1. The problem statement, all variables and given/known data Show that the FD distribution can be viewed as giving the probability that a given state ( of the prescribed energy) is occupied. 2. Relevant equations 3. The attempt at a solution Solution to this problem: I understand the solution,but I took a different approach which gave a different answer. For a quantum state i,denote Z_i as its partition function.Then for a single state distribution(2nd red box) : For a fermion,n_i = 0 or 1. I want to find P(1) for the state i,so by the 2nd box it is P(1) = e^β(μ-ε_i) / Z_i , which is not the same as the FD distribution?