There's uncertainty introduced by the idea that you don't know how big that range is supposed to be. Technically the M/n figure should be (M-L)/n, where L is th

There's uncertainty introduced by the idea that you don't know how big that range is supposed to be. Technically the M/n figure should be (M-L)/n, where L is the sample minimum observed, and the estimated population minimum would be represented by L + (M-L)/n - 2 [1]. [1] two degrees of freedom should be lost due to estimating both M and L, thus subtracting 2.