Is there anything about Strava’s Estimated FTP that would suggest that it’s a bad measure? Let’s assume all data used for the calculation came from a real power meter.
I tend to be more motivated on outdoor rides, so I think it’s reasonable that the estimation is higher than, say, from a ramp test.
Using this FTP for training is a different question…
@freerider while there is an algorithm that calculates this, it is essentially a guesstimate at best. This is what is bad about it, that it has little to no specificity and all power is calculated with factors including GPS, your weight, age and terrain that you ride on via an algorithm.
But if you really wanted to use it as a metric, compare it to your ftp numbers inside. Ie 20 mins test is 300w and on a 20min-ish climb outside it is 320w. There is a 20 watt difference in power…you can ROUGHLY use this to compare indoor and outdoor rides.
Though obviously it will be FAR less accurate and therefore rubbish for anything else than obtaining some kind of ballpark idea of training outside.
However, if you are interested there are ways you can train outdoors also. Just ask if this is what you are after 
It’s probably ball park but unless you are racing (and I mean contesting the race…not noodling around sipping mai tais at the back of the field like @NorCal_Cycling is a fan of) then the data probably doesn’t represent a true maximum/FTP test effort.
My current difference is 222 reported on Strava and 229 on last ramp test. My Garmin has been telling me 225 or so for the last few rides too.
A ramp test is also an estimate at best. I find Strava estimates are pretty dam close. Its got all my TR data, races etc.
In saying that, i do plenty of road races, and my group rides and outdoor rides in general have plenty of long efforts.
Given that FTP is a dynamic figure and fluctuates day to day, week to week, I’d say 222-229 is pretty damn good for variance.
Agreed. It’s not bad. Strava appears to be doing a floating 6 week average rather than a current point in time thing so that’s worth keeping in mind when using it to set zones. Ild expect the number to be very laggy if it’s doing what I think it is.
These are quite a bit closer than mine, actually, and my variance is the other way.
Using a power meter on my bike for both indoor and outdoor rides (i.e. consistent measuring device).
My rides (both races and recreational / training) definitely consist of all-out uninterrupted segments.
Ramp test gets me to ~255, Strava is saying ~275.
Conceptually, this makes sense. The level of motivation I have in my pain cave doing a ramp test is quite different than that racing my buddies up a climb.
Thats interesting. Which FTP number do you use for your indoor zones? Do you find your indoor training is effective and pushes you properly if you use the lower number? What’s your cooling setup like?
I use the ramp test data and feel sufficiently challenged.
I have 1 big fan in front of me. I definitely sweat buckets and often think about getting more fans.
Are you thinking that maybe overheating is killing indoor performance?
The FTP estimate is only going to be as good as the data you give it to sample. I believe Strava bases its FTP estimate on your ~20min max effort, so if you don’t go hard for at least 20min very often, it may not be that accurate. For example, my FTP based on the ramp test a month ago is 385w, but Strava shows me at 343w as a 6wk FTP estimate since I haven’t done a 20min effort in a while. 385 is much closer to reality, as my 20min max is close to 400w.
That was my thinking. Though maybe I’m the weirdo for having the two match closely
How big is big? I’m using a 20" offset to the side.
Ok now I’m seeing a huge swing the other way that makes no sense. I did a 20min effort yesterday @ 420w avg, and Strava is saying my estimated FTP is 416w… how does that make any sense. I’d love to get some insight into the algorithm they’re using.
I think that makes sense: you just did 20 minutes at 420. Strava says:
Thus, the easiest way to calculate your FTP is to test your best average power for 20 minutes. We believe 20 minutes is enough time to stress the same physiological systems as a 60-minute effort would and it is easier to consistently do within your season.
Perhaps they round down 
Ah, I didn’t realize they took the whole 20min avg, and not 95%.
I don’t believe it’s accurate for me, I’d say it’s over estimating mine by about 25W. Also, the new TR ramp test also over estimates for me*
*on the basis of my FTP being what I could sustain or 1 hour which I know isn’t the most recent accurate description of what FTP “is” or more to the point isn’t.
It’s very close to Monod.
You need recent, full-gas efforts from say 4min to 8min to 20min to 30min to get an accurate FTP. If you don’t have good critical power data for a range of durations, you won’t get a good curve.
When I was doing most of my riding outside, Strava’s estimated FTP seemed to be pretty accurate but lately, I’ve focused on TR workouts (currently in Build Phase) with fewer outdoor rides and it’s ~15% below my actual.
Also, since Ramp Test comparisons are being made, I’m doing my TR workouts around 5% higher than my Ramp Test results.
Strava, Training Peaks, and Zwift Power all estimate me around 255. I train at 272. Given how I feel at 270+ vs 255, I think the 272 is a better representation of my physiological FTP. As others have said, the Strava (et. al.) estimates are only as good as the data they get, so unless you’ve really gutted yourself for 20 or 30 min, or conducted a 20 min FTP test, odds are they’re going to come out low. If you don’t really gut yourself on a ramp test, THAT will come out low as well.
Don’t have that data ATM. I’ve been away from home for almost four months, so I’ll dropbox a bunch of data and see when I get back. I haven’t used a straight ramp test FTP estimate for several months, just making manual adjustments based on testing and training/racing performance. Training at 272, I’ve had one workout that went poorly over the course of a month, and that was coming off of back-to-back nights of poor sleep. But I’ve done 4x20 at 95%, 3x30 at 90%, and yesterday did almost 2 hours straight at 85%, without issue at 272. More in line with the above, my program estimated FTPs are all based on power files for events where I’m conserving energy (long road races, even crits), etc., or workouts which aren’t good representations of maximal efforts.
My 5-min MAP data is 327W, which puts me a shade over 120% of FTP at 272W.
I like to go as much by feel as the data anyway, and I know that conducting Over/Unders at 255 would be far too easy. I can recover at 255, and feel good burn after a couple of minutes at 270+. When pulling on the front of a long road race or century, I’ll target threshold watts and run at 270 or so outdoors without much issue.