Did a workout today that called for 50/50’s at 120% of threshold. On my Garmin I was using the workout screen which I believe is showing avg. power for the lap and I was above target (345) for every interval. When I review the analysis, it shows a higher Power (assuming this is avg.) than my NP power for each interval. Given these were 50 second all out efforts, I’m not sure how this is possible. I would expect them to be very close to the same. One example, NP is 338 and Power is 376. Happy to ignore this if NP is not used in any calculations, but if it is, this is way below what I actually did.
NP is basically smoothed data that can be helpful to understand the physiological impact of long “non-steady” efforts, but has little relevance for short intervals.
The first step of calculating NP is taking a 30 second rolling average. NP is not a thing for 30s or shorter intervals, and not useful for maybe 30s-2m intervals. For a 50s interval, a lot of the data that’s being averaged into the 30s rolling average is from before the interval started. Once you have an interval that’s either spikey enough for the NP to overcome the rolling average issue, or the interval is long enough that the first 30s is just noise, then it becomes useful.
Thanks for the methods used to calculate NP and makes sense that it’s not a good metric for short durations. Does anyone know if it’s used by TR in determining how well you did against your workout power targets?
We use NP and AP in different places. Generally speaking, you want your efforts for structured intervals to be smooth, so AP and NP should be relatively close.
If they differ significantly, that might be a sign to try to keep your efforts more smooth and controlled.
If your AP is at or above the target, though, I’d say you’re in a good place. 
