Author: Howard Exner
Date: 12:13:12 01/03/00
Go up one level in this thread
On January 03, 2000 at 13:38:12, Chessfun wrote: >On January 03, 2000 at 08:38:40, Howard Exner wrote: > >>On January 03, 2000 at 05:44:17, allan johnson wrote: >> >>>On January 03, 2000 at 01:29:06, Chessfun wrote: >>> >>>>On January 02, 2000 at 22:41:19, allan johnson wrote: >>>> >>>>> With all the talk on how strong programmes are today I >>>>>decided to put a few of my programmes to the test.The one >>>8>>I have submitted comes from The Big Book Of Combinations >>>>>edited by Eric Shiller.So far I have tested only Rebel Century >>>>>which fails to find the best move(recommended by Shiller)in >>>>>10 minutes.My machine is a Celeron 433 64 mb ram.Is this >>>>>position too complicated for computers to solve? I'd >>>>>appreciate feed back on responses from Fritz6 and Hiarcs7 >>>>>and the like. >>>>>Thanks Allan >>>>>r4k1r/1p3pp1/2b1p1np/p3P3/3PN1q1/4Q1P1/5P1P/R1R2BK1 white >>>>>to move >>>>>Rxc6! pxc6 >>>>>Be2 >>>> >>>>Hiarcs 7.32 Cel 433 >>>>1. f3..Qh5....line ends at depth 10/30 after 10 minutes. >>>> >>>>Fritz 5.32 >>>>1. f3..Qh5....line ends at depth 13/40 after 10 minutes. >>>> >>>>Fritz 6 >>>>1. f3..Qh5....line ends at depth 14/38 after 10 minutes. >>>> >>>>After feeding Hiarcs 7.32 1. Rxc6...bxc6 2. Be2 it evaluates: >>>>2.....Qh3. 3. Nd6 +0.71 after feeding these two moves in it cannot >>>>find many moves it likes for black. >>>> >>>>I myself think this type of position is easy for a human to see Rxc6 >>>>than for a computer. >>>> >>>>Thanks. >>> >>> Yes I agree but I'm puzzled as to why! Will programmers ever find a >>>means by which they can get their programmes to make temporary inferior >>>exchanges?Surely the plies required to determine Rxc6 is winning(admittedly >>>not everwhelmingly) is not that deep? >> >>1. f3 Qh5 2. Rc5 a4 3. Bb5 >>This short sequence of moves also looks good for white. This may be a line that >>computers are looking at. Black is in quite a bind here with the poor king >>position and lame rook on h8. >> >>It's a nice position as two different paths lead to an advantage. > >While I agree the computer line 1. f3...Qh5 2. Rc5...a4 3. Bb5 is also good for >white and black has enough problems to solve. The line with 1. Rxc6...bxc6 >2. Be2...Qh3 3. Nd6 appears to be even better. I think it appears that Rxc6 is >better than f3. 1. Rxc6 bxc6 2. Be2 Qh3 3. Nd6 Nh4! 4. Qe4 Qg2+ 5. Qxg2 Nxg2 6. Kxg2 Ke7 7. Nc4 a4 8. Nb6 Ra5 9. Rxa4 Rxa4 10. Nxa4 Rd8 Yet black is in fairly good shape after this line, so I'm not convinced of the superiority of Rxc6 over f3. But there could be some better way for white to proceed, although 3 ...Nh4 forces the play into a line that eliminates some of black's weaknesses(poor king safty, immobile h8 rook). What line are you getting with the Rxc6 choice? 1. f3 Qh5 2. Rc5 a4 3. Bb5 extending this further in a sample line could end up like 3 ... Ne7 4. Bxc6 (again getting rid of the nice black bishop) bxc6 5. Rc4 This I prefer over the initial Rxc6 line. At first I thought both were good but now I like the not so flashy f3. This is all from Schiller's book, "The Big Book of Combinations". I have this book and the one drawback to it is that it only gives the initial move as best with no further analysis. Perhaps the players missed the move Nh4, or then again maybe I'm missing something here also. Chess, as your handle sums up, is lots of fun for me when it comes to analysing interesting positions with other players (and computers). It's satisfying to come to an understanding of a position through this kind of exchange with other ccc members.
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.