🎉🎉🎉 Introducing AI FTP Detection 🎉🎉🎉

[gorillaclimber]

Just finished Build 1 (MV), so used the autodetection feature and moved from 356 → 365. Makes sense- I started TR roughly Nov1 with a 315 FTP and have been picking up 20w/ 6wk block (11/2, 12/21, 2/1, and now 2/28), so 10w/ 4 weeks seems about fair.

Progressions unchanged, though.

[gorillaclimber]

@rac1ngsnak3

So what’s the purpose of AI FTP detection?

Not everyone wants to actually complete a ramp test… some would rather accept an AI “best guess” and go on with another recommended workout

[mcneese.chad]

@rac1ngsnak3

So what’s the purpose of AI FTP detection?

• It can give you a reasonable FTP estimate for use in training with the TR workouts & Plans.

@rac1ngsnak3

What can it do better than an actual test?

• Not necessarily “better”, but it removes the need to test, and allows for replacing that test with an actual workout, that may be more “productive” in the direction of the training plan goals.
• Reduces / Eliminates test anxiety that is real for many people.
• It also seems decent at catching declines with lack of training for a range of times and reasons.
• Again, it does this without testing and doesn’t even require “maximal efforts” as is the case with other FTP detection methods.

@rac1ngsnak3

It seems redundant to actual FTP testing.

• It’s only redundant if you continue to test. Some will, some won’t.
• Several people have done the Detection, and then a test and found a range of result. But many have seemingly backed up the detection that I have seen. So, validation in a sense, at least in N=1 for bits of this topic.

@rac1ngsnak3

Are there any papers or science that says AI FTP is better than actual FTP?

• None that TR has shared.

[jjmc]

@WindWarrior

Given some of what has been written, perhaps a good first step is for TR to stop using the term FTP?

What I do know is that a user doesn’t need to pay any attention to progression levels. What a user needs to learn is how to get in touch with their feelings and rating workouts. If progression levels and workout levels weren’t displayed, would anything really be lost? You would still have a list of alternative workouts ranked by achievable, productive, and stretch.

I agree on both these points.

I am a big fan of what PL provides to my training experience, but the least important part is the number. I can look at what is recommended and decide with pretty good accuracy on how doable it is. The bins you note are the key part.

[jjmc]

@WindWarrior

But I know one thing, pre-AT the basic approach of TR Build without enough endurance riding left me crashing and burning. So I suppose AT is a gentler, kinder way of pushing hard?

Same experience. The key for me with AT are two things:

1. I have a better sense of minimum effective dose

2. I have zero fear about mucking with what is suggested and diverging from my goals.

Part of this is that I was just starting to figure things out on my own from mistakes pre-AT. But I still find it very easy to navigate through things and not stress as much if I am doing what I should be to get stronger.

The analysis can and should get a lot better. But it’s already a huge improvement for me. And it’s not from pushing harder - more the opposite. Not trying to drill myself every workout.

[WindWarrior]

Minimum effective dose is an interesting discussion, I wrote up a few thoughts here Sweet Spot Progression - #2478 by WindWarrior and after 2 years of using a “traditional modern” approach to training on 8-12 hours/week, I have a very different opinion on minimum effective dose. All I can say in summary vs TR AT approach, is that I’m doing a bit more weekly time and a lot less intervals (I last looked at TR AT plans 4 months ago). These types of comparisons are always difficult, but this approach I’ve nearly reached peak from fitness from 2017 (before TR).

It sure seems there is a lot of individual response to training, and requirements for dose and recovery. I imagine a ‘more is more’ approach might work if my aerobic capacity was higher, and I rolled out of an off-season with high fitness that just needed 6 hours/week of mostly intensity to quickly push it back up. But I’m not that athlete, and my aerobic base always requires a lot of boring work before I have enough of a foundation to layer on the really hard top-end aerobic power work.

[txrr9876]

@Nate_Pearson

New rider with lots of FTP growth room = Keep pushing up FTP and higher wattages

So much to unpack here and not an easy problem to solve given the heterogeneity within demographics

[jjmc]

@WindWarrior

Minimum effective dose is an interesting discussion, I wrote up a few thoughts here Sweet Spot Progression - #2478 by WindWarrior and after 2 years of using a “traditional modern” approach to training on 8-12 hours/week, I have a very different opinion on minimum effective dose. All I can say in summary vs TR AT approach, is that I’m doing a bit more weekly time and a lot less intervals (I last looked at TR AT plans 4 months ago). These types of comparisons are always difficult, but this approach I’ve nearly reached peak from fitness from 2017 (before TR).

It sure seems there is a lot of individual response to training, and requirements for dose and recovery. I imagine a ‘more is more’ approach might work if my aerobic capacity was higher, and I rolled out of an off-season with high fitness that just needed 6 hours/week of mostly intensity to quickly push it back up. But I’m not that athlete, and my aerobic base always requires a lot of boring work before I have enough of a foundation to layer on the really hard top-end aerobic power work.

Very nice. There are a few layers to MED, and I should emphasize that what I meant was I was messing up the most basic aspect - that it’s not a great idea to completely shell myself every workout. The goal for a hard workout is that it is hard - not very hard or all-out. Better to err on easier side than harder side often too.

Hence, I’m not advocating for any level of intensity that TR is giving compared to other coaching systems. can say that I find myself doing better on less intensity and more volume for sure, with intensity starting to be sprinkled in more as I get into specialty. I’ve been doing loads of aerobic, tempo, sweet spot work trying to maximize my volume in the time I have and results have been very good.

[HLaB]

Some folk test high with the ramp, similarly some folk test low and a lot of folk don’t want to test, especially during the season.

[AlphaDogCycling]

@gorillaclimber

TR “FTP” is not threshold power but just a stand-in for workout completion power

Not to start a philosophical battle, but any “FTP” is just a stand-in for workout / ride completion power. That is: FTP is a metric used to pace, figure out how hard to go in training, how much training stress you’ve put on your body, etc. Otherwise, besides purely bragging rights, how else is “FTP” used in a practical basis?

[jjmc]

@AlphaDogCycling

Not to start a philosophical battle, but any “FTP” is just a stand-in for workout / ride completion power. That is: FTP is a metric used to pace, figure out how hard to go in training, how much training stress you’ve put on your body, etc. Otherwise, besides purely bragging rights, how else is “FTP” used in a practical basis?

Super important point…it’s morphed into a bunch of uses. And it isn’t particularly great for any of those things either.

[lehnerm]

So my ego obviously decided to go with the ramp test result of 260w over my FTP AI detection of 241W.

After a week of training I have to conclude that while I have enough strength at VO2Max efforts to do 6min over-overs effectively that the detection result was a far more accurate in terms of proper training levels. Easily noticeable, since the low level threshold workouts have nice ramps that let you triangulate your true threshold. I’ll be using the AI feature from now on. Thanks for making it happen Amber & team!

Aside from getting a more sustainable training level there is one more thing why. Ramp tests are usually scheduled on Monday for me. I train in the morning and a maximal effort does make the rest of my day noticeably less productive.

[AlphaDogCycling]

@jjmc

Super important point…it’s morphed into a bunch of uses. And it isn’t particularly great for any of those things either.

Practically speaking, hasn’t FTP always been primarily used to set training targets / evaluate how well training is going (e.g., FTP is going up / w/KG going up)?

[jjmc]

Yes, but I guess better way is to say in a more narrowly defined sense that it has morphed into. and even then it wasn’t super narrowly defined.

[liam_mail]

Would this still be a number in Watts? Something you could use to establish pacing plans etc for racing etc?

[WindWarrior]

Yes in watts. For many users TTP™ is simply the ramp test estimate, or moving forward, the AI FTP estimate if you choose not to test with the ramp.

[pstalley]

Anyone else still getting this “hang tight” message when trying to use FTP detection?

[tomclune]

I still get that. But seems to go away after about 5 minutes.

[pstalley]

@ambermalika

I’ve been getting the “processing data…hang tight” message for the last half hour. I wait, but nothing seems to happen.

[ryanppax]

is your birthday set correctly?