Author: José de Jesús García Ruvalcaba
Date: 07:57:50 03/31/99
Go up one level in this thread
On March 30, 1999 at 16:10:07, James T. Walker wrote: >On March 30, 1999 at 10:56:30, Sylvain Renard wrote: > >>On March 30, 1999 at 08:38:23, James T. Walker wrote: >> >>> >>>Hello Sylvain, >>>I don't think the missing tablebase should matter ! If your tablebase would >>>simply tell your program to promote to a Rook from that point on it is a won >>>game and no tablebases are needed. >> >> It should not but it does! The conception of the Thompson databases >>is different from Edward's tablebases. You find the following information >>(see readme in Thompson CD's)in these databases : BLACK TO MOVE (the weaker >>side) and lose in some number of move. So, if you have the stronger side, >>you HAVE TO MOVE first, and then check for the value of the position. >>The advantage of this method is to avoid having 2 files : one for >>white to move and another one for black to move. One file is enough, >>the space needed is divided by 2. Of course you spend more time reading >>the database. >> >>After 1.f7 Rf2 you have to play 2.f7=R to know what happens after this >>move. But after this move, you are not in Pr_r.g anymore, the program >> MUST read Rr_r.g to have the correct information. The value of this >>position (white to move and win with f7=R) is NOT included in >>the Pr_r.g file! That's why Fritz cannot find the good move without >>the Rr_r.g file. >> >> Best regards, >> Sylvain > >Hello Sylvain, >I'm sorry but you are wrong. The move f7=R _IS_ included on the CD ROM. Just >look at the move list. It is the second move on the list. The apparent error >is in the fact that all moves listed are (0) 1.xxx. The move (0) 1.f8=B is just >the first move on the list so it is picked over the others. This is simply an >error on the CD ROM disk. If it was properly scored the winning move which is >(8) 1.f8=R would be played before abandoning the tablebase and would therefore >have a won endgame of KRRKR. At that point no tablebase is required. By the >way my Fritz 5.32 with it's own endgame CD ROM plays this move correctly. The >error comes from the old Fritz 5.16 endgame CD ROM which I also use when playing >auto232 games. The program using the old CD ROM is at a disadvantage but I >usually consider that both sides have "Perfect knowledge" and if the correct >score for the position is a draw, I end the game with a draw since this would be >the proper result (With no error in the tablebases). >Jim Walker Hi Jim, sorry, but the tablebases do not include moves, they only have scores. Or at least Thompson's, Edwards' and Nalimov's tablebases do not include moves. Sylvain's explanation (missing files) is perfectly fine. I do not have Fritz5.16 or Fritz5.32 endgame CD's; but it seems to me that the former misses some files. José.
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.