That rule tells you how likely you are to see one thing given N observations, but what we don't know is how many things there are.