I started incorporating strength training twice a week into my training and felt like it was going really well. I have had a few months of consistent riding and ftp jumps. When trainer road AI launched I got knocked down from 213 to 208 which was no big deal, my prediction for the next detection was 220 or something like that which seemed great. I got sick in the middle of the month and had to cut a ride short and missed one but my prediction was still 211. This week I think I had a hard ride Saturday and then I think I overdid the lifting a bit Sunday. Today my legs were completely dead and I failed a sweet spot ride 20 mins in. This knocked my ftp prediction all the way down to 196. I know I am probably over training but it feels like I just ruined my whole month in a three day span.
My lifting volume has been 8 working sets on legs on Sundays and only three on Wednesdays for lower body. My cycling volume has been pretty steady at 3.5 hrs a week but I did recently finish a masters plan and started a normal plan this week. Is this likely just a case of over training? Not even sure why I’m posting this because I think I know the answer but a 15 point drop in ftp prediction in one day was pretty demoralizing. (not blaming the software just feels like I ruined my whole month in a three day span)
1 Like
I don’t really know the answer to this but here are some theories based on what’s I’ve been skimming from the forum and podcast discussions:
- The model has correctly lowered your FTP because of lingering effects from sickness, weight lifting, and maybe other things you haven’t realized (diet, aging, stress, etc.), and:
- The lowered FTP is here to stay for awhile (months) because you take a long time to recover from sickness/get sick again/decide that continuing strength training is worth depressing your FTP rate of gain.
- The lowered FTP is only temporary (weeks) because you recover from sickness and your body gets used to strength training (and/or you schedule and dose it appropriately)
- The model has incorrectly lowered your FTP because
- For whatever reason, it’s too sensitive to recent data it’s received in your case
- It hasn’t received, or wasn’t built to incorporate, enough detailed data about your sickness and strength training and is effectively going with a worser-case scenario
- The model has correctly lowered the “TR AI FTP” but maybe not the “FTP” number you’re used to because:
- At the old FTP number, you’d rate the RPE for a threshold workout level 3 as “very high” instead of just “high” these days. However, your body can still do the work it did before and you’ll just be more out-of-breath at the end of a group ride.
I have been dipping into and out of threads discussing this stuff so maybe some of these are off. Just taking the opportunity to think through some of the possible explanations here.
2 Likes
Always helpful to share your calendar so we can provide feedback/advice.
Sure! Is screenshots the best way to do that?
1 Like
Here are some pictures from my phone lol
Don’t obsess about the prediction. Are you hitting workouts with higher watts? If so you are gettting stronger. If not, look at what you can do differently. The model is not “rewarding” or “penalizing” you for nailing or failing a workout. It is simply adjusting to give you the best baseline measure to pick and plan future workouts. Think about it this way: “cool I’m going to get better workouts now and not burn out. But let me think about what I can going forward to do better.”
Hope that helps!
Thanks for sharing. The photos definitely helped
- The sickness is going to take it out of you. HR/Power relationship will be off and the model is specifically trained to look at HR & Power relationship. My guess is this being off, as expected, is causing it to be lower
- On top of that noting the workout today as “failed” due to fatigue is going to impact it as well.
- On top of all of this doing 2 strength workouts a week now as a new stimulus may be impacting your recovery.
1 Like
the whole “thing” around Ai FTP has gotten out of hand.
your FTP hasn’t really changed. you ability to express FTP may have, due to fatigue, life stress etc.
your FTP is your body’s ability to produce power aerobically, maximally, without rapid fatigue. its a state within the body. its not 60 mins or 95% of 20mins, or a ramp test result. there is nothing magic about these numbers. when you ride at FTP, you’ll have that “i’ve got this” feeling for many minutes, and that comes at different points for the individual. the reasons for fatigue are many, and more aerobically fit riders can hold FTP longer than others, its time element is trainable, although some riders won’t necessarily benefit from more time at FTP, because of their events. FTP is structural, and requires serious hardware changes to improve, once the noob gains are gone. you should be able to feel changes in RPE if your FTP rises, and that’s the best indicator or prediction tool.
TR has now given us an FTP number which changes at every slight adjustment to training, or plan. that’s a problem, because the TR community hangs on that number. rather than cutting your FTP, the Ai should be spotting fatigue and adjusting training, whilst telling you why. i do believe it is adjusting your training, but by lowering your prediction, and that has gotten you concerned enough to post here.
i moved a few easy workouts about on my recovery week, and my FTP dropped 
with a few easy days, you will likely be back on track.
3 Likes
Thanks! I accepted the new ftp of 196 and now my the prediction in 31 days is 205. I just have trouble wrapping my head around the fact it dropped from 211 to 196 after one failed workout and now it’s not even expecting me to get back to that number in a month. I’ll keep at it though.
I like what cagiva said.
But note that it wasn’t just one failed workout. In your initial post you said you (1) cut a ride short and (2) skipped one. You (3) had a hard ride on Sat and we don’t know what TR saw your HR or power doing during that ride. So it could be that all of that plus (4) the failed workout is depressing the FTP prediction (rightly or wrongly).
1 Like
Yeah I guess it just feels like it was one workout because my Thursday and Saturday rides felt good and my Ai ftp prediction was 211 still then I failed the workout yesterday and it tanked me all the way to 196. I get what you are saying though
If your legs are completely dead, don’t do a workout! You can always push it to tomorrow (assuming your schedule permits). When my legs are dead I either (a) do a SUPER easy endurance ride. The goal is for your legs to feel better after the ride than when you started. Maybe 30 mins or an hour at ~50% FTP (b) if even this kinda mega chill ride would leave my legs feeling even worse, then just take the day off! Recovery is where you get stronger.
I do endurance rides on days where I don’t feel great, but for workouts with intensity you usually want to start the ride feeling good. Usually you can only get bang from your buck from intense stimulus if you’re feeling good enough to dig deep anyway - otherwise you’re just adding fatigue without getting much upside.
6 Likes
Hey @Yeager139 
Lots of good feedback already but I wanted to point out a few things just in case.
The change in FTP from 213 to 208 is a recalibration to the new system. Much like buying a new weight scale, you weigh the same, but the number reads a little different, so any comparisons moving froward should be to the 208 value.
That said, by looking at your Calendar it seems you’ve been struggling with fatigue and sickness over the last month by looking at your post workout surveys, so I would not fixate on the FTP value too much. Plus I already see your next prediction is looking positive!
Moving forward, it may be a good idea to focus on fully recovering energy and sickness wise so you can start training with consistency again. Piling on when we don’t feel good just has a spiral effect, and it’s always better in the long term to reset, feel good and start training again. 
Also, Workout Alternates are a good option to keep training the energy systems meant for the day but at lower intensity if needed.
Thanks! Yeah looking back I definitely should have not tried to push through some of those workouts so good lesson learned for the future. I was mostly just shocked how far the AI prediction dropped after that single failed workout. Reflecting on it I think my current number is actually a better place to be though so I’m going to trust the process. Thanks for the responses everyone!
1 Like
