Hi there,
I’ve noticed some discrepancies in power curve calculations across intervals, wahoo, and strava.
I originally synced the strava data and noticed a difference of ~100W, so downloaded my strava extract and uploaded via bulk but it’s still mis-reporting the power curve.
For 5 second power, Wahoo = 742, Strava = 749, intervals = 648.
See for comparison screenshot of power curve from wahoo and intervals. What could be the cause?
Its probably Intervals.icu “fixing” power spikes. You can use the activity list view to find these rides. Add the “Power Spike%” column and click it twice to sort:
Then go to the activity (click the date) and decide if the power is good. You can add the “Raw power” trace to see the unfixed power. The “fixed” bit gets a orange marker. If your PM was ok and you did the power then click the message at the top to raise the spike threshold:
You can change the default spike detection threshold in /settings.
Spike detection is relative to your FTP so it could be that was set a bit low.
Cheers David!
How would you go through the removing the “fix” stage. I’ve clicked on some of the markers at the top but doesn’t look obvious where to accept/reject the “fixing”
Also, do you have an FAQ or something where you would/wouldn’t want to overrule? I’m relatively new to working with power so not sure what to trust as beast legs and what to disregard as overcalculations.
It depends on how reliable your power meter is. You can usually tell just by looking at the spike and you know if you were blasting it there or not.
You need to click the orange “Power spikes” message and bump up the threshold. Then it will recompute and you should see the little orange blobs disappear from the graph.
