FTP test results different from best 20 minutes


(Christian Wiedmann [X] 50) #1

When I took the FTP test, my FTP came out somewhat lower than .95 * my best 20 minute power. I’m guessing ths is because the FTP measurement is using the exact segment and there is some lag before power ramps up at the beginning of the segment.

It would be nice if the FTP test used the best 20 minute power as the basis of the calculation instead of the exact segment. Maybe the best 20 minutes out of the previous 21 minutes 10 seconds after the finish line would allow instant feedback without requiring anticipating the start of the measurement.


(Jon Mayfield) #2

Hi Christian…We use the best 20 minutes out of any time in the workout, but even 5 seconds of 0’s at the start of 20 minutes would barely affect the final number.   How far off do you estimate your number was?


(Christian Wiedmann [X] 50) #3

Hmm, weird. I got 284 from the test, but according to Strava my best 20 minutes was 302 watts, implying 286.9 watt FTP. Here is the strava link: https://www.strava.com/activities/419651118


(Christian Wiedmann [X] 50) #4

Obviously this isn’t a difference that really matters - I’m just being a little obsessive about it.


(Jon Mayfield) #5

Interesting.  If you can, zip up your Documents/Zwift/cp folder and shoot it over to support along with your .fit file from that ride, and put “For Jon” in there somewhere.  We’ll check it out and see if its something worse than just a rounding error or not. We’d appreciate it.  


(Christian Wiedmann [X] 50) #6

No problem - I’ll do that tomorrow.


(Daren Chandisingh [Vision]) #7

Hi, I also found this. Zwift put my FTP at 221W. But on the same ride uploaded to Strava it said “Accomplished a goal of 220 W for 20 minutes (237 W)”. That works out at more like 225W for the FTP. https://www.strava.com/activities/419308250

That said, looking at the power curve I’m now wondering if Strava just means “accumulated time” rather than a single time period. https://www.strava.com/activities/419308250/power-curve


(Daren Chandisingh [Vision]) #8

Having read this - https://strava.zendesk.com/entries/22881980-Using-your-Best-Efforts-Power-Curve-Premium- - I’m still not sure if Strava is using a contiguous block of time, but I don’t think it can be. Looking at my 112 mile MTS activity’s power curve - https://www.strava.com/activities/410326520/power-curve - it reckons I spent 6h 45m over 129 watts. Now that may be true, but only for the moving parts. I had several breaks for a few minutes at a time, during which power would have been zero. So I think the “best 20 minutes” on Strava could be 12 minutes here, 5 minutes there and 3 minutes somewhere else.


(Francois Coppex [X][Velosophe]) #9

Small differences here as well: https://www.strava.com/activities/421332432/overview

Strava is giving 324W*0.95 = 307.8W, and Zwift 305W. Maybe either Strava or Zwift is using normalized power while the other is using average power or a different algorithm in order to compute normalized power.


(Jude Hung) #10

Just finished my FTP test(1:13 version), and Zwift estimated my FTP to be 205W; however, as I looked into the same ride uploaded to Strava, I actually did 242W on average over the the 20 mins FTP segment: https://www.strava.com/activities/421927310/analysis/2009/2931 

The difference between 242W*0.95 and 205W is quite huge IMHO.  

 


(Jon Mayfield) #11

Jude, the other cases have all been off by 1-3 watts and we found that there were some rounding issues (and strava may round in a different way from us as well).  However in your case something else may have happened. What we’d love for you to do is what I asked of Christian above and we could track down the issue.  Thanks!


(Jude Hung) #12

Hi Jon, thanks for the quick response, and I’ll do it soon.


(Joshua Saunders) #13

Yep, same issue for me. Strava says my 20 min power for the 73 min ftp test was 292W - and it’s definitely the “right” 20 minutes, a contiguous block. Calculated FTP was 269W, but by my calculations should be ~277W. 

FWIW the “average” power displayed throughout the test was in the 266-271W range


(T hird Wheel [KRT]) #14

Mine appears to differ as well for 20 min FTP test. 314W average Strava, 317W weighted Strava and then 297W Zwift FTP, neither the 317W nor the 314W * 0.95 give me 297W. I am hoping its the 317W*0.95=301W then I dont have to bother trying to crack the 300W again.  https://www.strava.com/activities/423179063


(Johnny van Gesink) #15

Same here. 245 watts for 20min, Zwift says ftp of 226, which is 92,5% of 245. 95% would be a ftp of 233.

https://www.strava.com/activities/430298303

 

 


(Joshua Saunders) #16

So as a follow up, I did a 1 hour ride on Zwift, with one “hot lap” of the Richmond course, in normal mode, not in a workout. Averaged 295W for 25 minutes, and although I wasn’t doing an ftp test, Zwift updated my FTP, says it’s now 282, which I think is much closer than 269, and is exactly 0.95 of my best 20 minutes from that ride. Happy with that.


(T hird Wheel [KRT]) #17

Looks like zwift is working much the same as strava’s power curve and estimated FTP now. I did 324W over 20 mins which adjusted to 307W FTP which is exactly 0.95 of 324W. Think zwift FTP is working ok now and from your post also uses the best 20mins to work out FTP similar to strava’s power curve and estimated ftp


(T hird Wheel [KRT]) #18

If u have premium strava just select your power curve under the main menu and move the cursor to 20mins then read off the watts and multiply by 0.95 to get FTP. Should be the same as zwift FTP


(Simone Chiaretta) #19

I also have the same problem.

Zwift says I’ve 216, but when uploaded to strava and training peaks I get different results: in both cases I have 243W as avg over the 20 minutes, so it should be 230W as FTP (243*0.95).

Here is my training peaks: http://tpks.ws/kKbCv

and here is strava: https://www.strava.com/activities/433973461


(Lee Davies) #20

Mine done the same too! Upon completion my ‘FTP Score’ popped up as 293W and saved to my profile (where I don’t know, but it saved). When I reviewed my effort on Strava and cropped my best 20mins I averaged 324W, which is exactly what the screen fed back to me during the test. Can you investigate please Zwift, I just want to know what my correct FTP is: https://www.strava.com/activities/433310626

Thanks,

Lee