I am just getting into intervals.icu and like it - almost too much data. Regarding the eFTP, how accurate is the estimate compared to actually doing a 20 minute FTP test?
it really depends how often do you ride hard if you are ridding easy most of the time it will be way off.
but on a positive note, from my experience if you go on a ride and do a max 1min or 5min effort, it will be much much closer.
1 Like
Agree with Lukasz: eFTP can only be as good as the hard efforts it sees. It is fitted to your best recent efforts, so if you mostly ride easy it has little to work with and usually sits low.
Worth knowing that the 20 min test is an estimate too: FTP = 95% of your 20 min average is a rule of thumb, and the real ratio varies from rider to rider (pacing, how fresh you are, how “anaerobic” you are). So neither number is the truth, they are two estimates of the same thing.
A practical way to compare them:
- Do one properly paced all-out effort of 3 to 5 min and one of 12 to 20 min within a couple of weeks (a test, a race or a hard climb all count).
- Then look at eFTP: with those efforts in the data, it is usually close to what a 20 min test would give you.
- After that, eFTP tracks you by itself, as long as you do some hard efforts every few weeks. If you go a month with only easy rides, take it with a pinch of salt.
For setting training zones, being within a few percent is fine either way. Consistency matters more than which method you pick.
4 Likes
But if you do not do the hard efforts shouldn’t your eFTP just be lowered to what your efforts reflect over the last few weeks?
Certainly not!
There is a decay built-in, but it’s not like FTP drops a huge amount in little time. In fact, if you are having a recovery week, your FTP might go up, because your body has the time to rebuild and become stronger.
If you had a long break in training, you should first re-assess your fitness level and adjust FTP.
This is important (both up and down), because FTP is the reference point to calculate the Load.
1 Like
Good question. The catch is that an easy ride says nothing about your ceiling. If you ride at 60% for two hours, that doesn’t mean you couldn’t hold more, so the model can’t treat it as evidence that your FTP dropped. Lowering eFTP to match only the efforts it sees would make it fall every time you have an easy block, even when your fitness hasn’t changed.
And as MedTechCD says, real fitness fades slowly: a few weeks of mostly easy riding (or a recovery week) won’t move your FTP much. After a real break (illness, weeks off the bike), that’s different: do a hard effort or a test to reset it rather than wait for the estimate to drift down.
1 Like
i was surprised how accurate (you know that we don’t talk rocketsience) it is in comparison to all the tests i did, if it’s FTP or AI Dignostics or Incyd … there are just some information’s missing like VO2max, HR, Fatmax, Lamax, and calculated zones.
If it could do that we would save a lot of money 
Same for running
1 Like
Ok so if I do a max effort, or attempt to, how does the model know it is a max effort or not?
If it’s better than what you’ve done before, it’s a best effort. If not, it can show on the sub-maximal efforts. Hover over the duration you set for your eFTP detection, shows the three settings on the show options.
edit: I replaced the desktop view as the phone is too small to see the finer details.
