Author: KarinsDad
Date: 14:56:44 03/02/99
Go up one level in this thread
On March 02, 1999 at 16:15:48, Marc van Hal wrote: >In computerchess it is just positional play and how to build up an >strategicalplan is what is the most important Actually in computer chess, there is very little difference between tactics and strategies. Now before everyone jumps on my case about this, let me explain. Every time a computer program can search one more ply deeper, regardless of whether it is due to an algorithmic improvement or a faster computer, the program gets slightly better (on average; a program could make an inferior move to a previous version of the program for certain positions). If good chess was based on tactics and great chess was based on strategy, this would not be the case. The programs would reach a tactical limit where they come up with a series of tactical moves which may or may not help the position strategically and they would still get crushed strategically by GMs. What is the true definition of a strategic move? It is a move which does not tactically harm the current position, but at the same time sets up a tactical advantage later in the game. 1. D4 is an example of such a move (this is also tactically good immediately by opening lines for other pieces). Most opening book moves (hence variations) are both tactical and strategic since the "strategically poor but tactically ok" moves get pruned out of the "opening books" when they are found. The reasons that humans "perceive" such a great difference between tactics and strategy are that 1) humans do not perform as deep a search as programs (on average) and 2) humans play on intuition and experience. If the position does not require an immediate tactical response, humans can modify their thinking to look for either a tactical possibility or a strategic possibility. Computer programs are (for most programs) always looking for a tactical possibility. However, due to special programming and/or pure circumstance (such as the best tactical move is also a good or the best strategic move), they can appear to be playing strategically. Since the human mind does not work like a chess program, it is easier for humans to gain strategies by studying tactics (as per KKs original assertation that tactics is the best thing for nonmasters to study) since human pattern recognition enables the human who has studied many tactical positions to make sacrifices and other strategic moves which s/he has not determined ahead of time to be tactically sound. The move just feels right based on the position and the human does not see an immediate tactical refutation. KarinsDad >In the games where Anand for example punished Fritz5 because the program puts >the pawns on the blackaquares whenhe has the black bischop also the thread of >capture is to big wich leads too unclear positions If you would ask if this treu >too the same Grandmasters they will agree but ofcourse Komputer Korner is right >that tactics is important for grandmasters International masters and high rated >club players but also for amateur players to learn first to learn how to play >positional and how to build up a strategicalplan and only the simple tactics is >more important to start with and onlythen learn to play and understand the more >advanced tactics >I know this from my one practice and from manyplayers who came higher rated at >their chessclub many times you here that it was the book Nimzowitsch mein system >or Juspov Devorsky's book positional play
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.