Author: José Carlos
Date: 06:46:12 12/12/00
Go up one level in this thread
On December 12, 2000 at 02:24:42, Uri Blass wrote:
>On December 11, 2000 at 20:27:43, Christophe Theron wrote:
>
>>On December 11, 2000 at 14:54:19, Mogens Larsen wrote:
>>
>>>On December 11, 2000 at 13:56:17, Christophe Theron wrote:
>>>
>>>>That's not true. Programs DO plan.
>>>
>>>No, they don't. Not by any definition of the word.
>>>
>>>Programs behave like a donkey following a carrot on a stick. Wherever the
>>>programmer points the stick the program follows. Changing the size of the carrot
>>>or the length of the stick doesn't matter.
>>
>>
>>
>>Excuse me Mogens, but I believe I have some knowledge about what's going on in a
>>chess program (mine for example), and I can assure you that programs DO plan.
>
>
>The question is what is your definition of planning.
>
>
>I can give you example of what I mean by planning when I have the impression
>that programs do not plan.
>
>[D]8/5k2/6p1/3K1p1p/1R3P1P/6P1/3b4/8 w - - 0 1
>
>I believe that Gandalf drew it with white against Fritz because it was not
>planning.
>
>I thought about the position and found the following idea
>
>White has to make progress in order to win.
>In order to make progress white has to push the king forward or to play g4 at
>some point.
>My plan is to push the king forward and to play Ke5,Kf6 because g4 seems not to
>help much.
>
>I cannot play these moves if the black bishop catch the long diagnol(a1-h8)
>so in order to win the game white has to prevent black to catch this diagnol.
>
>Here is the relevant analysis with Junior that suggest that my plan probably
>gives a win for white.
>
>83. Rb7+ Kf6 (
>83... Kf8 84. Rc7 Ba5 85. Rc5 Bb4 {2.21/21 173} 86. Rc4 Be1 {2.80/25 802} 87.
>Ke6 Bxg3 {2.80/22 24} 88. Kf6 Bxh4+ {2.80/21 16} 89. Kxg6 Bf2 {2.95/21 112} 90.
>Kxh5 Kf7 {3.20/22 91} 91. Kg5) 84. Rb6+ Kf7 85. Rc6 Be1 {1.91/24 309} 86. Ke5
>Kg7 {1.91/15 4} 87. Rc7+ Kh6 {1.92/18 15} 88. Kf6 Bxg3 {2.03/19 42} 89. Rg7
>Bxf4 {2.13/21 58} 90. Rxg6+ Kh7 {2.13/3 4} 91. Kxf5 Bc7 {2.75/23 82} 92. Rg5
>Bd8 {2.75/22 38} 93. Rxh5+ Kg7 {2.80/23 27} 94. Kg4
>
>
>White sacrifices pawns if you do not search deep enough but it helps it to make
>progress.
>
>The problem is not to find 83.Rb7+(Rc4 may be also winning and even the game
>move(Rb1) may win if you try to go back to the correct plan to get the bishop
>out of the diagnol when black may prevent it by the plan to put the king at
>f8,f7,g7 but in this case Ke6 with g4 may be good for white) but to do the plan
>to get the king forward.
>
>Uri
I agree with your definition of planning. So I disagree with Christophe's. I'd
like to see a similar example from him, explaning his point when he says
programs DO plan.
José C.
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.