Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fritz 9- Boom or Bust?

Author: Derek Paquette

Date: 13:18:22 09/02/05

Go up one level in this thread


On September 01, 2005 at 21:43:57, Tony Nichols wrote:

>On September 01, 2005 at 21:25:48, Derek Paquette wrote:
>
>>On September 01, 2005 at 19:17:56, Tony Nichols wrote:
>>
>>>Hello George,
>>>  I am not a programmer, but I definitley believe there is a difference between
>>>beating other programs and beating human players. All chess engines are strong
>>>in tactics. Many GM games are lost just because a player missed a tactic. If you
>>>design your program to value tactical positions it would probably do better
>>>against humans. However, Against other programs this would not be the best way
>>>to win. Other programs would very often handle the tactics well and then what
>>>have you got? It's funny to hear the statement from chessbase that their
>>>concentrating on beating human GM's. This might have been an interesting goal 10
>>>years ago. Today amatuer programs are beating GM's, So what is chessbase really
>>>saying? They also claim that their trying to make Fritz more of an instructive
>>>tool. I am all for this, But they don't really say how.
>>>Regards
>>>Tony
>>>
>>>P.S. I of course will buy Fritz 9 as soon as it comes out:)
>>
>>There is no evidence for this however, HiarcsX (hiarcs9) are argueable the
>>strongest positional programs and hiarcs couldn't even win a single game vs
>>Bareev who didn't have an army of grandmasters plotting his moves ahead of time
>>for him.
>>
>>It couldn't come up with the win in my opinion because it simply was not strong
>>enough.
>>
>>6 games is a lot of games to get atleast ONE win
>
>Based on my experience as a chessplayer, I have seen many human games lost
>because of tactics. Even at the highest level. If a program was designed to play
>for highly tactical positions it would surely get good results. Your right I
>have no proof other than the 1,000's of games lost because of tactical
>oversights.
>   You can't win a game of chess unless your opponent makes a mistake. Good
>players know what types of mistakes their opponent is prone to make. Humans make
>tactical mistakes more than any other kind. If you want your chess program to
>beat humans, Then play tactical positions. This seems self evident to me.
>Regards
>Tony

You know that does make a lot of sense, however I would like to add if you
create an all around program that can beat programs at a top level (toga, zappa,
fruit, shredder) then they inevitably know how to play a tactical game aswell as
a positional game, so when the human makes a mistake it is likely they will be
able to take advantage of it.



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.