AI FTP vs FTP Tests ... why?

Why waste a day of training if I don’t have to? AI FTP seems to be spot on for me so I don’t see the point of doing a test.


Need to work on that repeatability. :wink: :+1:

1 Like




You sir, are the definition of a masochist. I’m not sure whether to be impressed or disturbed. :rofl:


I hate the tests and rarely did them.

Did a 20min test. It was 1w different than AI.

I’ll stick with the AI for 1w variation.


Whilst I quite liked the Kolie Moore Test, like the 20 minute test you need some idea to get the pacing right. So I eventually came to the conclusion, I may as well self adjust and verify on a workout I know.

The Ramp Test always really underestimated me, compared to what I was doing. And I never really worked out why I’m spinning down after 15 minutes for accuracy, where as it was “don’t stop” on the ramp test!

I actually don’t mind going deep, and am something of a Type 2 “fun” fan, but I’d rather save that for races and events tbh. So give me something productive in my limited training time! I’m really not bothered whether it’s my “real” FTP or not, once (along with progression levels) it gives me appropriate training.

People are determined to find imaginary sticks to beat TR with! It’s getting quite tiresome for me.


Can’t tell if you’re trolling with this one, its literally defined as the hardest you can possibly go…

But I like AI FTP because it let’s me be more consistent with my training. I usually take a day or so to recover from a proper all out effort, so if an all out effort isn’t part of my training plan, I lose a day of training. (I know testing can be used as training, but sometimes it doesn’t fit in my plan).


Yes the ramp test is an all-out test to failure, however it is a substitute for doing an all-out 5 or 6 minute test to failure. If you empty the tank on the ramp test, its a lot easier for me, versus doing an all-out 5 minute test. And a properly executed ramp test is pretty good at estimating your maximum aerobic power.

But the ramp is really not a good substitute for doing a long test.

I’ve been using estimates for 5+ years, and only doing max effort long test a couple times a year. Years and years of productive training without having to burn a mental match on doing a long test. But at least a couple times a year it is good to make sure the brain’s RPE is calibrated correctly, plus its a valuable workout to help with pacing threshold efforts that are the basis of a handful of local Tue Worlds, Wed Worlds, and Thur Worlds.


Not an expert but my pennies worth, I was tired before doing a ramp test so felt it was not a 100% effort so this ai FTP seems to make sense; I still have the option to do one or accept AI FTP. Cheers Phil

I think this also depends on the type of rider you are. I know I’m more of a sprinter/attacker. So hard anaerobic efforts over and over are my strong suit. For me 30/30 VO2 Max session even at very high power output is a breeze. 20 minute sustained power makes me want to keel over. I loath Full Frontal, the 5 minute MAP test followed by the 20 minute FTP test with a short rest in between was the single most aweful thing I’ve done on the bike. I probably misspaced the 5 minute MAP effort, but there would be a big mental hurdle to overcome to get me to do that one again.


Testing is training… (well, except the ramp test… it’s kinda training I suppose, but I digress…)

Well, there’s not much reason to do any kind of FTP/threshold test every four weeks because FTP/threshold just doesn’t move that much that quickly unless you are just off the couch or coming off a long layoff. Most people using TR could probably (and maybe should probably) train at the same FTP for 2-3 months and progress in other ways instead of chasing 2W “FTP increases” via a ramp test OR AI FTPD.

But again, I digress…


Re: not much reason


2 - 3 times a year is plenty for many multi season althletes. After many years you just know from training,… I like to do 35 - 50 mintues at FTP every couple of months because its great training imo and I enjoy it, yes it hard but I can still go for a Z2 run later that day.

I think some people really need to frame testing differently, too much negative mental baggage, setting themselves up for a bad experience or in their head “failure”.

What it really is, is an assessment of current ability how its framed makes a big difference, it is what it is, no need to fear or stress about it.

Earlier this week I read someone was disappointed about a 2 - 3w decrease in their FTP, really? So nothing had changed then, margin of error and much less than daily / weekly fluctuations. (It wasn’t the person in this thread above, they didnt say they were disappointed, it was on facebook.)



All out 5 - 6 minute is awful for almost the entire duration (except the first ~20 - 30s.)

It is only the last 1 - 2 minutes of a ramp test thats really hard, for me it is much eaiser to go there as you know it wont last for long. Not too hard to do an hour Z2 after either after a 5 - 10 mintute break / active recovery.

After a true All out 5 - 6 minutes, a peroid of feeling like death (never again, but it passes) and then 10 - 15 minutes of of Z1 as active recovery is enough, I dont feel like doing anything else…


AI FTP says 270.
I did a ramp test : 269 + 1 muscle injury on the left quad

I’m 41 on a high volume with a bit of experience and a good bike fit.
Did the test with a “death wish”, rarely pushed this far.
Everytime I do max-max at something I get an injury.
Maybe not enough relaxed and too much tension or other parameters.

I train-cycle to work so I guess it’s better to not break the “machine”.

Thank you for the AI, arrived just in time I guess.
I’ll keep max for the races, hoping not to break bones. :crazy_face:

1 Like

This was/ is definitely me. But I don’t seem to carry that baggage into verifying, which is why the KM Protocol/ Lamarck worked where as my head was never properly in the game for the ramp test. And I tried a lot of different approaches.

We don’t hear as much about cognitive load now Amber isn’t around, but AI FTP has definitely reduced mine. I take the same approach around the training plans/ plan builder/ AT - all reduces my cognitive load overall!


Used to like testing, but…

  • ramp test - 3 out of the last 4 ended with terrible cramps (and the one that did not I quit early)
  • 20 min - I have to do clearly negative negative splits not to start too hard and blow up
  • 60 min - sustaining the same wattage indoors is mentally too much for me (hill climb / rolling TT is fine)

AI FTP detection just confirms what I already know based on how the training has gone. There would be little point to push the body (and mind) to a breaking point for no gain.


I don’t dread the discomfort as much as I dread the recovery. My work day is garbage after I do a test in the morning. The level of suffering you need to push yourselves to wreak havoc on my brain systems and I’m not as productive. I’m in a high-performance academic job right now, and so my mental fitness for work is something I have to factor in. I can’t have off days. So yeah, testing regularly isn’t something I like to do for the sake of work and family attentiveness.


This is legit reasoning right here. Kudos to you for having your priorities in balance.

1 Like

Yes, testing is training and what not. But a significant amount of people do not want to do “Tests”.

Yes, and many people want to continue eating snacks while losing weight and getting a six pack. Managing discomfort is an important part of most sports, especially endurance sports.


I agree on the KM tests. I think lots of people don’t like the longer tests because you need to learn to pace them. Most people probably go out too hard on a 20 minute test, fade, or fail and suffer miserably at the end.

I like the KM test better because it ramps up and forces you into a negative split. You’ll have a result and even if it’s not your best, it will be close. The KM test also teaches you to feel your FTP. You do need to have an idea of where your FTP should be ahead of time.