Author: Uri Blass
Date: 22:44:54 10/02/02
Go up one level in this thread
On October 02, 2002 at 18:51:24, martin fierz wrote: >On October 02, 2002 at 18:06:36, Joachim Rang wrote: > >>On October 01, 2002 at 21:41:34, John Merlino wrote: >> >>>And, how do they evaluate this positon that resulted from it? Larry played f3, >>>which certainly appeared to be good. Susan Polgar said that she thought Qh1+ >>>would lead to a won endgame. >>> >>>[D]8/p1p5/Pp1p2nk/1P1B3q/2P1Ppr1/1Q4p1/R5P1/3R1K2 b - - 0 57 >>> >>>And, finally, if Larry had played Qh1+, this position would have resulted. What >>>say your programs now? Do they play Qf3+, Ne5+ or Qh3? Any input from strong >>>humans would also be greatly appreciated.... >>> >>>[D]8/p1p5/Pp1p2nk/1P1B4/2P1Ppr1/1Q1K2p1/R5q1/3R4 b - - 0 59 >>> >> >> >>Great position! >> >>I let different programs play from this position and CT14 showed the winning >>line in a match against Fritz 7.008 >> >>I set a depth of 14 for each move, which takes on my duron 735 Mhz in the >>beginning about 30 min/move. >> >>Here is the game (from the position above): >> >> >> >>Fritz 7 - Chess Tiger 14.0 [A28] >>cm-christiansen-endgametest, 14ply/14pl JOACHIM (2), 02.10.2002 >> >>W=14.7 ply; 359kN/s >>B=14.7 ply; 220kN/s >>... 59... Qh3 60.Rg1 g2+ 61.Kc2 Rg3 62.Qb4 Ne5 63.c5 bxc5 -1.40/14 46:35 >>64.Qd2 -0.44/14 11:13 Kh7 -0.84/14 12:14 65.Kb1 0.00/14 13:57 f3 (Nf3) >>-1.60/14 13:52 66.Qd1 (Qf4) 0.00/14 10:10 66...Nd3 -1.82/14 7:24 67.b6 >>(Rd2) 0.00/14 6:01 67...cxb6 -2.30/14 3:53 68.Bg8+ -0.62/14 4:15 Kg7 >>(Kh8) -2.72/14 1:45 69.Bc4 -0.66/14 2:44 Ne5 (f2) -2.86/14 6:17 70.Bd5 >>(Bb5) -0.69/14 13:57 70...Qh6 (Qh2) -3.26/14 2:29 71.Qe1 -1.09/14 12:35 >>Qf4 -3.04/14 2:16 72.Rf2 -1.37/14 7:10 Kg6 -3.16/14 6:51 73.Qc3 (Qd1) >>-1.56/14 11:39 73...b5 -3.66/14 5:23 74.Ka2 (Kc2) -1.97/14 34:07 74...b4 >> -3.64/14 2:25 75.Qb3 (Qc2) -2.06/14 11:59 75...Ng4 (Kh5) -4.62/14 3:54 >>76.Qc2 -2.47/14 33:45 Nxf2 -4.65/14 7:47 77.Qxf2 adjud. -2.62/14 5:19 >>0-1 >> >> >> >> >> >>60 Kc2 is another possible line, but every engine I tried changes to Rg1 after >>few minutes. Tonight I will run another match with the move 60.Kc2 >> >>Do you see better defence for white, than Fritz 7.008 found? >> >> >>60.Rg1 g2+ 61.Kc2 (forced) Rg3. 62. Qb4 (or Qb2) Ne5 >> >>in this position most engines after a few minutes begin to realize, that black >>has an advantage. Most engines switch from the more logical Qd2 to c5 after few >>minutes. > >59. ...Qh3 certainly has it's merits although i still think that 59....Qf3+ is >the easier way to win (specially for a human!), leaving white no play at all. > >after 59. ...Qh3 60. Kc2 was my main line when i analysed this yesterday, and my >conclusion was that after 60... Qh2+ 61. Kb1 g2 62. Qf3 Kg5 63. Qd3 Ne5 64. Qd4 >g1=Q 65. Rxh2 Qxh2 black is winning. as a human, i would prefer the queen >exchange with 59...Qf3+, because "nothing can happen to black any more". >all in all, this position makes a great test position. is there any engine which >can see that black is winning (and i don't mean a score of +0.5...)? I think that it is easy to do it by increasing the score for passed pawns but I do not think that it is a good test position because I suspect that the engines who see it see it for the wrong reason. 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.