where is good stretch of road for 20min FTP test ?

[lechatnoir]

Speaking of how to estimate FTP, does anybody have some insight into how Strava estimates FTP on your power curve?

I did 'n bleeding eardrums 10min effort on a climb on Saturday and my FTP estimate jumped up by 30W. New number seems unreasonably high. I'm guessing if I had to implement the estimating algorithm I'd do some kind of least squares matching to smooth functions that represent realistic power curves, but maybe they are doing something simpler?

 

i suspect that if Strava gets it right, it's more by accident than design. I have no proof for this, and I don't know what algorithm they use, but there are so many variables that I have a very jaundiced view of those numbers. Perhaps because any estimate they make for me I feel are too low!!!!

[Jbr]

Strava and intervals.icu give a close estimate of my FTP. Intervals.icu will give you FTP bumps everytime you do a good effort, sometimes on a 5 mins, sometimes on a 10mins or more, not only on 20mins efforts. It will sometime give me a value that seems a bit high, but after a couple weeks it comes down to more reasonnable values.

 

You don't really need to know your FTP very accurately as you won't have the same result any day anyways, it is a good indicator, and most of the time different estimates will be within a 10w bracket (except for Garmin that will update your FTP everytime you ride at tempo for some duration, even though you weren't at all doing an all out effort)

 

At the moment my intervals.icu eFTP is 6w higher than my strava estimated FTP, and my training peaks is based on an FTP that is 1w lower than strava's estimate.

 

Haven't updated this value in months as I don't feel I have done a proper 20+mins all out effort that justifies using a higher value for now. I think my eFTP is a bit higher because I had a 10mins all out workout recently after a few weeks of 5x10mins intervals so logically I improved on that specific duration.

Edited April 19, 2021 by Jbr

[Chris_]

Where ever gives you the higher FTP - thats the one you use for bragging rights. 

 

The lower one you can use in your training.

[Jehosefat]

Speaking of how to estimate FTP, does anybody have some insight into how Strava estimates FTP on your power curve?

I did 'n bleeding eardrums 10min effort on a climb on Saturday and my FTP estimate jumped up by 30W. New number seems unreasonably high. I'm guessing if I had to implement the estimating algorithm I'd do some kind of least squares matching to smooth functions that represent realistic power curves, but maybe they are doing something simpler?

 

I'm pretty sure it's based on some modelled power curve (e.g. Morton's critical power model). And just like the 95% of 20 min max power, it's only going to be accurate for certain people. I've seen some papers that show, in testing, that while the 95% of 20min power is pretty close to the average, for individuals their true FTP lies anywhere between around 90% and 97% of their 20min max.

[Jbr]

Where ever gives you the higher FTP - thats the one you use for bragging rights. 

 

The lower one you can use in your training.

100%   

[DieselnDust]

Strava and intervals.icu give a close estimate of my FTP. Intervals.icu will give you FTP bumps everytime you do a good effort, sometimes on a 5 mins, sometimes on a 10mins or more, not only on 20mins efforts. It will sometime give me a value that seems a bit high, but after a couple weeks it comes down to more reasonnable values.

 

You don't really need to know your FTP very accurately as you won't have the same result any day anyways, it is a good indicator, and most of the time different estimates will be within a 10w bracket (except for Garmin that will update your FTP everytime you ride at tempo for some duration, even though you weren't at all doing an all out effort)

 

At the moment my intervals.icu eFTP is 6w higher than my strava estimated FTP, and my training peaks is based on an FTP that is 1w lower than strava's estimate.

 

Haven't updated this value in months as I don't feel I have done a proper 20+mins all out effort that justifies using a higher value for now. I think my eFTP is a bit higher because I had a 10mins all out workout recently after a few weeks of 5x10mins intervals so logically I improved on that specific duration.

 

 

 

Does Strava use the same FTP calculation and input data as Intervals.icu? I thought Strava was using average speed through segments vs Intervals.ICU using calculated average power at 10s, 30s 1min, 2min, 5min, 8min etc.

the difference can be accounted for if this is the case as I.ICU will account more for the variance in terrain than strava.

Strava has evolved their power calcs over the last 2 years so I'm not sure where they are in terms of accuracy anymore. 

[Jbr]

Does Strava use the same FTP calculation and input data as Intervals.icu? I thought Strava was using average speed through segments vs Intervals.ICU using calculated average power at 10s, 30s 1min, 2min, 5min, 8min etc.

the difference can be accounted for if this is the case as I.ICU will account more for the variance in terrain than strava.

Strava has evolved their power calcs over the last 2 years so I'm not sure where they are in terms of accuracy anymore. 

no idea what they use, but since both my bikes have power meters the estimate seems accurate. I train on a 320 base, Strava estimate is 321 and Intervals.icu 327.

Back when I used Xert it was also a bit high like Intervals.icu.

All very consistent data with my last few months/year results... A little too consistent if you ask me, I'd like to see those numbers go up for a change...

Edited April 19, 2021 by Jbr

[DieselnDust]

no idea what they use, but since both my bikes have power meters the estimate seems accurate. I train on a 320 base, Strava estimate is 321 and Intervals.icu 327.

Back when I used Xert it was also a bit high like Intervals.icu.

All very consistent data with my last few months/year results... A little too consistent if you ask me, I'd like to see those numbers go up for a change...

 

 

ah if you're using a power meter then your data spread be pretty tight. Its similar for me. FTP on the MTB is higher than on the road bike by a few Watts.

I've never tried using either. I checked some of my past data for when my PM was out of commission due to a battery cover that needed to be replaced and I.ICU doesn't calculate your FTP independently. It pulls whatever Strava calculates for you as the baseline. So when I didn't have the PM on the MTb it used the Strava Power (when using a premium account). When using a free account it didn't have any power data for my workouts.

So without a PM I reckon the Strava Premium power number is waaaay off the mark

[bleedToWin]

ah if you're using a power meter then your data spread be pretty tight. Its similar for me. FTP on the MTB is higher than on the road bike by a few Watts.

I've never tried using either. I checked some of my past data for when my PM was out of commission due to a battery cover that needed to be replaced and I.ICU doesn't calculate your FTP independently. It pulls whatever Strava calculates for you as the baseline. So when I didn't have the PM on the MTb it used the Strava Power (when using a premium account). When using a free account it didn't have any power data for my workouts.

So without a PM I reckon the Strava Premium power number is waaaay off the mark

 

 

Yeah, we're wondering about Strava estimates for FTP based off of PM data on the power curve, not Strava segment and ride power estimates without PM present.

[Jbr]

Slightly off topic, but Intervals.icu must be connected to your Garminconnect account, not strava, then it doesn't mess up your "laps" and you can use them to analyse your intervals

[DieselnDust]

Slightly off topic, but Intervals.icu must be connected to your Garminconnect account, not strava, then it doesn't mess up your "laps" and you can use them to analyse your intervals

 

 

didn't know that. Thanks for for tip <off to make the right connections>

[APJ88]

inside the corridors of hell....