# Talk:Extreme Spread

I know I'm being repetitive, but *"Assuming that the shots are Rayleigh distributed"* is a bit misleading: The shots (in the sense of (h,v) coordinates) are not assumed to follow a (univariate) Rayleigh distribution but a bivariate normal distribution. The distances of the shots the true COI follows a Rayleigh distribution if the bivariate normal distribution of the coordinates is restricted to have equal variances and 0 correlation.