The interval power average of this activity is incorrect : Intervals.icu
306W instead of 335 and I can’t figure out why.
On the power data, I have 337W for 5min.
I made comparison of all curves in several tools, I have also 335-337W for 5min, only the interval power average is showing 306W
I made a dual recording (power is really close between 2 records) and the power average in the same interval is correct. See second interval of activity Intervals.icu
Is it because the 30s before the interval was ridden at very low watts and possibly that 30s was included in a rolling average? Just an idea that might be worth checking
Avg Power is just average for the entire Interval and does not use any rolling avg. Only NP uses rolling avg.
It might be worth checking the power data stream for the interval. If there are dropouts recorded as 0, this will influence the standard avg in Intervals because average power for intervals includes zeroes.
Zoom in to the interval and click on Actions - Edit data to check the data-points.
I’ve tried to edit data, to replace “0” cadence where there is power>0 by another cadence to check, etc.
But I see no difference between both activities
I’m wondering something similar from my July 23rd activity. I did a 1 hour test indoors, and while intervals says my 1hour average is 391W vs. the 393W reported by the recording app and strava (could be a kJ average vs. linear average difference), my calculated FTP is showing 383W from “41min at 388W”, which doesn’t make sense to me. Any thoughts?
Intervals.icu does interpolate to close small gaps but 17s in a 1s/tick activity is too much.
You can see the moving time is only 48m because of those gaps. This is also causing the total work for the activity to be less and lowering the power curve (work / elapsed time). The interval display is more lenient with respect to gaps.
I have this activity - an easy run with 10x 30 seconds hill sprints - where intervals data are very different than what displayed on Suunto App. I extracted the fit file and analysed “laps” in Excel, they correspond to Suunto App. What could be the issue?
Intervals.icu uses the start and end of each lap but computes all the stats from the power, heart rate and other traces. This is to support editing the power etc. and editing the intervals.
Unfortunately often this does not match the data computed by the watch and stored in the laps. That is “higher resolution” than what is available via the activity data traces.
Thanks for the quick reply @david
So is there a way for me to edit those intervals to match the actual ones? I tried by editing start/end time of an interval, but it automatically changes the lap distance, too, and the result is not what I’m looking for.
Intervals.icu should already have got them as close as possible to the laps. If you download the streams CSV from the activity data page and open it in Excel you will see what I mean. Intervals.icu uses the start time of the interval to find the start index. Sometimes this is between time ticks. It uses the larger of total elapsed or timer time to compute the end time and hence the end index (exclusive). Again this might fall between ticks. Then these indexes are used to calculate average power, duration etc. from the streams. So you can see how it might be different to the laps.
