woods@hao.UUCP (Greg Woods) (02/13/84)
This is a common fallacy. Just because there are 365 "chances" does *not* mean the probability of winning is 1. When your "chance" total hit 365 exactly, that means you have a 50% chance, not a 100% chance, of winning. If you are going to play that game for money, you better have enough people so that your "chances" are a lot higher than 365! This is one of those things where it is easier to calculate the probability of *losing* and subtract it from one. As you can see, the chance of winning does not become 100% until you have more than 365 people, which is what your intuition should tell you. The chance of losing is number of outcomes leading to a loss divided by total possible outcomes. The total possible outcomes is easy: if you have N people, it is 365^N (ignoring leap-years for simplicity's sake. It doesn't change the result significantly). The number of possible ways to lose is a little more tricky, but not too bad. The first person can have any of 365 birthdays. The second, if you are to lose, can have any of the remaining 364. The third, 363, and so on. Therefore the number of ways to lose is 365*364*....*(365-N+1). This can be expressed mathematically as 365!/(365-N)! where "!" means factorial, fact(n) { n=0 ? return(1) : return(n*fact(n-1)); } Therefore, your chance of losing is (365!/(365-N)!)/(365^N). This correctly predicts your chance of losing to be 1 if N=1, and nearly 0 if N=365. The formula requires defining -1! if N>365, but surely you can see that it really is impossible to lose if you have 366 or more people! :-) GREG -- {ucbvax!hplabs | allegra!nbires | decvax!stcvax | harpo!seismo | ihnp4!stcvax} !hao!woods
darrelj@sdcrdcf.UUCP (Darrel VanBuer) (02/13/84)
The probability of adding the m-th person to a set of people with different birthday is (roughly) (366-m)/365 [ignores leap year, assumes all dates equally likely]. So the probabily of a set of m all different is (364 * 363 * 362 * ... 366-m)/365**(m-1). This works out to about 50% at 25 people and not quite certain at 50 people. -- Darrel J. Van Buer, PhD System Development Corp. 2500 Colorado Ave Santa Monica, CA 90406 (213)820-4111 x5449 ...{allegra,burdvax,cbosgd,hplabs,ihnp4,sdccsu3,trw-unix}!sdcrdcf!darrelj VANBUER@USC-ECL.ARPA
srini@ut-sally.UUCP (Srinivasan Sundararajan) (02/21/84)
The Pigeon-Hole Principle ========================= Why bother with probabilities at all if all you need to be sure is a 100% chance of winning. For those have not heard of pigeon-holing : 1) Assume 365 pigeon-holes - one for each day of the year. 2) Assume P persons in the room. 2) Each person claims a pigeon-hole for himself. ( 2 persons might have the same birthday and thus claim the same hole ). 3) If P = 365, it is possible ( rather slim ) , that they all choose different holes, BUT, if P >= 366, then we are guaranteed that at least 2 persons pick the same box. From: srini@ut-sally