home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: alt.folklore.science
- Path: sparky!uunet!elroy.jpl.nasa.gov!sideshow!urban
- From: urban@sideshow.jpl.nasa.gov (Michael P Urban)
- Subject: Re: Monty Hall/Shell Game..
- Message-ID: <1992Dec22.155234.28333@elroy.jpl.nasa.gov>
- Sender: news@elroy.jpl.nasa.gov (Usenet)
- Nntp-Posting-Host: sideshow.jpl.nasa.gov
- Reply-To: urban@cobra.jpl.nasa.gov (Michael P Urban)
- Organization: Jet Propulsion Laboratory, Pasadena, CA
- References: <aPJ6VB1w164w@perseus.spk.wa.us>
- Date: Tue, 22 Dec 1992 15:52:34 GMT
- Lines: 29
-
- The remarkable thing about the Monty Hall Problem or shell game or
- whatever is, perhaps, not its mathematics but its psychology. The
- correct answer (the odds are 2:1 that you should switch) is
- _most_ counter-intuitive for people with some mathematical
- training--the sort of people who might apply a heuristic like `now you
- have reduced the problem to two doors'. And because the mathematical
- savants know what they are doing, once they have reached the wrong
- conclusion it is astonishingly hard to persuade them to the other
- side. This is why the problem surfaces again and again in places like
- alt.folklore.science or the syndicated column `The Straight Dope'
- (where it has appeared at least three times) and similar forums.
-
- In `The Straight Dope', the columnist (Adams?) once came up with an
- explanation that might be helpful: extend the problem to one million
- doors. You pick one. Now Monty goes and opens 999,998 of the other
- doors, showing you that none of those doors has the valuable prize.
- At the rate of a door per second, he finishes this task in about 11
- days; this is why we are not using goats :-) You may now switch with
- that one other door that remains. The point here is that 999,999
- times out of a million (i.e., if you did not pick the right door to
- begin with), Monty ONLY HAS ONE CHOICE as to which doors he opens, and
- as a result, HE IS SHOWING YOU WHERE THE PRIZE IS.
-
- Does this help anyone? No? Oh well...
-
- Mike Urban
-
- urban@cobra.jpl.nasa.gov
-
-