Newsgroups: comp.lang.ada,comp.lang.apl,comp.lang.basic,comp.lang.c,comp.lang.fortran,comp.lang.perl.misc,comp.lang.pascal.misc,comp.lang.smalltalk
Path: cantaloupe.srv.cs.cmu.edu!rochester!cornellcs!newsstand.cit.cornell.edu!portc01.blue.aol.com!news-peer.gsl.net!news.gsl.net!howland.erols.net!EU.net!sun4nl!cwi.nl!dik
From: dik@cwi.nl (Dik T. Winter)
Subject: Re: Results of my test: Re: Friday 13th, try it yourself
Message-ID: <DyE0Jo.64B@cwi.nl>
Sender: news@cwi.nl (The Daily Dross)
Nntp-Posting-Host: rood.cwi.nl
Organization: CWI, Amsterdam
References: <52bmft$f3l@news1.halcyon.com> <Pine.SOL.3.91.960925162108.6568A-100000@serv1.jump.net>
Date: Fri, 27 Sep 1996 10:45:24 GMT
Lines: 26
Xref: glinda.oz.cs.cmu.edu comp.lang.ada:51514 comp.lang.apl:9577 comp.lang.c:208498 comp.lang.fortran:43509 comp.lang.perl.misc:46953 comp.lang.pascal.misc:9108 comp.lang.smalltalk:44131

In article <Pine.SOL.3.91.960925162108.6568A-100000@serv1.jump.net> Lee Crites <adonai@jump.net> writes:
...
 > Summary of Friday dates for 400 years:
...
 > If I'd made a wild guess of what the results would have been before I
 > started, I'd have come up with something similar to the above.  Uniform
 > distribution across all dates -- with diminished numbers from the 29th
 > through 31st. 

Yup, it looks uniform.  But because these are exact figures, valid for
every period of 400 years, they are *not* uniform but biased in favour
of Friday 13 (and 6, 20 and 27).

 > Realizing these results would not make the it's-not-uniform-distribution
 > group happy, I ran it again based upon 1800 instead of 1753, with exactly
 > the same results.

Which ought to have shown you that the figures are exact for every period
of 400 years.  The situation is similar to a pseudo-random number generator
that repeats a sequence of 100 digits 0 and 1 with 49 occurrences of 0
and 51 of 1.  When you look at it without background information it looks
uniform, but because you know it repeats you know also that it is biased
in favour of 1 and so not uniform.
-- 
dik t. winter, cwi, kruislaan 413, 1098 sj  amsterdam, nederland, +31205924131
home: bovenover 215, 1025 jn  amsterdam, nederland; http://www.cwi.nl/~dik/
