Author: Roberto Waldteufel
Date: 17:38:28 06/01/98
Go up one level in this thread
You are not alone in feeling like a beginner! You have done well to reach the stage of having a good running program. I would suggest that you read some of the more recent posts from the archives, as these topics have been discussed recently. There are free endgame databases (known as tablebases), but there structure is quite complex. Alternatively, you could try to construct your own: you first have to identify and score the terminal positions (eg mate, stalemate, exchange to a simpler won ending etc), and then recursively backtrack to find " win in 1", "win in 2" etc positions until you make a pass without scoring any new positions, at which point all the remaining positions must be draws. Regarding openings, there are essentially two approaches, what might be call fast and dumb or slow and smart. The easiest method - probably the only practical one if you want a large opening book, is to read in a large number of top notch master and grandmaster games from a PGN file, and statistically analyse all the positions from all the games up to some arbitrary depth (eg 20 or 30 plies from the start position), noting how often each move was made from that position. In doing this, you should probably only include the moves by the side that won the game (the loser must have played at least one bad move). You can then use these positions for your opening book, with the masters' moves stored with the positions. It is up to the program to decide between popular alternative moves from the same position. The other method is very slow and tedious, and requires quite advanced chess judgement. It is to select yourself which opening suit best the programs own style, and then construct the program's opening repertoire around its own strengths. This is more akin to the way opening reoertoires used to be put together by top grandmasters in the old Soviet Union for up and coming young talents in the so called "Pioneers Palaces", where most of the top Russian players studied the game. I think this is the best method, but it would take years to do it. Hope some of this is helpful, Roberto
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.