Author: Peter Fendrich
Date: 13:26:38 05/29/01
Go up one level in this thread
On May 29, 2001 at 15:42:45, Andrew Williams wrote:
>On May 29, 2001 at 15:04:54, Peter Fendrich wrote:
>
>>On May 29, 2001 at 14:42:43, Miguel A. Ballicora wrote:
>>
>>>On May 29, 2001 at 14:21:13, Peter Fendrich wrote:
>>>
>>>>This position appeared in an ICC game: Guitje - TerraPi
>>>>
>>>>[D]2r3kb/5q1p/4pPp1/1p1pP3/p2P2Q1/PrP5/1P3R1R/2B3K1 w - - 0 42
>>>>
>>>>the last move from Terra was 41... b5
>>>>Some moves earlier when white made g6, Terra decided to move the bishop from g7
>>>>to h8.
>>>>Both Rb3 and Bh8 are locked in and Terra is absolutely unaware of the
>>>>hopeless situation. Yes, Terra lost...
>>>>
>>>>I see two different but connected problems for computer programs here:
>>>> - the closed position itself
>>>> - the two blocked pieces.
>>>>
>>>>How do you solve/avoid positions like this? Terra has about equal eval for this
>>>>position.
>>>>It is of course possible to punish pawn chains like this in order to avoid
>>>>closed positions but is that really a solution and how to deal with blocked
>>>>positions?
>>>>
>>>>//Peter
>>>
>>>Mobility helps a little in this respect. My program Gaviota has an small
>>>evaluation where the mobility component is a big % of it. I gives in this
>>>position +1.75 for white.
>>
>>I try to minimise the mobility factor and find other terms in order to achieve
>>about the same effect. In many positions a heavy weight on mobility will give
>>strange play for Terra. But herre it would help.
>>One problem here is the K-positions. Terra gives a lot of punishment for the
>>white K position that doesn't have any pawn shelter. In an open position that
>>would be ok but here it is not.
>>//Peter
>
>You're right. This is a nasty position.
>
>PM scores it statically as +0.21 for Black, which is obviously not helpful.
>It has the same problem as Terra, insofar as it thinks that White's King is in
>more danger than Black's. Part of the problem is that it's difficult for PM's
>dynamic king safety factors to kick in. All it can see is that White's King
>has less pawn shelter than Black's. How many plies from this position does it
>take for Terra to see that Black is in serious trouble? Mine sees White at
>+1.5 at depth nine, but that takes 2 seconds even on my Athlon 1200.
>
>Andrew
Terra needs depth=11 to really find that. Between 9 and 11 he is slowly changing
his mind. On my PII 300 MHz it takes more than 2 minutes to get it:
Depth Time Eval Nodes
----- ---- ----- -------
11 2:24 248 8730379 Rxh7 Kxh7 Rh2 Kg8 Qh3 Bxf6 exf6 Kf8 Qh8 Qg8 f7
Kxf7 Rh7 Qxh7 Qxh7 Kf6
//Peter
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.