Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: To Uri and Hyatt

Author: Uri Blass

Date: 20:39:06 01/18/05

Go up one level in this thread


On January 18, 2005 at 23:05:47, Robert Hyatt wrote:

>On January 18, 2005 at 11:48:31, Uri Blass wrote:
>
>>On January 18, 2005 at 11:16:16, Robert Hyatt wrote:
>>
>>>On January 18, 2005 at 08:17:27, Madhavan wrote:
>>>
>>>>
>>>>>>i deny that,strongest program running on a faster hardware should not get a draw
>>>>>>or lose in many games against super grandmasters,if it does then it is
>>>>>>considered as not solved
>>>>>
>>>>>Solution of chess only means that the solver will never lose a match.
>>>>>It does not mean not getting a draw or a loss.
>>>>
>>>>
>>>>then consider there are 5 grandmasters and 1 program playing in the tournament
>>>>that program is said to be "chess solvable",it does not lose to any of the
>>>>grandmasters but drew with 2 grandmaster
>>>>one of the grandmaster drew the machine but won all the game against other
>>>>grandmasters,then that grandmaster will be declared as event winner,but loses
>>>>few game in another event then what is your point?
>>>
>>>
>>>That wasn't my definition.  When chess is solved, we will know, from the initial
>>>position, whether white wins or loses or draws.  And for any white first move,
>>>we will know the correct black first move to preserve the game status as defined
>>>by white's first move.  Repeat forever.
>>
>>
>>We can know the result of the initial position without knowing everything later.
>>
>>You can know that chess is a draw if you only know 1 and 2:
>>
>>1)1.e4 leads to draw.
>>2)For every alternative of white black has a move that achieves at least a
>>draw(we do not need to know if it is a win or a draw)
>>
>>It means that you may know only that 1.d4 Nf6 is at least a draw for black and
>>know nothing about the results of 1.d4 d5 so we do not know the best reply
>>against 1.d4 but only that 1...Nf6 is good enough to draw(1...d5 may win and be
>>better than 1...Nf6).
>>
>>Uri
>
>
>Solved doesn't just mean I know it is won for white.  It means I can _win_ it
>with white against any black defense.  Hence I need all the information about
>the tree to play the game perfectly.  Otherwise I play 1.e4 and if black doesn't
>resign because that is the only winning move for white, what do I do to ensure I
>win?  :)

If you are interested only in the theoretical result of the game
You need to know to win after 1.e4(otherwise I do not see how you prove that
white wins) but you need to know nothing for black after 1.d4

You may not play perfect with black but you will never lose a fair match (in the
worst case you may lose instead of draw after 1.d4 of the opponent but you will
always win with white).

Uri



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.