Author: Robert Hyatt
Date: 06:46:13 09/08/01
Go up one level in this thread
On September 08, 2001 at 01:01:03, Peter Kappler wrote:
>On September 07, 2001 at 23:14:33, Robert Hyatt wrote:
>
>>On September 07, 2001 at 17:19:07, José Carlos wrote:
>>
>>> Last night, in one of my test games, Averno played ...Bxf4 in this position
>>>(against another program):
>>>
>>>[D]4R3/p2r3r/1pNbRp2/2p2k1p/2P2PpP/PP4P1/K7/8 b - - 0 1
>>>
>>> The game ended soon:
>>>
>>>57. gxf4 g3 58. Re2 Rhg7 59. Re1 g2 60. Rg1 Rd2+ 61. Ka1 Rf2
>>>62. Ree1 Kxf4 63. a4 Kg3 64. a5 Rc7 65. Nd8 bxa5 66. Kb1 Kxh4 67. Ka1 Kg3
>>>68. Kb1 h4 69. Ne6 Rc6 70. Nd8 Rb6 71. Ne6 Rxb3+
>>>{White resigns} 0-1
>>>
>>> 59. Re1 is a blunder, but I wonder if the sac is correct at all. It seems that
>>>59. Rg2 is good enough for white (Averno plays Bxf4 by eval, it doesn't see the
>>>pawn queening at all in search). Is my passed pawn eval too high here? What do
>>>other programs think?
>>>
>>> Thanks in advance.
>>>
>>> José C.
>>
>>
>>My first thought is that it is a bad idea. This kind of sac is not really safe
>>unless you can see the king and pawn advancing pretty quickly. Otherwise the
>>extra piece your opponent has is going to be used to win pawns and the game.
>>
>
>I think this is a case where the king and pawns advance pretty quickly. White's
>extra knight looks pretty useless.
>
>
>>Crafty won't play it, but it refuses because it doesn't like the idea of
>>giving its opponent an extra piece to play with here...
>
>How long did you let Crafty search?
>
>It might take a *long* time, especially since some of the lines involve another
>exchange sac by Black. Here's a nice example:
>
>1 ... Bxf4!
>2. gxf4 g3
>3. Re1/Re2 Kg4
>4. Ne7 Rxe7!
>5. Rxe7 Rxe7
>6. Rxe7 Kxh4 -+
>
>This isn't White's only try, but I fiddled around in Chess Tiger for a few
>minutes and couldn't find a decent defense.
>
>-Peter
I let it search a few minutes. I based my analysis on (a) my own feeling about
tossing a piece when there are still 4 rooks left on the board, which is
generally enough to either stop the passer or force a repetition; (b) the
programs that would play this were doing so with a near-zero evaluation, which
(IMHO) is very suspect.
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.