If Strava was going to remove zeros it would do it for the entire ride, not jut a portion of it. There doesn’t appear to be any part of the actual work interval that you weren’t pedalling so there’s no zeros to remove! I’m not sure I follow your point anymore!!! Sorry
I had a similar (but opposite) experience when I imported historic rides from my Wahoo to TR via Strava.
I dug into it for a while, but gave up.
The average lap power that was recorded on the wahoo was higher than what was imported to TR.
Once I had my FTP the same across all apps/devices it sorted itself out.
Happy to leave it, essentially the question could be re-framed as:
Why is the lap data on strava higher than the lap data showing on my Wahoo?
Ok, makes more sense. FWIW, I wouldn’t worry about it. Those differences aren’t really significant. Only about 1-2% difference.
1 Like
Regarding that specific ride, it would be interesting to pull the actual raw data file and look at the power figures recorded. You could then run a few different averages to try to get a sense of what the different softwares are seeing/ignoring.
What @AndyGajda said. Those kinds of differences could just be due to the ambiguity of whether you include the first and/or last data samples in the lap averages. Remember data is usually sampled once per second, so this means up to 2 seconds-worth of data could be affected by this ambiguity alone AND for intervals this is exactly where power is changing the most.
1 Like
What power meter are you using and is the firmware up to date on your wahoo?
Google “wahoo average power too high” - there were definitely issues with how wahoo dealt with average power in the past.
My Wahoo is actually too low and it updates automatically and quite often, same with my Stages PM so I think I am upto date. I’ll certainly double check.
I had a look at these and I think its a framing issue with the lap data not matching the power output for each interval. The intervals are started a little late and don’t “fit into” the laps. So the Wahoo interval data is incorrect. There aren’t any zero power points during the intervals so its not “Strava removing zeros”.
If you take the power points for the first interval (ignoring the laps, just the 60s of hard work) the average is 617w which is much closer to Strava 611w (and Intervals.icu 617w) than Wahoo 563w.
1 Like
Thank you for this, I did notice when I briefly look down at the Wahoo at the start of the lap that things can take a bit to respond so this backs up your findings.
I’m actually getting a new Bolt next week due to an issue with the twist lock so it’ll be interesting to see if a new device works differently.
Ok blast from the past here and maybe related…
Today i did a workout on Zwift using my PM as the power source, after uploading the data to strava the interval power is much lower compared to intervals.icu @davidtinker and also to training peaks, both of which match with each other. Golden Cheetah which I also use (linked to strava) has the same data as shown in strava. These were 30s on 30s off intervals, I hit the pedals hard at 1-2s to go and during the interval I’m above target power (sometimes quite a bit above) so I know that the higher power output is correct. Intervals.icu and Training Peaks correlate with my expectation, Strava and GC do not.
In the original post I think it was agreed to be a framing issue in that the Wahoo was out of sync a bit with the interval start and end, this time I am just using one system, in this case Zwift.
I’d welcome thoughts again on this.
Agreed but in this instance it’s one lot of data. 
Intervals.icu doesn’t use the lap data by default and detects the intervals from the power trace. So thats why it is getting them right. I would also be a bit peeved to do a set like that and have the power under-reported.
1 Like