central-limit-theorem

The sample mean of many independent finite-variance draws becomes Gaussian, and its spread shrinks like σ/√n — a rate that follows from how variance adds, and that breaks precisely when its hypotheses do.