Author: Herman Hesse
Date: 03:57:42 07/29/99
Go up one level in this thread
On July 27, 1999 at 20:43:32, syed wrote: >How necessary is it to be a good chess player in order to be a good chess >programmer? In my understanding the lietmotif of the chess engines is high quantities of low quality. From the one simple evaluation over millions of tree nodes, come entsch. complex concepts. The programmer does not / need not understand these, nor need to program them. The programmer codifies the rules of chess moves and simple evaluations. Search algorithm, already known, coded, and freely available, does the rest. Little chess playing knowledge is needed by the programmer. More important are programming techniques of speed and efficiency. Until. Then later, the programmer discover that chess is not all tactical and solvable, and that general rules can make application. Some rules are simple to reduce to the mathematik that is easy for programmer understanding, like square of pawn. Some rules need simple pattern recognition, like bishop trap on a7, pawn b6 and pawn c7. Save search of many plys. All easy so far for 1700-1900 chess programmer. With simple rule and pattern program no longer fall in stupid traps and tactic and understanding from search component is enough to defeat even close to IM's. Except. Stronger players than these understand how program thinks. Understand of thinking is enough. But programmer is so far from IM plus understanding, that cannot see general rules and patterns needed. No also can see complex ideas needed. Cannot code. Only now reliant on making search space larger or MTD(f) or SSS* or new paper idea. All same idea. But no problem. Quantity of opponents are no IMs. All are other programs or weak humans on ICC. Big fight between all at low levels. Mud is stirred round and round. Looks if as never know. Never look even. No matter. Not important. Herman. >I figure the basic core of the program is pretty much standard using >common algorithms. I am referring to the debugging process. If I find my program >losing a game, how do I know that my judgement of where it went wrong is 100% >technically correct, so as to go in amd make necessary adjustments. Looking at >the posts here, on analyzed computer games, I see quite a few people making a >level of analysis which would be equal to that of a senior club player. > >Arshad
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.