I thought the Kickr uses slope mode? Which should help account for what you had happen
it automatically changes to slope mode during the testing portions, otherwise it'll be using erg which mechanically smooths.
This is normal (not mine, for what it's worth)... what you have is pretty jaggy.
What's your power smoothing on? I run mine at 3 seconds for pretty much everything other than crazy hard sprint intervals, when I run it at 1 or 2.
i was under the impression that power smoothing only effects the onscreen data, not the raw data? at least that's how it works on my garmin head unit. my power smoothing is set to zero on trainerroad.
and aren't you using virtual power? the farther away the power meter is from the crank arms, the smoother the results will look. virtual power is extrapolating power based on the flywheel of the trainer, so its results will look smoother regardless of any setting. and since it's not directly reading power, it simply can't be as sensitive as a dedicated power meter so it doesn't take into account the wild fluctuations you see in my graph. not a single person can produce power as smooth as you do (or the person in the chart whom i assume is using virtual power)
for refrence, skim down to the last portion of this
article. it explains why despite the two graphs appearing different, the results are the same. and also this
graph is useful in illustrating how virtual power smooths data in comparison to raw power data