If I have a distribution of points, how may samples do I need in order to draw an area that encompasses (say) 90% of the points, to (say) 90% confidence?
I am running a monte carlo simulation to predict the accuracy of a footballer's penalty kick. I am specifically after the area in which the ball will end up, in the plane of the goal, 90% of the time. I want to construct a confidence interval, and I am after the sample size required. The problem is I do not know the variance or any other statistical data as yet