Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Did GM Michael Adams 2737 honestly lose?

Author: Mark Young

Date: 18:11:02 06/26/05

Go up one level in this thread


On June 26, 2005 at 21:06:15, Rolf Tueschen wrote:

>On June 26, 2005 at 20:03:28, Mark Young wrote:
>
>>On June 26, 2005 at 18:54:24, Rolf Tueschen wrote:
>>
>>>On June 26, 2005 at 18:08:49, Mark Young wrote:
>>>
>>>>On June 26, 2005 at 16:25:35, Rolf Tueschen wrote:
>>>>
>>>>>On June 26, 2005 at 15:02:56, I Hart SanQuentin wrote:
>>>>>
>>>>>>On June 26, 2005 at 14:27:44, stuart taylor wrote:
>>>>>>
>>>>>>>No, of cause not! Not enough games! It needs atLEAST 500 games between the two
>>>>>>>to get a rough idea.
>>>>>>>
>>>>>>>S.Taylor
>>>>>>
>>>>>>No not 500 games a 1,275,326 games otherwise there is no likelyhood that the
>>>>>>comp is better than mickey, it was just the most amazing streak of bad luck you
>>>>>>ever saw that he is down 4.5 to .5
>>>>>
>>>>>
>>>>>In case you made ironic remarks let me state that indeed such a result does not
>>>>>mean anything (in reality and also in statistics). You dont believe me? Well,
>>>>>remember the match between JUNIOR and FRITZ a couple of years ago. JUNIOR was
>>>>>clearly leading and FRITZ still won the match. In short: 6 games is nothing
>>>>>relevant.
>>>>
>>>>As you most likely know. It is not just the number of games. It is the score
>>>>that also must be considered. The closer the score the more games that are
>>>>needed to show which player is really stronger.
>>>>
>>>>If GM Adams draws tomorrows game, and the score ends 5-1 for Hydra. There is a
>>>>almost a 90% certainty that Hydra was stronger then GM Adams. Who is #7 in the
>>>>world.
>>>>
>>>>If GM Adam loses, then we can say with even higher certainty then 90%. That
>>>>Hydra is stronger then GM Adams. IF Hydra loses tomorrow with a 4.5 - 1.5
>>>>winning score for Hydra, I agree. The results are nothing relevant. Meaning you
>>>>could not say that Hydra is stronger then GM Adams with any kind of great
>>>>certainty.
>>>
>>>
>>>Mark, you are wrong with your statement. For 6 games any score is without
>>>relevant meaning. You cant assume any certainty with 6 games. But of course
>>>score means something and you need less games for a clear score. Again, dont you
>>>remember the famous qualification match in Cadaques between JUNIOR and FRITZ?
>>>JUNIOR led by 4-0 or was it 5-0? FRITZ still won.
>>>
>>>But honestly in case of Adams I wouldn't expect anything in his favour. :)
>>
>>I am sorry.... I gave that stats, and they are correct. With a draw tomorrow you
>>can say with 90% confidence that Hydra is better then GM Adams. That means a 1
>>in 10 chance,  GM Adams my be better then Hydra. With a win tomorrow for Hydra
>>the chances are even lower for GM Adams.  GM Adams must win tomorrow for this
>>match to not have much meaning.
>>
>>You can accept this or reject it.....be the stats are correct.
>
>
>No. You forget about the combination of the two factors score and number of
>games.

I am sorry again, but you are incorrect. To calculate the stats. You must have
the number of games, and the score. The stats are the stats.

As I said before. The stats are correct, but I see you want to reject the
stats....



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.