Strava:
ICU:
Garmin shows:
Not sure how the same data file ends up w/three different average power numbers. The Garmin and Strava numbers are in the same ballpark. However, the ICU avg power values for each lap are ~50W higher.
Any ideas here?
I will have a look at this tonight. Could you please msg me a link to the ride on Intervals.icu. Tx.
The normalised power numbers between ICU and GC are very close 203w vs 201w which could be ICU fixing some drop outs and so on. Not sure why Strava has a lower number (192w). Might have something to do with how the different services handle stops and there are several in this ride. Intervals ICU skips “not moving” data points when calculating normalised watts.
The ICU average power for the whole ride is 143w vs 159w from Strava. ICU divides the total work done in Joules by the “recording time” which is the elapsed time of this ride because your Garmin was running during the stops. Strava has used moving time.
The ICU average power for each lap/interval is reported as average excluding zeros by default. The reason for this is so it can report the zone you were riding in correctly even when you have to stop for traffic lights and so on. So you were in Z2 (201w) when you were actually riding in the 1st interval and 212w in the second.
You can add a “Real average power” field to see the actual average (calculated Joules / elapsed time as described above).
Then you get 138w for 1st lap and 160w for 2nd.
These calculations are more complex than they seem for rides with stops!
Eyeballing it I figured it was due to either not including zeroes in the calculation or something w/the moving and/or elapsed time. But, I was too lazy to dig into it.
Thanks for not being lazy and having a good look 
