Computer Chess Club Archives


Search

Terms

Messages

Subject: correction

Author: Martin Giepmans

Date: 17:26:11 12/19/01

Go up one level in this thread


On December 19, 2001 at 19:05:41, Martin Giepmans wrote:

>On December 19, 2001 at 16:57:46, Michel Langeveld wrote:
>
>>Problem = 1 - ( 16777216! / (16746008! * 16777216 ^ 31208) )
>>
>>16777216 = Number of records in my hashtable
>>31208    = Number of nodes searched
>>
>>Windows calculator says BOOM.
>>
>>If Problem > 0.5 then I've spend 15 hours debugging for nothing :-)
>
>
>problem = 1 - f(b)
>
>where f(b)=a!/((a-b)!*a^b)
>
>"15 hours for nothing" if problem>0.5
>this is true if f(b)<=0.5
>
>well, f(b+1)=a!/((a-b-1)!*a^(b+1)) = a!/((a-b)!/((a-b)*a^b*a))
>=((a-b)/a)*f(b)  ok?
>
>in your case: a-b/a <0.999
>
>f(1)=1
>f(2)=((a-b)/a)*f(1)<0.999*f(1)=0.999
>f(3)=((a-b)/a)*f(2)<0.999^f(2)<0.999^2
>....
>f(b)<0.999^(b-1)
>
>
>f(1000)<0.999^(1000-1)
>this is allready <=0.5 and your f(b) is even smaller, so ...
>
>Martin

"in your case: a-b/a<0.999" is not correct
It should be:
for your a and b>10000 is (a-b)/a < 0.9995

now f(10000)<1
and f(20001)<0.9995^10000*f(10000) (same reasoning as above).
this is less then 0.003 and
f(b) for your b is even smaller, so the conclusion is correct.




This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.