An update:
| Edge 530 eFTP | Date | Workout |
|---|---|---|
| 267W | 16 March 2022 | 2 hour Wed ‘social’ ride |
| 265W | 18 March 2022 | 3 hour early season kitchen sink ride |
| 258W | 21 March 2022 | 2 hours with 2x15-min at 92% + 20-min at 85% |
| 258W | 2 April 2022 | 1.75 hour with 4x10-min ‘sweet spot w/1-min hard starts’ |
| 265W | 13 April 2022 | 2 hour Wed ‘social’ ride |
| 267W | 15 April 2022 | 2 hour with 3x5-min at 103-105% + 30/30s |
| 261W | 18 April 2022 | 1.5 hour with 5x6-min 95-100% |
Last night I easily finished 30-min total at 270W average (6-min at 270W, 2-min at 150-200W). Barely touched LTHR at end of last interval. It was essentially a pre-field test pacing effort. Finished with high confidence I could go out and do 45-60 minutes at 270W.
Other estimates:
- 243W Strava from last 6 weeks
- 268W Strava for last 90-days (just like WKO)
- 260W Intervals.icu
- 248W Xert (after switching from no decay to low decay a month or two ago, and now HIE is ridiculous 30kJ which is causing low FTP estimate)
- 268W WKO
My 32-min @ 275W pre-field test pacing effort from January 22 will drop out of the 90-day window in 3 days. The 90-day window is used by WKO (and Strava if you setup custom date range).
From posts earlier in this thread, Garmin is using machine learning with HR, HRV, and eVO2max (power-to-HR). I’m going to claim that TR, WKO, Strava, Intervals, and Xert all use some form of max effort power based estimates.
What I find interesting is that my Garmin estimates do not require a max effort, and even shorter threshold intervals will produce good enough estimates (always a bit below what I can do). In addition I see reasonable estimates from tempo workouts.