Attached is an image of data scattered around a power curve.
The yellow line is a trend line (power curve) representing "the best fit" for all the bubbles that you see
My intention is to test what the average deviation from the curve is for each bubble in the data set
I would like to draw two new power curves that run parallel to the trend line. The idea here is to demonstrate that X% of all the bubbles fall within this "band"
I hope this is sufficient to explain my dilemma