While we are at it, we can again "go backwards" and figure out the mean and variance if given some probabilities.