Author: Tony Werten
Date: 14:37:09 07/07/01
Go up one level in this thread
On July 07, 2001 at 15:51:46, Peter McKenzie wrote: >[D]2r2rk1/1p2bppp/1qn1p3/p2pPn2/1P1P1P1P/P2Q1NP1/1B2N3/2R1K2R w K - > >This position was reached in a game on ICC, with LambChop black and a human >white. LambChop just played a5, which allows the opponent to attack on the king >side with g4 Nh6 Ng5 g6 h5 and white is winning I think. > >I was a little surprised that it takes Chop 11ply to play g4 (initially it >preferred b5), and even after a 13ply search the score is still slightly >negative (0.2 in black's favour). > >So I'm wondering how other programs do here? XiniX (AMD333, 32Mb hash) gives about the same result. Biggest difference is that is suggests black recaptures on b4 with the bishop. ply: 8/19 score:-0,700 time:33s 47Kn/s 1.b5 Na7 2.a4 Rxc1+ 3.Bxc1 Rc8 4.O-O Rc4 5.Qb3 Nc6 ply: 9/22 score:-0,564 time:110s 47Kn/s 1.b5 Na7 2.a4 Rxc1+ 3.Bxc1 Rc8 4.O-O Rc4 5.Qb3 Nc6 6.Be3 ply:10/22 score:-0,606 time:181s 48Kn/s 1.b5 Na7 2.a4 Rxc1+ 3.Bxc1 Rc8 4.O-O Rc4 5.Qb3 Rb4 6.Qc2 Rc4 ply:11/25 score:-0,487 time:548s 49Kn/s 1.b5 Na7 2.a4 Rxc1+ 3.Bxc1 Rc8 4.O-O Rc4 5.Qb3 g6 ply:11/26 score:-0,387 g4 fail high ply:11/27 score:-0,232 time:1019s 48Kn/s 1.g4 Bxb4+ 2.axb4 Nxb4 3.Qb1 Ne3 4.Bc3 Nxg4 5.Bxb4 I guess white king safety together with the amount of pieces still on the board gives the problems. Tony > >I think that black has some chances to muddy the waters with sacrifices on b4 at >various points. While I don't think these are any good, they do have some merit >and could push the horizon out a bit. > >Any suggestions of what sort of evaluation terms or extensions might help here >would be welcome :-) > >cheers, >Peter > >p.s. Now I see that there is another possible defence: g4 axb4 gxf5 bxa3, not >sure if this works but it's interesting. The position after g4: > >[D]2r2rk1/1p2bppp/1qn1p3/p2pPn2/1P1P1PPP/P2Q1N2/1B2N3/2R1K2R b K - > >In this position, Chop needs 12ply to switch to axb4, score is +0.44. I >wouldn't treat that score as gospel, but I think axb4 is a much better try than >Nh6.
This page took 0.01 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.