It helps us to turn around conditional probabilities
It is super useful to see if a variable is independent from the other. For example, if the probability of having cancer given you drink coffee is similar to the probability of having cancer probably both are independent.
It is possible to derive it if you think of a sample number of people and doing the intuitive approach of finding how many are in each population