Hey David,
I noticed to day that the eFTP is being displayed differently on the ‘Power’ screen than on the ‘Fitness’ screens. My real FTP is ~330 and my eFTP is ~325 or so. It is displayed correctly here:
But not here on the ‘Power’ page:
Or here:
(However, that percentile value is correct based on the chart to the left)
Hopefully those pictures aren’t too close to see where they are located. But I can give some bigger screen shots if needed.
Thanks!
Yes this is a little unfortunate. The eFTP displayed on the power page is different to the “continuously updated” eFTP displayed elsewhere in the app. On the power page you get an eFTP value for each selected power curve that is computed using only data for that curve (last 42d, 84d, last season etc.). The “continuously updated” eFTP value gradually decays if you reduce training load a bit and is updated every time you complete a ride.
You can also change the algorithm (default is “Mortons 3P with FFT curves”) on the power page. “Mortons 3P” is good if you have good inputs and will give you a good W’ number as well. However if your inputs aren’t good then it is all over the place which is why “Mortons 3P with FFT curves” is the default. It is much more stable and only needs a single max effort, however it assumes everyone on the same FTP has the same W’.
The eFTP value shown on the little table with rider types is incorrect. That should be the eFTP shown elsewhere in the app because that is what is used for the percentile rating. I will fix that!
I was also confused by this. It is still to be fixed, is it not?
The eFTP value shown on the little table with rider types is taken from the power curves selected. I decided not to use the rolling eFTP value because the rest of the data comes from the selected power curves.
Sorry to resurrect such an old thread, but this has been really frustrating me recently as well. I really dislike interrupting my typical ride schedules for taking FTP Tests, and was hoping that I could just trend off of one of these values. When I joined here, they were right on top of each other and represented what I was seeing in Zwift as well; however, they are now very far apart. I blame inconsistency of outdoor group events for this partially. I have two questions about this… Which should I trust (for workout inputs)? and Why did they both just go down after completing the hardest outdoor race I’ve ever done yesterday?
PS - Activity matching isn’t working great with my 3 connected data sources [Dropbox (from my Wahoo), Garmin (from my watch), Strava (potentially from both)], so I delete one recording from Strava, and ignore everything from one here - but is that potentially not be accounted for in the eFTP calculations being performed?
Races often don’t give good numbers because the efforts are on/off which results in lower average power that what you can achieve with a consistent effort. You can look at the “Ride eFTP” number which is eFTP calculated using that ride alone. Your race power is very spikey so hence the lower than expected eFTP.
The continuously updated eFTP value decays slowly if you reduce training load. Of course Intervals.icu doesn’t know if it has really gone down. You have to do a long enough max effort to get an update measure. At least you don’t have to plan for that, if you are feeling good on a hill just go for it.
The one on the /power page uses data from the selected curve and displays highest number achieved in the last 42 days or whatever.
The dup detection only works vs Strava activities. It doesn’t try to pick between Garmin Connect and Dropbox. But the ones from Strava should be auto deleted.
Bottom to top: Strava shows up matched to both Dropbox and Garmin (which is fine). I think I read that if I’m not regularly testing, I should trust the eFTP on the Power Page (perhaps you should consider giving these different names to avoid confusion) with more confidence. FYI - Garmin said that after my race my FTP went up to 277W (you say it went down from 271 to 265W). Thanks for the feedback.