Hello everyone,
This is my first post on the TrainerRoad forum. I’ve learned a lot from this community over the last year, so thank you all for the informative discussions.
I am looking for some perspective on my recent experience with AI FTP Detection. For background, I used TR structured workouts for over a year with good results until a four-month hiatus due to a foot injury last fall. My FTP was 351 when I stopped riding. I maintained cardio over the break using a rower and ski-erg.
I returned to the bike in January to start a Gran Fondo mid-volume plan for Leadville. Despite the break, the AI predicted my FTP remained at 351. I found the subsequent workouts nearly impossible to finish without dropping intensity by 10%, so I manually adjusted my FTP down to 331. At that level, my performance and heart rate zones felt correct, and I successfully completed a full base round.
Out of curiosity, before starting a new base period, i turned on AI FTP. The tool suggested a jump to 384—a 16% increase over what I feel is a near perfect FTP. This would require me to perform sessions like 4x21 at 94% off the bat (because my schedule didn’t adjust), which is currently unattainable. I have since manually set my FTP to 341 (a 3% increase). Again, my scheduled workouts have not adjusted.
Questions for the group:
-
Does it seem realistic for an FTP to remain at 351 after four months off the bike?
-
Does a jump to from 351 to 384 after two months of consistent base training seem accurate? I am surprised my recent data hasn’t been weighted more heavily to show that even 351 is too high.
From all the reading I’ve done, it seems FTP and AIFTP are not equivalent. So AIFTP may predict a value of 384 but prescribe workouts that are perfectly dialed for my current fitness of ~341 FTP. So perhaps I need to just accept 384 and let it do its bouncy-bouncy thing to give me ideal workouts that will help me progress.
I am not chasing a specific number; I am chasing progress and am concerned that accepting these inflated suggestions will lead to failed workouts. I’ve reached out to TR staff, but while I wait for a response, I would appreciate any insight or advice from the community.
What should I do? Another round with FTP at 341? Or allow AIFTP to do its thing?
Best,
Devon
P.S. Here is a link to my calendar: TrainerRoad
1 Like
I don’t think it’s probable that FTP remained constant with that large of a brake. Is Planbuilder prescribing those double workouts? It looked like some days had 2 fairly intense workouts on the same day.
If you’re referring to Feb 19, 24, and 26: I added a warmup, cooldown, and warmup, respectively. Cold-pike jumps almost straight into 94% and keeps you there for 84 minutes with 1 min breaks, so I’ve found self-added warmups beneficial. For pointer, I took a 30 minute cool down to spin thr poison out of my legs.
Good lord. 
You’re demolishing those workouts at the FTP you’ve been using. The problem is that you’ll run out of hard workouts (this is partly why you keep getting Cold Pike).
It’s a tricky call, because if you accept the FTP Detection you’ll end up with SS/Threshold workouts that have shorter intervals at higher power, which may or may not be what will progress you faster and regardless of that may or may not be what you want.
But I think you’re probably going to have to increase the FTP you’re using in order to get decent VO2 Max workouts.
Final point. Rather than manually adding a 30 minute warmup to a 90 minute workout, just tell it you want a 2 hour workout and you should get a decent warmup. 
1 Like
Dont just accept it… celebrate it!
1 Like
Thank you for the suggestions! I bumped 3% and will see how this week goes. With a max hr of 189’ish, I feel like vo2 work should get me into the 170s and even 180s. With pointer, I felt the workout was hard, but I only remember hitting low 170s at the end. My legs sure as heck felt it though, haha.
I know people say “don’t go by HR” because of variability. I’ve been very careful with sleep and fueling in hopes that I can keep variability low(er). Mostly I’m looking to see normal HR drift over the workout and no decoupling by the end. I’m also monitoring my breath. Pre-panic-breathing starts at “very hard”-workouts and turns into panic-breathing with “I feel like this might kill me leg burn”-power output over longer intervals.
Regarding higher power over shorter intervals: I’m on board for that. But I do see the value in those long SS sessions. Durability was (is?) the buzzword, and I’m trying to get stronger but want to be able to hold higher power for long periods (say, 9 hours ;)). The two are interlinked right?
TR does seem to making workouts longer (it’s topped out at 1.5 hrs right now), but I’m not seeing it add anything on my days off. Is that because I’m on a mid-volume plan? Should I consider 1-1.5 hr endurance rides on those days? I want to make sure I’m healing up on the off days so I’m good for the hard days. But I also want to get plenty of volume. Maybe Sunday’s will become long days (3-4 hours) to make up for having m-w-f as rest days, at least initially.
I really appreciate the feedback thus far. Thank you everyone!
At those power levels, it would be surprising to have the gain you’re describing. Did you switch trainers or power meters?
No, I didn’t change the meters or any other piece of equipment during that last block. And nothing inside the block would have tested me in a way that would hint of an ftp north of 340-350. 384 is lala land.
Unless… I’m comparing apples and oranges. It could be that 384 just cracks open a more varied set of workouts that allows me to progress better than stagnating on Cold Pike at a much lower but more accurate FTP (in the classical sense of the acronym).
I still don’t fully understand how to use the new AI FTP beyond just strapping in for the ride.
Cheers, and thank you for the response!
After a break of that length your FTP has most certainly gone down.
We all know those massive increases in FTP are not possible thinking of traditional FTP, people are still quoting WKG and its even used in TR marteting blurb. Its not FTP, soon enough I’ll be claiming to be 5 wkg on club rides and get lashed out of it by cat 1 who are genuinely 5wkg 
Ive just had a 10 watt increase over 4 weeks with low compliance and low volume, in 15 years of training that has never happened.
But workouts are spot so maybe you need to accept the inflated values and see how it goes.
For me it appears im in an endless loop of threshold workouts of 3.3 _ 4 level workouts capped at 1 hour. Similar with other zones. Unless I mark a workout below the predicted difficulty im likely to stay in this loop. Which is unlikely given the AI is pretty good at predicting difficulty.
Because we no longer have the old ramp ftp result comparisions are no longer possible.
2 Likes
You’re at the top end of the workout ranges for your FTP. To me that reads as your FTP being too low.
Take the TR prediction - it should drop your workout levels such that you get higher efforts for shorter durations.
But given you’re maxing it out - your jump is likely to be quite large!
1 Like
… and so a pragmatic approach for the next 4 weeks might be to split the difference and set a manual FTP half way between where they think they are now and where it thinks they are now.
And maybe rinse and repeat the block after as well.
2 Likes
Your current FTP definitely doesn’t seem too high based on the way AIFTP currently calculates it.
I think part of what you are observing is the move from old AIFTP to the current one. You seem to fit into the bucket of athletes that would have received a large FTP increase with the new system with respect to the old one.
So the AIFTP bump going to the new system is balancing out the detraining when it comes to the reported FTP value.
i.e. Its likely that had you been training at the time of the system change then the AI would have given you a jump in FTP to around 380w.
1 Like