Author: Uri Blass
Date: 08:04:50 06/05/02
Go up one level in this thread
On June 05, 2002 at 10:56:08, Uri Blass wrote: >On June 04, 2002 at 13:12:40, J. Wesley Cleveland wrote: > >>On June 03, 2002 at 22:50:10, John Merlino wrote: >> >>>On June 03, 2002 at 22:39:35, John Merlino wrote: >>> >>>>On June 03, 2002 at 19:06:31, Dieter Buerssner wrote: >>>> >>>>>On June 02, 2002 at 22:32:25, Dana Turnmire wrote: >>>>> >>>>>>-- -- -- -- -- -- -- -- >>>>>>-- -- -- -- -- -- -- -- >>>>>>-- -- -- -- -- -- BK -- >>>>>>-- -- BP -- BP -- -- -- >>>>>>-- -- WP -- WP -- WK -- >>>>>>-- -- -- WN -- -- -- -- >>>>>>-- -- -- -- -- -- -- -- >>>>>>BB -- -- -- -- -- -- -- >>>>>> >>>>>>Here is the test position found in "The Mammouth Book of Chess." >>>>>> >>>>>>1Nxc5? >>>>>> >>>>>>This obvious, materialistic move throws away the win. White's c-pawn cannot be >>>>>>advanced to the queening square without the help of the king, but this allows >>>>>>Black to counter by attacking the e-pawn. White should not take the c-pawn, but >>>>>>instead make progress on the kingside by manoeuvring his knight. >>>>> >>>>>Can you please show a (sample) winning line. My program went also for Nxc5 for 4 >>>>>hours (on rather slow hardware). I tried to understand the pos myself. All I >>>>>came up with - I cannot find improvement for white after this move, and I think, >>>>>it is draw. However, I have no better idea, that secures the win. >>>>> >>>>>Regards, >>>>>Dieter >>>> >>>>We just happened to have that book laying around, and since I hadn't done ENOUGH >>>>typing today, I figured I'd quote the entire passage from "The Mammoth Book of >>>>Chess", pg. 386, under the section entitled "Computer Chess". >>>> >>>>------------ >>>>From Pachman-Hromadka, Prague Ch, 1944 >>>> >>>> Nxc5? This obvious, materialistic move throws away the win! "White's c-pawn >>>>cannot be advanced to the queening square without the help of the king, but this >>>>allows Black to counter by attacking the e-pawn" - Pachman >>>> Supposing it were instead Black to move, the following variation is >>>>enlightening: >>>>1...Bd4 2.Ne1 Bf2 3.Nf3 Kf6 (or 3...Bd4 4.Nh4+ Kf6 5.Nf5) 4.Kh5 Bg3 5.Nh4! Bf2 >>>>6.Nf5 Bg1 7.Nh6 Bd4 8.Ng4+ Ke6 9.Ng6 >>> >>>The last move here should be 9.Kg6, of course.... Sorry about that.... >>> >>>> >>>[D]8/8/4k1K1/2p1p3/2PbP1N1/8/8/8 w - - >>>> >>>> White will now play Nf6-h7-g5+, etc., and win easily. However, if there were >>>>no black pawn on c5, then Black would have sufficient counterplay to hold the >>>>draw, since his king could use the c5-square to attack White's pawns. >>>> Thus, in our start position, White should not take the c-pawn, but instead >>>>make progress on the kingside by manoeuvring his knight. >>>>------------ >>>> >>>>So, the question is, is that correct? Anybody care to throw some serious CPU >>>>cycles at it? >>>> >>>>jm >> >>I let crafty run overnight and got nothing. > >Did you tell it not to use null move pruning or does it automatically avoid null >move pruning in KBPP vs KNPP when the pawns are blocked? > >I did not analyze the position but it is clear that if white can win it without >capturing the pawn it is only by a lot of zunzwangs. > >Uri Not sure about it but from computer point of view there are going to be a lot of moves that seems not to have a threat so it is probably better not to use the null move algorithm for solving of that position. Uri
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.