Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: what program is best to play correspondece chess against humans?

Author: Kurt Utzinger

Date: 08:42:11 03/10/03

Go up one level in this thread


On March 10, 2003 at 10:37:36, Uri Blass wrote:

>On March 10, 2003 at 10:33:31, emerson tan wrote:
>
>>On March 10, 2003 at 07:43:25, Uri Blass wrote:
>>
>>>On March 10, 2003 at 07:27:45, Jorge Pichard wrote:
>>>
>>>>On March 10, 2003 at 06:40:07, Drexel,Michael wrote:
>>>>
>>>>>On March 10, 2003 at 05:58:13, Jorge Pichard wrote:
>>>>>
>>>>>>On March 10, 2003 at 05:34:07, emerson tan wrote:
>>>>>>
>>>>>>>what program is best to play correspondece chess against humans?
>>>>>>
>>>>>>Unless you are playing against Victor Palciauskas, Shredder 7 or Deep Fritz
>>>>>>should be excellent programs to analyze any position.
>>>>>
>>>>>Not exactly true
>>>>>
>>>>>Now the challenge would be
>>>>>>the chess knowledge the the user has, in case the program choose the wrong move,
>>>>>>since you have to be able to select the best move based on your knowledge
>>>>>>of the position. This is where the combination of a good program such as
>>>>>>Shredder 7.04 and a good correspondence chess player like Steve Ham can make a
>>>>>>world of a difference.
>>>>>>
>>>>>>Jorge
>>>>>
>>>>>Shredder 7 is not the best for Correspondence game analysis.
>>>>>The evaluations are not stable and therefore you cant trust them.
>>>>>It show you +1.22 in a position and you think this has to be big advantage for
>>>>>white and when you make the move suggested you get -0.09. You make another move
>>>>>and you will get +1.34 again. and so on.
>>>>>
>>>>>Fritz 8.008 or Fritz 7.008 are much better for Analysis mode.
>>>>>Shredder 7 might be good for over night analysis.
>>>>>It is also good for the endgame and it might be good to find ideas in the
>>>>>middlegame.
>>>>
>>>>
>>>>Probably I was wrong to recomend Shredder 7.04, but in correspondence Chess even
>>>>if you choose Fritz 8.0008 or Hiarcs 8 for Analysis mode it won't be sufficient
>>>>against a strong correspondence player; not to even mention  Victor
>>>>the World correspondence chess champion. I remembered back when Mike Valvo
>>>>crushed pretty badly DT in a 2 games match. Mike crushed it even though DT was
>>>>thinking many hours per move, and hitting 30 plies in its search.
>>>
>>>I do not believe that deep thought hit 30 plies brute force in their search.
>>>I even do not believe 20 plies in the middle game for deep thought.
>>>
>>> I remember
>>>>reading how shocked the programmers were at how badly the program lost. But
>>>>probably by now chess programs such as: Hiarcs 8 and Deep Fritz have progressed
>>>>to the point where they don't get crushed by IM like Mike Valvo, but players
>>>>like Victor Palciauskas (Correspondence World Champion) can still beat any
>>>>program in Correspondence Chess.
>>>>
>>>>Jorge
>>>
>>>I doubt if they can beat it without computer help.
>>>It is known that top correspondence chess players use
>>>computers to help them.
>>>
>>>I also have doubts if Palciauskas can be sure of beating chess programs
>>>even when computer help is allowed.
>>>I suspect that the computers have good chances to get at least a draw.
>>>
>>>I think that a player with many computers only to analyze chess positions may
>>>have a significant advantage in correspondence games because I believe that most
>>>of the correspondence
>>>players use only one or maybe 2 computers for their games and they do not
>>>give their computer a full time job in analyzing chess position.
>>>
>>>It may be interesting to know how many computers does Palciauskas use for his
>>>games(I will not be surprised if we find  that he has one computer for every
>>>opponent).
>>>
>>>Uri
>>
>>
>>I think a chess player who has chess knowlege and is good at generating ideas
>>will become a good correspondence player, you dont need to be tactically gifted
>>(able to visualize part) because you can work it out by moving the pieces. The
>>computers cannot still generate ideas and they are still limited by the horizon
>>effect, running the computer 24 hours can hardly reach 25plies on early
>>positions. Correspondence chess is the next challenge of computer after it beats
>>all human players over the board.
>
>I think that running the computer for 24 hours can help.
>I do not say that chess knowledge cannot help but these things do not
>contradict.
>
>It is possible that the same player also think about ideas in the same time that
>his programs work and decide which positions to analyze based on his ideas.
>
>Uri

To strictly follow the computer moves [even after a thinking time of many hours]
is nonsense. It is much more important to check ideas and tactics. In
correspondence chess you can go move by move deeper and deeper just analysing
each [own or suggested computer] move. But you should analyise with the computer
over the board. It's of course a time consuming matter, but that's
correspondence chess anyway.
Kurt



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.