If you are using adaptive training I’d leave it where it is and let it figure you out. If you raise your FTP manually you run a real risk of starting to fail workouts. Are you accepting the adaptations suggested for you?
But that’s the strength of AT. If you fail workouts, it will adjust you down and you are good to go again. Alternatively, accept the lower FTP and you will get higher PL workouts. Alternatively, don’t change the FTP.
You’re gonna get workouts that are in the right ballpark one way or another. So no need to worry too much at all.
Make the hard days hard, and the easy days easy. Follow a plan. Be consistent. Do those things and you’re gonna be doing really well without any need to really worry about what the FTP input is. It will sort itself out over time.
2 Likes
Agree. However if he raises his FTP and starts to fail and doesn’t accept the adaptations suggested (sometimes we see those as negative comments on our fitness I think) - then the machine can’t help him.
Does anyone know how the RPE feedback after workouts is used by machine learning to adjust - or is it used much at all? Does the machine learning look at accomplishment in a running way to make future adjustments ?
Sure - it is best to accept adaptations. Or adapt yourself. I do both. But not actually necessary.
The system is pretty (rightfully) conservative on RPE adjustments. It is that + natural decay rates + changing fitness that will eventually get the PL back on track. The biggest drain there is the mental. With a possible negative physical toll if someone is being stubborn and going very hard/all out before bailing. So there is a user dependency on how fast things get on track. But there are guardrails.
The thing is - even if you don’t know your VO2MAX or FTP, your body does. And TR helps implicitly pick up on those over time based on every workout - much much better than any single assessment. And there’s not any magic there. There is a guess at what you should be able to do based on what you’ve done, feedback on if it lined up with expectations, and comparing of how well you actually tracked with expectations from power output.
Anyone can choose to fully ignore the information provided. Or take it very literally. Neither is a great idea. But you are still better off in either extreme case in an AT world than not.
Quite big actually. And this is independent of how you determine FTP. You have a statistical measurement error.
- Most power meters have an accuracy of +/- 1.5 %, although some have +/- 1 % while others +/- 2%.
The rest of the errors are systematic:
- I cannot quantify the error, but quality of sleep the night before (and actually, nights before) I test have a huge impact. Ditto for all other life stresses that influence my daily form.
- All methods to measure FTP are indirect: with a 20-minute test, the recommendation is to subtract 5–15 % — that’s a span of 10 %. Similarly, the ramp test where you compute FTP from MAP has a similar range of, I think, 68–78 %. Note that a TTE done by the book includes a ramp test to gauge the power you should shoot for, so that isn’t any better.
In scientific publications, AFAIK you always include the protocol you use to deduce FTP and it is understood that you cannot directly compare FTPs measured in different ways. So, e. g. FTP20 is the actual power from the 20-minute FTP test without any “correction” subtracted. Ditto for MAP, that’s just your maximum aerobic power in that test. IMHO a TTE is a test for a different metric already — which can be a useful performance metric, but I wouldn’t use that number as the basis to scale my workouts.
To compute the total expected error, I’d have to know more about the distributions, etc. But I reckon it is easily 5 % or more. For reference, the last time I had to manually tweak my FTP, it was by 2.4 %. If you include an error of +/- 1.5 %, that’s not too much. Fortunately, the fudge factors for one fixed athlete doesn’t change too rapidly over time. And with experience, I find you can pin point your FTP with ease. Over-under or threshold workouts are a good option to determine in what direction and by how much you should tweak your FTP value.
The method you use doesn’t matter much here, as long as you have garnered enough experience with it.
Just to add to this: I find that AT has had a very positive impact on my workouts. I think I understand the system and I have learnt to trust it to some degree, but know enough to override it when it doesn’t. Overall, I’m a fan, because it works for me.
That being said, yes, AT can mitigate an incorrectly set FTP to some degree. But regular testing (and tweaking) does really help AT to do its job as well as it can.
1 Like
This sounds pretty normal. I started as an indoor novice (retiree) in March. My FTP has gone 162-172-184-192-196-192-196-202.
The drop gave me chance to consolidate and be ready to improve further. I don’t really mind what the absolute numbers say, I’m working, I’m keeping fit, I’m having fun. All is good.
4 Likes
Assuming we are using a precise measurement - which is reasonable - you can ignore PM error. The absolute number doesn’t matter to me. Just a consistent offset.
I’d agree on say ± 5% for a 95% confidence level masses have on their FTP.
-
I think that’s huge error! Riding at 105 of FTP or 95 of FTP are very different than 100 of FTP.
-
AT handles this error with ease! I’d argue you could be up to 10% off on FTP and still be ok depending on the training focus (SSB I’m looking at you). But as long as you are ± 5% you shouldn’t worry at all. Which means you can pretty much trust AT to get you in the right ballpark and go from there if you want.
-
TR makes no explicit mention on this error. They vaguely talk around it. But it’s an important point in the present example. It is very possible for FTP to go up, but an assessment to go down because the fitness change was less than error of an assessment. And fitness increase does not always equal increasing FTP (this point TR does discuss better than the former, just not as clearly as they could).
My critique of TR on this matter is not that I think they are nefarious or have a bad system. It’s that I think they are missing a great opportunity to highlight a strength of their system. If you are an inexperienced user/new to structured training with power you can set it and go.
Now I’m at the opposite end and still love it. Because I understand it enough to work around some of the rough edges and also to sort through the imprecise language used. It is that middle ground where they are lacking. The people that kind of get it, but don’t know enough not to take everything very literally. These could also use some help by TR messaging and it would greatly improve trust in the system.
1 Like
Absolutely. And I’m not aware of any measurement method that is more accurate, as easy as a ramp or 20-minute test and as easy to understand as either.
I completely agree that AT can handle that, and that is one of the reasons why is such a big deal IMHO. Plus, AT can accommodate differences in power profiles, e. g. if I have particular strengths or weaknesses. Still, in my experience, training is nicer if you keep an eye on your FTP and tweak it or retest it if necessary.
The topic of FTP tests comes up regularly in the podcast, and while I don’t think they have phrased it the way we have, they have put emphasis on athletes tweaking their set FTP for optimal results.
They said they are working on an FTP estimator, but there is no ETA on that. It sounded like a longer-term project and not something they intend to release soon. I think they’d want outdoor riding data for sure, because with indoor rides, I am “limited” by whatever TR selects as a workout for me.
Yes. This. Such a good point.
I think one weakness of TR is fitness assessment: now that we have different dimensions of fitness with progression levels for different power regions, I think they should work on performance metrics for users that are linked to the training plan they chose (and perhaps some other metrics).
Even the humble W/kg isn’t emphasized as much. I have seen several posts where people were disappointed with how their FTP has (not) grown during a training cycle, completely forgetting that they have lost 10 kg in the process and gained all sorts of other fitness.
I have sent a question to the TR team for the podcast along those lines, but no takers. Perhaps I should write to support again. I don’t think you can find a history of your progression levels.
1 Like
I read the first 50 odd posts in here. I think almost everyone in here should listen to the recent Fast Talk labs podcast on “Is Perceived Exertion the most important metric?”
If you all are training exactly to your tested FTP from x weeks ago - especially mindlessly in ERG mode, you can easily be leaving gains on the table or start to fail workouts. I did this for far too long and felt I was getting nowhere until my coach finally sussed out what I was doing wrong. FTP tests kind of get you “in the ballpark” of where to train. It is helpful but not everything. You should actually develop a “feel” with experience for what is a legitimate vo2 max level intensity vs. threshold vs. SS or tempo etc.
My suggestion would be to leave your FTP where it is from when you tested it, (setting the ego aside - I know it’s hard!! 
) then do your workouts (at least the steady state ones), with the ERG mode off - i.e. in resistance mode. For example - in a 4x4 vo2 workout, keep to the suggested power level for interval 1, assessing how it is feeling. Was it really hard so that you almost couldn’t finish the 4 minutes as it should be, or did your power drop below the blue line? Was it too easy? Adjust the next 3 blocks accordingly.
3 Likes
I know there are those that are not fond of erg mode…but what is described can still be done in erg mode via intensity on the interval.
1 Like
I’m not fond of working out on the trainer, and when I do, prefer to have it behave like doing a workout outside. When forced inside, my legs and brain are a lot happier with a smart trainer and sim mode! Set it, forget it, and ride on! If you love erg, and it motivates you, just do it!
2 Likes
Thanks for sharing your experience
There is nothing wrong with ERG mode! I like it too. My suggestion to go by feel in resistance mode was a temporary thing to do in situations like the original poster’s, whose ramp test didn’t feel accurate.
Or just anytime you’re stagnating, in general.
1 Like
How do you fail a 20 min ramp test? My experience of this has been over estimating my ability, then just getting slower, and slower, and slower… but I didn’t stop. So does “failing” mean actually stopping, saying*%£$-it and climbing off and having a strong drink, with the intention of having another shot(not the drink) the next day?
Or do you mean you were so demoralised by your effort ,you vowed never to do another one? 
I think many people would have a higher avg power for the 20min if they evenly paced their power (basically stable, flat-line power) or slight negative split (more power in second half) vs going out too hot and slowing down over the course of the test. At least that is my experience and pretty much everyone I’ve talked to.
1 Like
When you pace too aggressively or way too conservatively. It doesn’t mean you cut the 20-minute effort short necessarily, just that your average power is below what you could have done, had you paced more conservatively or more aggressively.
2 Likes
I think there is maybe confusion by the poster on “20 min test” and “ramp test”.
2 Likes