home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.skeptic
- Path: sparky!uunet!haven.umd.edu!darwin.sura.net!zaphod.mps.ohio-state.edu!sdd.hp.com!think.com!rpi!newsserver.pixel.kodak.com!psinntp!psinntp!wrldlnk!usenet
- From: "James F. Tims" <p00168@psilink.com>
- Subject: Re: Innumeracy, humorous ... maybe.
- In-Reply-To: <1992Nov17.025722.6032@pages.com>
- Message-ID: <2931105518.0.p00168@psilink.com>
- Sender: usenet@worldlink.com
- Nntp-Posting-Host: 127.0.0.1
- Organization: Semper Excelsior
- Date: Tue, 17 Nov 1992 16:10:37 GMT
- X-Mailer: PSILink (3.2)
- Lines: 64
-
- >DATE: 17 Nov 92 02:57:22 GMT
- >FROM: Bruce F. Webster <bwebster@pages.com>
- >
- >> In article <BxoJK2.KzI@ccu.umanitoba.ca>
- >> vnelson@ccu.umanitoba.ca (Gerald Vernon Nelson) writes:
- >> >
- >> >Speaking of inummeracy...
- >> >just didn't understand that the odds of winning that lottery were
- >> >about 10 billion to 1 (49 * 48 * 47 *46 * 45 * 44).
- Anybody want to figure out how many winners per century?
-
- >> >
- >> >He was of course, inumerate himself, as the odds of winning are
- >> >actually about 14 million to 1.
- >> >
- >
- >Well, in a society where high school graduates can't make correct change, I'd
- >be a bit hesitant to call someone who doesn't correctly apply statistics
- >"innumerate"--but given the tenor of his article, it might be just. :-)
- >
- >In article <1992Nov16.045407.29782@udel.edu> mccoy@pecan.cns.udel.edu (Don
- >McCoy) writes:
- >>
- >> Uh, I could be wrong here, but I believe 10 billion to 1 is right.
- >> How did you arrive at the 14 million to 1 figure???
- >>
- >
- >The fallacy of the author's calculations is an easy one to fall into, but
- >here's a quick thought experiment which will illustrate it: if the lottery
- >required that you correctly picked 48 out of 49 numbers, would your odds of
- >winning be (49*48*47...*4*3*2) to 1? Obviously not; they would be merely 49 to
- >1, the same as correctly picking 1 out of 49 numbers. So the calculations have
- >to be symmetric and work appropriately for any selection of numbers.
- >
- >The correct caculation for the number of possible ways of taking n objects r at
- >a time is n!/[r! * (n-r)!]. For picking 6 out of 49 numbers, that leaves you
- >(after factoring out 43! from 49!):
- > (49*48*47*46*45*44)/(6*5*4*3*2*1) = 13,983,816 possible combinations.
- >
- >Thus, for any random combination, the odds of having chosen that one is 14
- >million to one.
- >
- >If you're still stuck back on "but the odds should be based on how many numbers
- >are left", here's the fallacy the author of the news article made: the odds of
- >the first lottery ball chosen being correct are not 1/49, they are 6/49. You've
- >picked six numbers ahead of time, remember--any one of those six can match the
- >first lottery ball. If one does, then the odds of any of the second ball
- >matching any of the remaining five numbers is 5/48; for the third ball, 4/47;
- >and so on. Multiplying these odds together, you get a familiar value:
- > (6*5*4*3*2*1)/(49*48*47*46*45*44) or 1/13,983,816.
- >
- >Q.E.D. ..bruce..
- >
-
- Don't the numbers pop out of six separate containers?
- Assuming that every jar has all 49 numbers, the odds become 6!/49**6,
- or 1/19,223,881. Or am I missing something?
-
- --
- jim tims
- "And if you're a miner, when you're too tired and old and sick and stupid
- to do your job properly, you have to go, whereas just the opposite applies
- with the judges." Beyond the Fringe
-
-