Author: Uri Blass
Date: 08:00:12 10/06/01
Go up one level in this thread
On October 06, 2001 at 10:11:41, Dieter Buerssner wrote: >On October 05, 2001 at 15:53:32, Uri Blass wrote: > >[Rearanged and snipped] > >>Zarkov4.5T - Yace >>[D]1nr1rbk1/qp3ppp/4b3/pQ1pPpB1/N2P4/P1R2N2/1P3PPP/4R1K1 b - - 0 1 > >>I think that yace blundered by 20...Bd7 > >I thought about this as well. Here in few moves the score dropped by about 1.5 >pawns in few moves (pawn=0.8 in Yace). However, it seems that after 21. Qxd5, >21...h6 was even worse. When I analyse the game from back to front (for about >one minute each move), after 21...h6 I get (German notation, S->N, L->B, T->R, >D->Q): > > 11.01 1:01 +1.41 22.Txc8 Lxc8 23.Ld2 Ld7 24.Db3 b5 25.Sc3 b4 26.axb4 axb4 >27.Sd5 Sc6 28.Dc4 Tc8 29.Le3 (9.191.711) 148.7 > >After 21...Bc6 I get similar scores, than you showed for Fritz. > >>Here is Deep Fritz's analysis >>Deep Fritz needs some hours on p800 to avoid 20...Bd7 > >Then Yace is at least in good company (also after reading Ernst's post) ... >After 21...h6 I get the following: > > 9.01 0:15 +0.44 21.Ld2 Txc3 22.Sxc3 Td8 23.Tc1 Sc6 24.Se2 Tc8 25.b4 axb4 >(1.892.588) 122.6 > 9.03 0:20 +0.45++ 21.Txc8 Txc8 22.Ld2 Ld7 23.Db6 Sc6 24.Dxa7 Lxa3 >(2.498.144) 124.7 > 9.03 0:21 +0.54 21.Txc8 Txc8 22.Ld2 Sc6 23.Sb6 Td8 24.Tc1 Le7 25.b4 a4 >(2.713.248) 124.0 > 9.54 0:27 +0.54 21.Txc8 Txc8 22.Ld2 Sc6 23.Sb6 Td8 24.Tc1 Le7 25.b4 a4 >(3.502.052) 125.2 > 10.01 0:40 +0.67 21.Txc8 Txc8 22.Sb6 hxg5 23.Sxc8 Lxc8 24.Sxg5 b6 25.Dxd5 >Dd7 26.Dc4 Lb7 (5.003.418) 125.0 > 10.54 0:54 +0.67 21.Txc8 Txc8 22.Sb6 hxg5 23.Sxc8 Lxc8 24.Sxg5 b6 25.Dxd5 >Dd7 26.Dc4 Lb7 (6.898.065) 126.8 > 11.01 1:07 +0.67 21.Txc8 Txc8 22.Sb6 hxg5 23.Sxc8 Lxc8 24.Sxg5 b6 25.Dxd5 >De7 26.Dc4 Lb7 27.Sf3 Lxf3 (8.703.943) 128.1 > 11.54 1:41 +0.67 21.Txc8 Txc8 22.Sb6 hxg5 23.Sxc8 Lxc8 24.Sxg5 b6 25.Dxd5 >De7 26.Dc4 Lb7 27.Sf3 Lxf3 (13.563.372) 133.0 > 12.01 3:29 +0.52 21.Txc8 Txc8 22.Ld2 Ld7 23.Db6 Da6 24.Dxa6 bxa6 25.Sb6 Tc2 >26.Tb1 Le6 27.Lxa5 Sc6 (27.052.046) 129.0 > 12.02 4:43 +0.53++ 21.Ld2 Txc3 22.Lxc3 Sc6 23.Sb6 Da6 24.Dxa6 bxa6 25.Tc1 >Tb8 26.Sa4 Le7 27.Sc5 f4 (35.599.897) 125.3 > 12.02 5:35 +0.54 21.Ld2 Txc3 22.Lxc3 Sc6 23.Sb6 Da6 24.Dxa6 bxa6 25.Tc1 Tb8 >26.Sa4 Tb5 (41.533.410) 123.9 > 12.54 8:23 +0.54 21.Ld2 Txc3 22.Lxc3 Sc6 23.Sb6 Da6 24.Dxa6 bxa6 25.Tc1 Tb8 >26.Sa4 Tb5 (63.050.673) 125.1 > 13.01 11:59 +0.50 21.Ld2 Txc3 22.Lxc3 Sc6 23.Sb6 Td8 24.Tc1 Le7 25.b3 f6 >26.exf6 Lxf6 27.Te1 Lf7 28.a4 (89.295.123) 124.0 > 13.02 13:10 +0.51++ 21.Txc8 Txc8 22.Ld2 Ld7 23.Db6 Da6 24.Dxa6 bxa6 25.Sb6 >Tc2 26.Tb1 Le6 27.Lxa5 Ld6 (98.011.398) 124.0 > 13.02 16:16 +0.52 21.Txc8 Txc8 22.Ld2 Ld7 23.Db6 (120.829.623) 123.7 > 13.54 25:12 +0.52 21.Txc8 Txc8 22.Ld2 Ld7 23.Db6 (189.580.066) 125.3 > 14.01 35:33 +0.52 21.Txc8 Txc8 22.Ld2 Sc6 23.Sb6 Tb8 24.Tc1 g5 25.Lc3 Da6 >26.Db3 Sb4 27.Sa4 Sd3 (266.284.994) 124.7 > 14.54 56:31 +0.52 21.Txc8 Txc8 22.Ld2 Sc6 23.Sb6 Tb8 24.Tc1 g5 25.Lc3 Da6 >26.Db3 Sb4 27.Sa4 Sd3 (430.776.766) 127.0 > 15.01 82:01 +0.51 21.Txc8 Txc8 22.Ld2 Sc6 23.Sb6 Tb8 24.Tc1 g5 25.Sxd5 Lxd5 >26.Dxd5 g4 27.Sh4 Td8 28.Db5 Txd4 29.Le3 (613.901.494) 124.7 > 15.54 148:18 +0.51 21.Txc8 Txc8 22.Ld2 Sc6 23.Sb6 Tb8 24.Tc1 g5 25.Sxd5 Lxd5 >26.Dxd5 g4 27.Sh4 Td8 28.Db5 Txd4 29.Le3 (1.121.208.459) 125.9 > 16.01 226:42 +0.39 21.Txc8 Txc8 22.Ld2 Sc6 23.Sb6 Tb8 24.Tc1 g5 25.Lc3 Da6 >26.Db3 Sb4 27.Sa4 Sd3 (1.658.039.763) 121.8 > 16.02 294:38 +0.40++ 21.Ld2 Txc3 22.Sxc3 Td8 23.Tc1 Da6 24.Dxa6 bxa6 >(2.091.868.565) 118.3 > 16.02 362:40 +0.41 21.Ld2 Txc3 22.Sxc3 Td8 23.Tc1 Sc6 24.Se2 f6 25.exf6 >(2.535.344.699) 116.5 > 16.42 525:30 +0.41 21.Ld2 Txc3 22.Sxc3 Td8 23.Tc1 Sc6 24.Se2 f6 25.exf6 >(3.601.191.337) 114.2 > >It seem not easy. From the programmer point of view, I cannot think of much what >to do here. Perhaps, like I user once suggested "you must reach more depth". >Well, if that's all :-) > >>I suspect that only material evaluation can find that 20...Bd7 is losing a pawn >>when 20...h6 and it may be interesting to see if I am right. > >I cannot understand. You mean that h6 does not lose material? I tried for 10 >minutes, and there could be seen no loss of material for both moves. > >Regards, >Dieter Did you try only material evaluation and saw 0.00 for both moves. If this is the case then I was wrong with my suspect. I suspected that Bd7 lose material from only material evaluation and that yace saw positional compensation for the pawn so it chose Bd7 but I have no program with only material evaluation near me to verify if it is truth. Bd7 let white to capture the pawn at d5. I also saw that in part of the main lines of Deep Fritz white was a pawn up from only material evaluation and this was the reason that I suspected that Bd7 is losing material from only material evaluation point of view. 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.