Let's say you have 50 phone books, one for each state in the US. For each last name in each phone book, you compute the frequency with which that name appears. So "Smith" might comprise 1% of the phone book, while "Spears" might be 0.01%.
Now let's say I gave you a group of 10 last names, and asked you to determine which state these people were most likely from, based on the phone book counts.
How do we go about solving this problem?
Even if you can point me to the area of probability where this problem lies, I'd be happy to go look it up and teach myself, but I do not know where to begin.