Author: Eelco de Groot
Date: 02:07:17 06/14/01
Go up one level in this thread
On June 13, 2001 at 10:16:42, Terry McCracken wrote: >On June 13, 2001 at 04:17:17, Eelco de Groot wrote: > >>Terry, >> >>In case you or or other competitors missed it I posted a snippet of computer >>analysis on 36. ...Bc6 here: >>http://www.icdchess.com/forums/1/message.shtml?174727 >> >>Last lines after 30 hours it is still 36. ..Bc6: >> >>[D]6k1/5p1p/6p1/3p2Pn/b2P1P2/pnN1BB2/7P/6K1 b - - 0 1 >> >> >>Rebel Engine for ECTool. (c) Ed Schröder >> >>Engine version : REBEL CENTURY 2 >>Rebel personality : "Q5T" >>Opening book file : RebelC.mvs >>Hash table size : 12 MB >>Analysis mode : Analyzing next move >>Refresh interval : 500 ms >> >> >> >>00:00 06.00 0.09 1..Bb5 2.Bxd5 Ba4 3.Bc4 Na5 4.Bd3 >>00:01 06.02 1.38 1..Nxd4 2.Bxd4 Nxf4 3.Bxd5 Nxd5 4.Nxd5 h5 5.gxh6 f5 6.Nf6+ >>Kh8 7.Nd5+ >>00:01 07.00 0.80 1..Nxd4 2.Bxd4 Nxf4 3.Bxd5 Nxd5 4.Nxd5 a2 5.Nf6+ Kf8 6.Nxh7+ >>Ke7 7.Nf6 Bc6 8.Kf1 >>00:02 07.11 1.45 1..Ng7 2.Bxd5 Nf5 3.Kf2 Nxe3 4.Kxe3 Nc1 5.Nxa4 >>00:02 08.00 0.93 1..Ng7 2.Bxd5 Nf5 3.Kf2 Kg7 4.Bc4 Nxe3 5.Kxe3 Nc1 6.Nxa4 >>00:03 08.08 1.22 1..Bc6 2.Bd1 Nxd4 3.Bxd4 Nxf4 4.Kf2 Ne6 5.Ke3 Nxg5 >>00:04 09.01 1.20 1..Ng7 >>00:09 10.00 1.70 1..Bc6 2.Bxh5 gxh5 3.Na2 Bd7 4.Nc1+ Na5 5.Bd2 Nc6 6.Bc3 Kg7 >>00:16 11.00 1.25 1..Bc6 2.Bd1 Nxd4 3.Bxd4 Nxf4 4.Kf2 Nd3+ 5.Ke3 >>00:18 11.01 1.36 1..Ng7 2.Bxd5 Nf5 3.Bxb3 Bxb3 4.Bc1 a2 5.Bb2 Kf8 6.d5 Ke7 >>7.h3 >>00:33 12.00 1.34 1..Ng7 2.Bxd5 Nf5 3.Bxb3 Bxb3 4.Bc1 a2 5.Bb2 Kf8 6.d5 >>01:08 13.00 1.24 1..Ng7 2.Bxd5 Nf5 3.Bxb3 Bxb3 4.Bc1 a2 5.Bb2 Nxd4 6.Nxa2 >>01:49 13.01 1.24 1..Bc6 >>02:07 13.01 1.46 1..Bc6 2.Kf2 Nxf4 3.Bxf4 Nxd4 4.Bg2 Ne6 5.Be5 d4 >>05:34 14.00 1.19 1..Bc6 2.Kf2 Ng7 3.Bd1 Na5 4.Bc1 Nc4 5.Bb3 Nf5 6.Ne2 >>27:51 15.00 1.75 1..Bc6 2.Kf2 f5 3.Bxh5 gxh5 4.Na2 Bb5 5.Nc1 Na1 >>55:23 16.00 1.74 1..Bc6 2.Kf2 f5 3.Bd1 Nxd4 4.Bxd4 Nxf4 5.Be3 Nd3+ 6.Ke2 Nb4 >>52:53 17.00 1.44 1..Bc6 2.Kf2 Ng7 3.Bd1 Na5 4.Bc1 Nc4 5.Bb3 >>27:26 18.00 1.60 1..Bc6 2.Kf2 h6 3.h4 Nxf4 4.Bxf4 Nxd4 5.Bg2 >>39:10 19.00 1.34 1..Bc6 2.Kf2 h6 3.f5 hxg5 4.Bd1 (20:39:10) >> >>Not that I quickly see why the comp thinks this is any better than Ng7 but just >>for comparison. But I think it is too soon to declare it a draw unless not >>enough persons want to play on. >> >> Eelco > >Eelco I agree that it's up to the voters as a whole, whether to play on or not >regardless if it's a draw. > >However, "Q5T" doesn't know what GT 2.0 will do with 36..Bc6 and I do. >Unless it changes it's moves at a very high depth, if 36..Bc6 37.Bd1 if 37..Ba4 >38.Bf3..Ba4 39.Bf3..Ba4 and draw by 3 Fold Repeat. (Unless the voters want to >revert to the rejected Na5?! line which _does_ draw without hope for more.) >GT 2.0 doesn't consider Kf2! It goes for the draw! > >If people want to try for a win then I would highly recommend 36..Ng7 with an >edge for Black. >The only 2 programms to my knowledge that at high plys go for 36..Ng7 are GT 2.0 >and Junior 6. There must be others? >Fritz 6 only likes 36..Ng7 when you substitute the move in place of 36..Bc6 >otherwise it just appears to keep working on 36..Bc6 till "The Cows Come >Home"!;) Maybe at ply 30 it would change it's "mind"? Who knows. But even at ply >23 to 24 it's still on 36..Bc6. >Gambit Tiger II changes by ply 21 or 22 to 36..Ng7 giving an evaluation of -/+ >38. So it likes 36..Ng7 as it puts more wait on the play and appears to avoid >draws with the exception of repeating moves. >Junior 6 gives an evaluation of -/+ 30 for Black with 36..Ng7 at depth 23. >Very close to Gambit Tiger II. >My recommendation then to attempt to play for a win would be 36..Ng7 since I >know how GT 2.0 evaluates this position. >Or play 36..Bc6 37.Bd1..Ba4 38.Bf3..Ng7, just to stall but nothing more. >However, that would buy some analizing time!:-) >The game is most complex after Black plays Ng7. Will it win, not likely, but >programms are fallible and it might take a wrong turn in the endgame. >It will prolong the game at least 20 more moves before people gave up at the >very least. Depends on the people playing of course;) > >Regards, > Terry I see Terry thanks for explaining and the lines. I let Q5T analyze without the game history and that may change things. If I enter the same position again and then one repetition Q5T now goes for Ng7 (36. ..Bc6 37.Bd1 Ba4 38.Bf3 Ng7) So I think there is a chance that humans, or computers, might want to do the same. Difficult to predict what is the best move to vote for to comply with the majority in this case I'm afraid.. Eelco
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.