Trying to understand Sweet Spot

As a coach, I use the above in practice, largely power distributions from multiple rides almost always including some kind of proxy MAP test (5min power, e.g.) alongside longer form testing efforts of at least 30 minutes (or preferably more) and via WKO5 modeling. Once I have some history on an athlete, I’ll also use HR response to tests and intervals to help me pinpoint threshold and when we exceeded it, etc., but that’s inexact and less about the raw number and more about the trend in longer intervals.


Hey, I’m just a chatbot, remember?



While you’re at it, ask him why Brooks’ study using the lactate clamp provides no support for his claim that elevated lactate lactate levels suppress lipolysis during exericse.

“It’s a challenging dilemma”, I tell you.*

*Kudos to anyone who can figure out the allusion.

1 Like

You probably mean ‘estimate’ but I don’t know why anyone in this community has a problem with a 40km TT once or twice a year.

I mean, are you here to get faster or not?


Recently had a friend tell me his FTP was 343W. I was dubious. Less than a week later he did a 54min TT at 282W (Merckx style). :face_with_monocle:

Race of truth.


This AI stuff is amazing!

If anyone reading does do comparisons of their TR AI FTP and that obtained from another protocol (I’m interested mostly in Kolie Moore protocols), how do those comparisons come out please?

At the moment I am of the view that the TR AI FTP is the right number (for me) to use for TR purposes. I believe it is specific to my TR history on the equipment that I use.

I don’t expect the TR AI FTP number to match up exactly with my on the road 40km TT power and I don’t expect it to match up exactly to the number achieved in any other protocol, or to stay exactly the same if I make a significant change to the equipment set up I use on TR.

Anecdotally, my RPE, HR & breathing rate responses to, say, Baxter and, say, Taylor, are pretty much in line with what I would expect them to be based on the stated nature of those workouts.


Ai FTP last time I was curious (I just clickd it - I don’t follow plans): 302W
Last 40 minute threshold workout (was pretty cooked by the end, and done in resistance mode): 291W average power - don’t think I could have done this 10 watts higher at 300.

1 Like

There are quite a few existing comparisons to review:


Yeah a couple of times I’ve rejected AI FTP and done a ramp test, both times came out the same.


Speak for yourself. Some of us are here to get better at going longer distances. Yes, I’m faster at that since I started structured training. But I still have no interest in time trials. (I’m pretty sure there are racers who like racing in a group but hate time trials–though they can chime in if they want to). There’s a difference also between physiological capability and psychologically being able to make yourself do it. I have not motivation to do a TT, or really to experience the fatigue (in the lay sense) that comes with doing anything in higher zones (er…levels?). I do it during structured training to get faster in lower zones and capable, but I don’t have any interest in doing it at other times. That’s not my kind of pain (my kind of pain comes from being in the saddle for very long).

Personally, I think my FTP is approximately right because over unders feel the way that Chad TR tells me it will in the work out text. It might be a bit off (and it’s undoubtedly over exact) but it’s close enough for training purposes, which is all I care about. I don’t know how accurate a subjective sense like that is.

1 Like

This may vary from trainer to trainer, but I swear that it’s easier to do a set wattage in resistance mode than in ERG mode. At least, I’ve found it easier to comply with a workout structure indoors since I started doing it in resistance.

I wasn’t shooting for a set wattage - I was doing it at an RPE of 7-8/10 and let the chips fall where they may. I don’t look at the wattage during FTP/SST intervals - go purely based off RPE. So I’m not sure what you mean about “setting” a wattage in resistance mode.

Also, TR doesn’t smooth power when doing workouts so you don’t really know you wattage because it jumps around so much during the interval

For any ERG vs RES vs STD vs LVL mode comparisons on trainers, you MUST control for bike gearing which relates to flywheel speed. Unless you match that bike gearing & flywheel speed between trainer mode tests, you are introducing two variables (trainer mode + flywheel speed) rather than one. Proper testing approach is to only alter one variable, so you really know if trainer mode leads to different results.

You mean wheelsuckers? :rofl:


Of course. I do though use the same gearing both ERG and resistance (except very occasionally for things like sprints which I don’t have often in my training plan). I should say, I still use ERG for rest intervals and warmups/cooldowns, and then switch to resistance for intervals so I don’t have to adjust the resistance level too much.

1 Like
  1. it’s fun to read people arguing the “bot” @The_Cog.
  2. for the comparison of different training software. I use TR and Xert at the same time. TR AI says my FTP is 297W, XERT says it’s 277W. I have hit several brakethroughs on XERT over the last month but the FTP will never match. So at the end of the day it seems that the FTP given by software is an arbitrary number and to find the most accurate FTP you should still go out and do a full send for 1h.
  3. also it is possible that people are different and one testing protocol suits better for one person than the other!?

And every workout is carrying a different level of fatigue/ load/ sleep/ nutrition/ life stress etc into it. I rated a threshold workout “very hard”, but the end of week three, a hard race last weekend, a shít week in work (especially yesterday, which led to poor sleep), very warm (for Ireland) mancave this morning etc etc.

I may or may not be able to hold my AI FTP for an hour, or sweet spot based off it - I probably could if I tapered into it for an hours TT, but I’m not really interested in that when that’s not my goal. Since AI FTP, I fail far fewer workouts, once I rate the surveys honestly.

So now that we understand sweet spot (I think), what are we supposed to do with that information? I’m certainly not going to use it as a training zone. How can it benefit my training?