Author: Anthony Cozzie
Date: 10:19:50 07/09/04
Go up one level in this thread
On July 09, 2004 at 13:00:10, Daniel Jackson wrote: >On July 09, 2004 at 10:11:52, Anthony Cozzie wrote: > >>On July 09, 2004 at 09:37:20, Daniel Jackson wrote: >> >>>The Uzbekistani GM was low on time and had played a very incautious move (Rd6). >>>After checking the lines for a few minutes Adams struck out: 31.Rxb4! axb4 >>>32.Ne7+ Qxe7 33.Qxd6. Now instead of defending the b-pawn (with 33...Qxd6 >>>34.Rxd6 Rb8 or even 33...Qe4) Kasimdzhanov went for 22...Qe2 after which Adams >>>could pick up the second pawn with 34.Rd4 h5 35.Qxb4. Everone expected Black to >>>resign, but Kasimdzhanov fought on desperately until the time control and a few >>>moves after that. 35...Qf3 36.h4 Rc8 37.Qd2 Rc3 38.Rf4 Qc6 39.Kh2 Rxb3 40.Qd8+ >>>Kh7 41.Qd1 Rb7 42.Qxh5+ Kg8 43.Rd4 Qf6 44.Rd2 Ra7 45.a5 g6 46.Qb5 Kh7 47.Qb6 >>>1-0. >>> >>>A resulting position had Kasimdzhanov exchanged Queens and played 34..Rb8. >>> >>>The play that would have ensued, 35.Rd5!..Kf8. Does your engine find (36.Rb5!!)? >>> >>>[D]1r3k2/5ppp/8/3R4/Pp6/1P4P1/5P1P/6K1 w - - 0 36 >>> >>>Daniel >> >>It looks like Black will be in zugzwang after Rb5 Rxb5 axb5 Ke7 Kf1 Kd6 Ke2 Kc5 >>Kd3 Kxb5 Kd4, but it depends who gets the move after the pawn pushes on the >>kingside. >> >>I would play Rd7 in blitz and expect to win easily. White would win the b-pawn >>long before black could extricate his king. >> >>anthony > >Rd7 certainly wins, and computers would choose that move, is my guess, I know S8 >does. I haven't tested my other programs. > >S8 will first look at Rd5 a move earlier then switch to Rd7. Rd5! and Rb5! are >"Human" moves, and lead to the simplest win. Computers focus on what is the >highest value returned, unless there is specific "knowledge" to tell them >otherwise. > >In blitz many people would play Rd7 as it's clear, it wins. However, with >experience, people are still likely to play Rd5! and Rb5! winning the ending >after Rooks come off the board. > >Years ago Rd7 would have been my choice as well, today Rd5! with the idea Rb5! >and winning the ending appears first in my mind, so even at blitz I would play >it. > > >You said, "It looks like Black will be in zugzwang after Rb5 Rxb5 axb5 Ke7 Kf1 >Kd6 Ke2 Kc5 Kd3 Kxb5 Kd4, but it depends who gets the move after the pawn pushes >on the kingside." > >Now look at the position. How can Black's K-Side pawns ever hope to defeat >White's? Of course Black is in zugzwang. You don't need to calculate it, to >realize this. > >[D]8/5ppp/8/1k6/1p1K4/1P4P1/5P1P/8 b - - 0 41 > >This is something that remains untouchable by computers, intuition. When >programs can emulate that, then machines will take our last refuge in supremacy >at chess. > >Daniel I did it all in my head; after looking at it it is obvious white can just shuttle Kd4-d5 until black runs out of moves. It might be an interesting extension: if(king-pawn-endgame && all_king_moves_lose material) -> extend; anthony
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.