Author: Nicholas Cooper
Date: 19:18:31 01/04/04
Go up one level in this thread
On January 04, 2004 at 09:48:10, George Tsavdaris wrote: >On January 04, 2004 at 06:15:37, Mark Young wrote: >>>NO. You can NEVER say Shredder is better than Bam Bam! (meaning that the >>>probability to be better is 100%). You can only say Shredder is better than >>>Bam Bam with a probability 99.99999999878%. Although some people say this is >>>100%, this is wrong. >> >>You need to stop being cute and read what I posted. I never said better, I said >>stronger. And I never said 100%. I said I don't need 10,000 games. > >1) What do you mean when you say: "Shredder is stronger than Bam Bam"? >It must mean 100%. If not, then you have to say how much stronger. For example: >"Shredder is C % stronger than Bam Bam". >It isn't reasonable if you say:"Shredder is stronger than Bam Bam" and you mean >"Shredder is stronger than Bam Bam with a probability of 88%" . >Sorry if i am amazingly pedantic but ..... > >2)I think the statement "A is better chess player than B" is exactly the same >with "A is stronger in chess than B". I can't find a difference and i can't >understand where do you find the difference. > >> >>I'm not 100% sure your were trying to be cute only 99.997%.:) > >I don't try to be cute, only as possible accurate as i can :) More like anal! Yes, you are being amazingly pedantic as there are few statements one can make with 100% confidence. That is why a confidence interval is chosen, which is typically 95% or 99% to make a statistically significant statement. The basic thrust of this thread, namely that many games are usually required to reach a statistically significant conclusion is valid, albeit with the qualification Mark mentioned about less games being required when the programs are of vastly different strengths. If we all tried to be as accurate and pedantic as you obviously enjoy, then it would take a long time to say anything useful owing to the large number of qualifications required! It is clear to me that Mark understands what he is talking about, though I should qualify that by saying I have no statistical basis for making such a claim, only an intuitive one based upon what he has written on the subject! ;)
This page took 0.01 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.