out if you did the workout as prescribed or for any of the adaptations. They did however have the data the whole time and have said a few times that they still thought it wasn’t going to be useful but once they applied it to the model it increased the accuracy. I’m not sure how you can say that not using it would reduce error when they have said the exact opposite.
They did say it wasn’t needed quite a bit in the past, so yeah I get that the flip flop is unexpected but they have in fact talked about it several times on here and the podcast.
In the discussions explaining new features and new AI recently with Nate.
I gathered they were using it yes for trying to infer efforts of indoor rides with power/HR… with ride outdoor without power. 20min SwSpt efforts indoor gets HR up to 150→160…. so then we see 15min outdoor without power at 155…. must be at Sweetspot.
Also made allusions of things like flat non-responsive HR or very high HR for a given power compared to before. Etc.
Maybe it was my jumping to conclusions… but Nate and Johnathon have been talking a lot about the new AI and HR… weird. maybe good. maybe
Ah if it was most recent one that would explain it had it playing on youtube while working so barely paying attention. Just added to queue for the drive home will listen again.
