Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: A case when Movei earn 2 plies from verified null move pruning

Author: Sune Fischer

Date: 03:15:27 11/23/02

Go up one level in this thread


On November 23, 2002 at 06:03:50, Uri Blass wrote:

>On November 23, 2002 at 05:52:51, Sune Fischer wrote:
>
>>On November 23, 2002 at 05:48:25, Uri Blass wrote:
>>
>>>On November 23, 2002 at 05:45:19, Sune Fischer wrote:
>>>
>>>>On November 23, 2002 at 05:02:40, Uri Blass wrote:
>>>>
>>>>>The repetition is tactics but the fact that you do not see a positive score for
>>>>>other moves is evaluation.
>>>>>
>>>>>Movei see positive score for other moves.
>>>>
>>>>Other moves, you mean c3?
>>>>I guess c3 is only considered at low depth where it can't see the tactics.
>>>>
>>>>-S.
>>>
>>>I mean that movei gives positive score for moves like 1...Rb8 or 1...Rb7 so it
>>>is not going to be happy with 1...Rxb2 in case of a draw score.
>>
>>I don't see why 1..Rb8 or Rb7 should give a positive score, if you don't intend
>>Rxb2! then I see no way forward for black. What is the PV?
>
>Here is the pv of movei without verifyied null move pruning before finding Rxb2
>
>depth=14 +0.59 b3b8 e3e4 f5e4 f3e4 b8e8 e4d4 e8e2 d4c3 e2d2 c3d2 h7h6 h2h4 h6h5
>d2e3 f6e6 e3d2 e6d5
>Nodes: 12019985 NPS: 188341
>Time: 00:01:03.82
>
>depth=14 +0.61 b3b6 h2h4 h7h5 d2g2 b6b7 g2d2 b7g7 e3e4 g7g1 e4e5 f6e6 d2h2 e6d5
>f3e3 g1g3 e3d2
>Nodes: 22524141 NPS: 184881
>Time: 00:02:01.83

I gave the position after Rb6, it can only see a draw now.
Remember black can still play Rxb2, so white has to "do something about it".
So I think h4 looks a bit odd.

7	28	11	22671		1.e4 fxe4+ 2.Kxe4 Re6+ 3.Kf3 Kf5 4.Rg2 Rg6 5.Rg3
8	26	21	56608		1.e4 fxe4+ 2.Kxe4 Re6+ 3.Kf3 Kf5 4.Rg2 Rg6 5.Rf2 Rg1
9	0	47	146578		1.e4 fxe4+ 2.Kxe4 Re6+ 3.Kf3 Kf5 4.Rg2 Re2 5.Rg5+ Kf6 6.Rg2 Kf5
9	0	120	416379		1.h3 Re6 2.Rg2 Rd6 3.Rd2 Rh6
10	0	222	800215		1.h3 Rxb2 2.Rxb2 c3 3.Rb6+ Ke7 4.Rb7+ Kd6 5.Rb6+ Ke7
11	0	495	1832059		1.h3 Re6 2.Rg2 Rd6 3.Rd2 Re6 4.e4 Rh6
12	-4	1153	4256456		1.h3 Re6 2.Rd1 Kg6 3.Rc1 Rd6 4.Rd1 Kh5 5.Rd2 Kh4 6.Kg2 Rb6
12	-3	1547	5745396		1.Rg2
12	0	1745	6498785		1.Kf2 Rb3 2.Ke1
13	0	4217	15612657	1.Kf2 Rb3 2.Kf3 Rc6
14	0	12196	43569716	1.Kf2 Rb3 2.Kf3 Kg6 3.Kf3

>>Staticly I don't see a reason to give a high score for black, the only passed
>>pawn is blocked.
>
>protected passed pawn by a pawn is a good reason to give bonus for black even if
>it is blocked.

yes that is true, a small bonus, but not a high score for black IMO.

>The reason for movei are different
>
>h2 f4 get the same score as h7 f5 by piece square table
>a4 c4 d3 get better score than a3 e3 b2 based on piece square table
>
>The black king can get into the center so it get better score than the white
>king based pievce square table.
>
>The black rook has better mobility than the white rook so black get a bonus for
>better mobility.

Well black has terrain, but it's hard to say if the pawns are weakened or
strong, afterall white's king is closer, so staticly...

-S.
>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.