Hey @wintermute! I’ll step in for Ivy while she’s out recovering from illness.
Do your post-workout surveys impact AiFTP?
No, your post-workout surveys won’t impact your AI FTP Detection in the context you’re describing. They will, however, provide important feedback for Adaptive Training to better tailor your training plan to how you have been feeling and performing.
Indirectly, as you work through higher Progression Levels, you’ll get faster, which will probably result in a change in your FTP. As you train, AI FTP Detection will look at all of your rides – both indoors and outdoors. As your FTP changes, your Progression Levels will change as well, so Adaptive Training will keep you dialed in with the right workouts for your current fitness levels.
Hope that makes sense! Let me know if you need any clarification and keep up the great work with your training!
Yes, it does. I get really accurate HR data from my watch and it’s what I use both with the Edge and TR. I’m just not sure if you specifically have to use a chest HRM to get the FTP and other performance estimations/metrics that are not just calculated as 0.95*20min Power etc
Keep in mind that your Progression Levels are relative to your FTP. As your fitness increases (expressed in your measured FTP), your Progression Levels are recalibrating in relation to that specific fitness.
This way, the intensity and overall difficulty of your workouts are always reflected accurately through your Progression Levels.
Let me know if you have further questions about that, I know the FTP to PL relationship is a bit tricky but we’re happy to provide clarity and context when we can!
Please keep @'ing us! I’m glad Zackery was able to step in, but please feel free to mention us whenever you have a question we can help out with. We love it!
Update: I’m just going to trust the AI from now on… I’ve just had a three week break from the turbo during which I did random exercise in the only gym I could access, or no exercise. Restarting, the AI predicted a new FTP 213 and, considering that threshold sessions set at 207 had been a struggle in December, I was doubtful but thought I’d give it a try. For my forst session, the plan appropriately shortened the o/u intervals and so the session was very managable, even at higher FTP. Moral of the story: breaks from exercise can be helpful and the AI prediction knows what it’s doing.
Latest update is that AI FTP continues to provide me with 295/296 which appears high for me personally although I have in the past trained off this range.
So just finished SSB1 HV which has taken Sweetspot PL from 3.5 to 5.2. I usually replace one of the weekend workouts with a long endurance spin and drop the Friday endurance workout.
All sweetspot workouts were marked as hard (some bordering on moderate) to be conservative.
So progress through the PL is slow and will continue with SSB2 HV next week and hopefully finish that block with PL closer to 7.
As a training number and the feel of the workouts being assigned I am pretty happy overall with AI FTP and AT.
Wanted to make an observation or provide a data point.
I’ve been doing a year of trainerroad Low Volume. I’m the guy who gets the workout, does the prescribed workout in Erg Mode, rates workout appropriately, and have been extremely consistent. I’ve always wondered about the disconnect between the estimated FTP in Intervals.icu, more recently mFTP in TrainingPeaks, knowing there was a discrepancy as I literally had done almost ZERO max power efforts at any interval.
A week or two ago AI FTP raised me to 291. Today, I did my first 1 hour power / TTE test. Did it at 278W, finished the workout strong, was upping power at the end, even threw in a quick sprint where I maxxed out at close to 900W for 5s even though I forgot to turn erg mode off (Yes, it caught up)
But anyways, I felt strong, I left feeling like I could go back and complete that 1 Hour at 285, not sure about 290 though. Maybe.
But, has me feeling like for me, AIFTP is pretty accurate based on that test.
It’s an easier week for me and then will be doing 5m, 1m, 20m, and sprint tests. Never done those before, and have spent almost no time in resistance mode.
Anyways, so far, my thinking is I’m close, but give me a week…
If you have enough gas left for a sprint at the end of a 1-hour effort, I’d say 278 W is significantly lower than your FTP. I wouldn’t be surprised if you have an FTP that is 5–10 % higher than that. (My rationale is that 278 W sounds like an easy sweet spot effort, and easy sweet spot is around 90 % FTP.) The 5 % figure matches with AI FTP’s prediction within 1 W.
In my experience, AI FTP tries to be conservative with its estimates, i. e. in my experience it tends to err on the lower side.
I came in real fresh, good sleep, well fueled, and motivated to crush the workout.
AIFTP has me at 291, so this was slated as a 95% of FTP effort which started me out at 276 (Keeping in mind, I’d never tried to ride an hour interval like this, I figured I’d see how it went and adjust in ride). After 30 minutes upped it to 279, in the last 10 minutes started bumping it up into the 280’s, and finished with a minute or so at ~300-310. Then threw in the sprint after a couple minutes of recovery. NP came out at 278 for the interval.
I “feel” like 291 is close. When I do over/unders, it feels like I’m clearing under, building over.
Going to do some more tests, but I feel like for me I’m right in the ballpark. But then again, I don’t know if I’ve ever failed a workout, so maybe my limit is a little higher than I think it is and I could push myself a little more?
I’m just going by your description. I don’t think you’d be able to build an effort and top it off with a 900 W sprint if you were close to your actual FTP.
I like that you are verifying your numbers with workouts. Nevertheless, I’d encourage you to try to bump your next threshold over-under workout by 3 % and see what happens. I wouldn’t be surprised if you could complete it based on the numbers you presented in this thread.
If that’s the case, I’d definitely try to push myself — not for the sake of “getting a higher FTP”, but experiencing what too much feels like. You gotta cross the line to know what too far actually feels like.
(In regular training it isn’t bad to keep something in the tank. IMHO that’s a good thing in most circumstances and helps you stay consistent.)
Had a very interesting FTP experience today. Just finished SSBMV I, just easy rides this past week. Did AIFTP this morning, gave me 290 up 2W fro 288 (whatever, didn’t bother to accept it). I recently have also been tracking on Xert, and have been using Goldencheetah for years. Xert previously had my TP (xert FTP equivalent) at 260. Goldencheetah had CP (we’ll just assume that’s basically FTP equivalent) at 269.
I did Kolie Moore “baseline” FTP test, cutting power back to 96% thinking better to start conservatively. First 10 min were fine. When it bumped up at the 10min mark, got tough, and I was switching back and forth between ERG and slope modes. I got through the next 7 min near target, then had to cut back for a few minutes to near the original 96% power (276) or below. Then at the end of the 15 min block, tried to ramp slowly. HR got to 155 which is 90% HRmax for me, at basically low end VO2max HR. I managed another 5 min, getting to about 300W, but could only get HR to 158. I had to quit at that point, basically due to burning lungs despite I had a goal to get to 165 which is 95% HRmax, but couldn’t make it. The average was 274 for the effort, which Moore says I should take as FTP. Xert gives updated estimate of TP at 265, and GC give CP of 267
all that said, in SSBMVI, I didn’t feel the workouts were too hard, most I rated “moderate”, and I added a bunch of Z2. So it seems AIFTP is ok for setting TR zones, except I know from previously doing some TrainNow VO2max workouts, I have to turn that down 5%
anway, tldr; AIFTP is higher than my actual sustainable power, but seems to be ok for setting zones for TR workouts, except near VO2max
Not sure how that will impact my planning for the next block, but that would get off topic
Here are my simple thoughts - just thoughts no deep dive into scientific studies.
I am not my FTP
My FTP merely sets my training zones
I use progression level as a visual progress toward creating less brittle fitness.
I have seen my fitness become more stable with AI FTP and Progression level because I fail less workouts- therefore I’m more motivated and consistent with training.
For me it’s not more complicated than that.——
However I’m an average Joe.
One point of doing this test is to learn to “feel ftp” which is the tipping point between stable and unstable HR/breathing. Erg is NOT recommended.
Sounds like you should have backed off before 300W, because your HR (and breathing?) were increasing.
Why did you have a goal of 95% HRmax? If you believe that is vo2max HR, why would you want to have that as a goal when riding at threshold and you HR and breathing should be stable?
It seems unlikely your threshold HR is 95% HRmax. I’m an outlier with a high ftp relative to vo2max. And my threshold HR is 161bpm on max of 175bpm, or 92% HRmax.
I ride with people a lot faster than me. I’m old and slow. They force me to figure out that tipping point between:
stable and going for 40+ minutes
and
unstable and blowing up
Not more complicated than that. Really. Its not about have an FTP too high because some I’ll make up for it by lowering intensity on my workouts.
This wasn’t a threshold ride or testing TTE. It was the “baseline” FTP test. As described, after the second 15 mintes at target threshold, the idea is to slowly ramp up over another 10-15 min “until exhaustion”. I feel like that should be getting near max HR, no? Even if I didn’t ramp up, after long enough, even threshold becomes VO2max, no? So by quitting before HR got to even 95% of HRmax, I felt I was quitting early, really for no good physiologic reason, just mentally couldn’t take it anymore. I’m not at all saying my threshold HR is 95% max. I’d say it’s actually around 145-150 which is like 83-85%. And I switched out of ERG, maybe you missed that bit.
Anyway, I’m not quibbling with the test…274 “feels” like my FTP. I’m saying that AIFTP (which is saying it’s 290) seems to be ok to set levels for TR workouts since I can successfully complete them at a generally “moderate” to “hard” RPE with it set there, but it ain’t my FTP
Maybe I didn’t express myself clearly, I think your comments were somewhat beside the points I was trying to make. Appreciate the input though.
edit: just re-read the “new FTP testing protocols” argument, and he says “everyone who prefers the baseline or progression 1 test sets a new 20 minute power”…well, despite consistent training, I did 295 for 20 minutes in March of 22, and 267 for 60 min on Dec 4 which was at Alpe du Zwift at the end of the “4 horsemen” route – already climbed “hilly KOM”, “volcano KOM”, and “epic KOM”, so 274 for 35 min essentially fresh does seem low, but…for whatever reason, don’t think I could do better at the current time. Maybe that’s a knock on SSBMVI
Looping back to this with some data. Started working with a coach, so did a series of max effort tests as input into WKO5 to get a better estimate of my FTP and structure an upcoming training block.
TR AIFTP has me at 291W.
Mentioned this earlier in the thread, but first did 1 Hour @ 278w and finished real strong, room for ramping up the power and a sprint test after a couple minutes of recovery. Definitely not a max effort.
Recovered and then did 5 Minute Power (383W), 1 Minute Power (599W), 20 Minute Power (317W), and a Sprint Test (1010W). Sprint test was the only one that I think needs to be re-done, inside seated on a trainer and not feeling it that day. But, not sure that’s relevant to FTP anyways.
WKO5 has me at 295W mFTP. Intevals.icu has me at 308w eFTP. 308W feels too high - that’d be an “over” interval based on feel.
Net - looks like AIFTP is a little bit of an underestimate for me, but not by much. Pretty close. Going to be training at the 295W FTP level. And who knows, maybe AIFTP will update based on those max effort tests too, have another couple weeks before I can re-run I think.
Separate from my results above, I’m going to share my theory on TR AIFTP.
I think there probably is a physiological component, but I think that a lot of it’s accuracy will come down to garbage in, garbage out.
In my case, I’ve been using TR for a year now, and have been very compliant. I do my “workouts” in TR, on the Trainer, in Erg Mode. Head down, hit my targets, rate them appropriately, recover well. On top of that, I’m not blowing myself up outside or with other work that would impact my ability to do those workouts, and I’m doing a good mix of VO2 Max, Threshold, and Sweet Spot. (I do endurance and fun rides outside, but am careful not to impact workouts). Basically, I think if you feed the model good data, compliance, all that - you’re going to fit the model better and get a more accurate result.
If you do lots of workouts outside, and “match” workouts that may not really be a match, don’t hit power targets or TIZ as closely, or do endurance rides that are actually threshold, VO2 Max, aggressive group rides, etc. etc. - TR is going to infer one thing, when the reality is actually something else. That by nature is going to result in a worse “Fit” and more inaccuracy in prediction.
Given your 5-minute power, 308 W is not an unreasonable guess, although on the high side: 383 W/1.25 = 306 W.
However, I’d definitely err on the side of feel, and if you say that 308 W feels too high, but, say, 300 W, feels about threshold, I’d stick to 300 W. Trust, but verify.
No idea on time interval for eftp. Haven’t changed anything from whatever the default is in intervals.
For what it’s worth as context to my numbers, I’d consider myself more of an anaerobic athlete. Strong, more muscle and heavier than a typical “cyclist physique” I’m 5’10" / 178cm, 177# / 80kg and decently lean. Pretty easy for me to start putting on strength and muscle if I start doing squats and deadlifts, have a background in ski racing. If I have a weakness, it’s probably endurance and sustained power. Not sure, but makes me suspicious that my shorter duration efforts could skew my FTP artificially high which is why the 295 feels right.
Training for Leadville, at least I picked a race that caters to my strengths
(Side note, need to drop a little weight over the next 6 months!)
This comment aligns with my suspicion, I was going to say based on your 1min and 5min power and the drop off to the 20min that 308W seems optimistic.
You might cope fine with it as a training number, especially if you’re focusing on shorter intervals. However, with Leadville being the goal I’d stick with your ~290W and focus on building interval length.
