How are you estimating the serve velocity accuracy? How accurate is the serve velocity?
My error analysis is a little rusty -
Usually a measuring instrument has a reproducibility error and a systematic error. I'm not sure how to add the conditions under which the measurement is made, camera to ball distance, etc.......?
For the reproducibility error -
How reproducible is the measurement when the instrument measures identical serve velocities? +/-1% or +/-10% etc.
For the systematic error -
Does the instrument have a systematic error whereby it measures all serve velocities with a bias, for example, all are 10% high. This might vary depending on the conditions, distance from the camera to the serve, etc. ?
As an example, suppose that you have a ruler that is supposed to be 12" long but it is really only 11" long. You might reproducibly record lengths to +/- <0.3 % or +/- <0.036". But because your ruler is too short a correction has to be applied.
Last edited by Chas Tennis; 12-11-2012 at 08:06 AM.