Author: Matthew White
Date: 11:19:42 07/02/03
Go up one level in this thread
On July 02, 2003 at 13:55:45, Vincent Diepeveen wrote: >Because "i'm good with computers" i'll be contributing from different >grandmasters and national trainers of Netherlands here. > >At the phone i have IM Hans Bouwmeester. > >He prefers to call it: > >PlayLikeCapaBlanca > queen = 9 > rook = 5 > pawn = 1 > knight = 3 > bishop = 3.5 > opening = 1.e4,e5 2.Nf3,Nc6 (forced) > acceptsacrafices = forced > capturing = forced > grabbing = good > arrogance = 100 > surrendering = never >Ugh > >The idea behind this is because i had said something to him which has caused him >to write for a bunch of magazines a big article. I said that the games of the >old grandmasters and world champs were not so good when checked with the >computer. In fact i said that todays world top is a lot stronger, apart from >that they know more from openings of course. > >I said in fact it is because of training, better knowledge, more practicing and >more experience in the different openings where it no longer is forced to play >e4 e5 nf3 nc6. Of course Hans notes that accepting sacrafices at the time >was forced because it would insult possibly the opponent if you didn't take it. > >Therefore he says capturing is forced. > >Of course because i humbly write this down for my good old team member, the >meastro of the dutch federation who still played last season masterclass >competition far in his 70s. > >Now having a scripting language that accepts any variable names is a good idea >because all the code is going to run smoothly then. Some people like Chrilly >Donninger are not so good in english. That's why he has put it in german. > >PlayLikeNimzo > Damen = 11.5 > Turm = 5.6 > Laufer = 3.7 > Springer = 3.5 > Bauern = 1.0 >Ugh > >Nimzo doesn't have much more language than this in its evaluation as it is a >preprocessor and to quote chrilly as he said to me in 1998 when che was on >topic: "nimzo played better when i removed knowledge. I do not believe in >knowledge". > Interesting that he says that nimzo plays better without the knowledge, so he removed it. I wonder why he didn't question the knowledge, rather than saying that knowledge is bad, so he should remove it. In the Che++ code that is included with Nimzo, there is code to the effect of "If there is a blocked center, then launch a wing attack." I'm no master, but I do know that a wing attack is worthless if you don't have piece support. I have, in fact, seen Nimzo launch a wing attack with no pieces supporting the pawns. Doing so shows a total lack of understanding of the point of a wing attack... It's (generally) to break through with the pawns to give the pieces an inroad to the king! However, since Nimzo doesn't know this, it is perfectly content to push the pawns forward until the opponent blocks up the position and turns a win for Nimzo (given proper piece support to the attack) into a sure draw (since there is no longer a way to break through). Chrilly's aversion to knowledge might also explain why Nimzo is so weak in endgames. Even with the in-memory endgame tables, a program won't be able to play endgames effectively if it doesn't realize that you don't want the opponent's king blocking your only pawn if you are going for the win... I don't think the knowledge itself is the problem, but rather the QUALITY of the knowledge :). Matt
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.