Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: an endgame test position

Author: Uri Blass

Date: 07:09:35 01/13/03

Go up one level in this thread


On January 13, 2003 at 09:02:40, Richard Pijl wrote:

>On January 13, 2003 at 07:23:39, Uri Blass wrote:
>
>>
>>40 minutes/40 moves
>>
>>Movei played Ra2+ and even was able even to lose the game
>>
>>[D]8/2p5/1b6/1P1K1B2/1P1p4/R1r3P1/3k3P/8 w - - 0 103 bm Rxc3
>>
>>
>
>Baron doesn't play Rxc3 either because of the odd bishop endgame that results
>from it. Instead, Baron plays Ra1 and starts pushing the g-pawn. Analysis on a
>PII-500:
>
> ply       time     nodes  score  pv
>
>  4( 9)&      0      4635    136  a3a2 d2e1 a2a1 e1d2 g3g4 d4d3 h2h4
>  4( 9).      0      5441    136  a3a2 d2e1 a2a1 e1d2 g3g4 d4d3 h2h4
>  5(11)&      0     14342    106  a3a2 d2e3 h2h4 d4d3 g3g4 d3d2 a2a1
>  5(11).      0     16806    106  a3a2 d2e3 h2h4 d4d3 g3g4 d3d2 a2a1
>  6(13)&      0     23703     81  a3a2 d2e3 h2h4 d4d3 h4h5 d3d2 f5g4 b6d4
>  6(13)&      0     44264     82  a3a1 d4d3 g3g4 d2e2 a1a2 d3d2 h2h4
>  6(13).      0     44264     82  a3a1 d4d3 g3g4 d2e2 a1a2 d3d2 h2h4
>  7(15)&      0     76384     75  a3a1 d4d3 g3g4 c3b3 a1a2 d2c3 h2h4 b6d4 g4g5
>  7(15)&      0    108476     87  a3a2 d2e3 a2a1 d4d3 a1d1 c3a3 h2h4 b6d4 d1e1
>e3f2
>  7(15).      0    116894     87  a3a2 d2e3 a2a1 d4d3 a1d1 c3a3 h2h4 b6d4 d1e1
>e3f2
>  8(18)&      1    205344     78  a3a2 d2e3 a2a1 d4d3 a1e1 e3d2 e1e8 c3b3 e8e4
>b3c3 f5e6
>  8(18).      2    322922     78  a3a2 d2e3 a2a1 d4d3 a1e1 e3d2 e1e8 c3b3 h2h4
>b3b4 e8g8
>  9(19)&      4    581286     74  a3a2 d2e3 a2a1 d4d3 a1d1 d3d2 h2h3 c3b3 d5c4
>b3a3 h3h4 e3e2 f5g4 e2e3
>  9(19).      7    850025     74  a3a2 d2e3 a2a1 d4d3 a1d1 d3d2 h2h3 c3b3 d5c4
>b3a3 h3h4 e3e2 f5g4 e2e3
> 10(20)&     11   1353120     61  a3a2 d2e3 a2a1 d4d3 a1e1 e3d2 e1e8 c3b3 f5g4
> 10(21).     26   3128964     61  a3a2 d2e3 a2a1 d4d3 a1e1 e3d2 e1e8 c3c1 e8e4
>c1h1 h2h4 b6f2 e4e2 d2e2 f5e4
> 11(22)&     38   4690962     28  a3a2 d2e3 h2h4 d4d3 a2a1 d3d2 a1d1 e3e2 f5g4
>e2d3 d1a1 b6f2 g4d1
> 11(22)&     99  11661195     39  a3a1 d4d3 h2h4 d2e2 h4h5 b6e3 f5g4 e2f2 g4d1
>d3d2 g3g4 c3c1 a1c1
> 11(22).     99  11740636     39  a3a1 d4d3 h2h4 d2e2 h4h5 b6e3 f5g4 e2f2 g4d1
>d3d2 g3g4 c3c1 a1c1
> 12(25)&    238  27410320     47  a3a1 d4d3 g3g4 d2e3 g4g5 d3d2 a1d1 c3c1 f5g4
>e3f4 g5g6 f4g4 g6g7 c7c6 b5c6 c1d1
> 12(25).    287  33360222     47  a3a1 d4d3 g3g4 d2e3 g4g5 d3d2 a1d1 c3c1 f5g4
>e3f4 g5g6 f4g4 g6g7 c7c6 b5c6 c1d1
> 13(25)&    621  70177602     62  a3a1 d4d3 g3g4 d2e3 g4g5 d3d2 a1d1 c3c1 f5g4
>e3f4 g5g6 f4g4 g6g7 c1d1
> 13(25).    705  79851113     62  a3a1 d4d3 g3g4 d2e3 g4g5 d3d2 a1d1 c3c1 f5g4
>e3f4 g5g6 f4g4 g6g7 c1d1
> 14(27)&   1733 188194665     61  a3a1 d4d3 g3g4 d2e3 g4g5 d3d2 a1d1 c3c1 f5g4
>e3f4 g5g6 f4g4 g6g7 c1d1 g7g8q g4f3 g8g3
> 14(27).   2000 218842963     61  a3a1 d4d3 g3g4 d2e3 g4g5 d3d2 a1d1 c3c1 f5g4
>e3f4 g5g6 f4g4 g6g7 c1d1 g7g8q g4f3 g8g3
>
>Richard.

Interesting line but I think that the final position is a draw inspite of the
fact that moveiXX say +2.10 for white

depth=10 +2.17 g8f7 f3e2 f7h5 e2e1 h5e5 e1f2 e5f6 f2e2 f6e6 e2f2 e6f5 f2e2 f5g4
e2e1 g4e4 e1f2 e4f4 f2e1 f4g3 e1e2 g3g2 e2e1 g2h1 e1e2 h1e4 b6e3
Nodes: 81636908 NPS: 198741
Time: 00:06:50.77

depth=11 +2.10 g8f7 f3e2 f7h5 e2e1 h5e5 e1f1 e5f6 f1e2 f6e6 e2f2 e6f5 f2e2 f5e4
e2f2 e4d3 f2e1 d3g3 e1e2 g3g4 e2e1 g4f3
Nodes: 140416222 NPS: 196556
Time: 00:11:54.38

Movei also does not like to push the g pawn in the initial position

The pawn at g3 is protected passed pawn so the bonus for protected passed pawn
encourage it not to push the pawn forward unless it can also push the h pawn to
protect it.

piece square table encourages it to push g forward but not enough.
I have ideas how to fix it and I may do it.

Today it is not clear if the bonus for protected passed pawn is productive or
counter productive.

Here is movei's opinion of the root position after some minutes

depth=12 +1.53 a3a2 d2e3 a2a1 c3b3 d5c6 b3b4 a1a3 e3f2 f5d3 b4b2 d3e4 f2g1 h2h4
g1f2
Nodes: 32980545 NPS: 209933
Time: 00:02:37.10
depth=13 +1.51 a3a2 d2e3 a2a1 c3b3 d5c6 d4d3 a1e1 e3f2 e1d1 f2e3 c6d5 d3d2 d5c4
b3b2
Nodes: 49517807 NPS: 210785
Time: 00:03:54.92
depth=13 +1.51 a3a2 d2e3 a2a1 c3b3 d5c6 d4d3 a1e1 e3f2 e1d1 f2e3 c6d5 d3d2 d5c4
b3b2
Nodes: 74123956 NPS: 212401
Time: 00:05:48.98
depth=14 +1.47 a3a2 d2e3 a2a1 c3b3 a1e1 e3f2 e1e8 b3b4 f5d3 b4b3 d3c4 b3b2 d5e4
f2g1 h2h4 g1f2
Nodes: 125787051 NPS: 209505
Time: 00:10:00.40

After Ra1 movei does not expect d3 and prefers Rb3

depth=12 -1.45 c3b3 a1b1 b3b1 f5b1 d2c3 d5e4 c3b4 b1d3 b4c3 d3e2 b6c5 e2d3
Nodes: 27649110 NPS: 213523
Time: 00:02:09.49
depth=13 -1.36 c3b3 a1b1 d2c3 b1b3 c3b3 h2h4 b3b4 f5d3 b4c3 d5e4 b6c5 g3g4 c5d6
h4h5
Nodes: 71847546 NPS: 213381
Time: 00:05:36.71



2 notes about bugs and mistakes:

1)Note that I know that Movei has a bug in
the pv and it may report wrong pv in the qsearch(it may sometimes say mate in 2
but gives a pv that is more than 3 plies)

I know about the problem but Movei never is wrong in the moves not in the
qsearch that include the first move so I did not consider this bug as very
important to fix and the only negative effect of it in games may be wrong order
of moves in the last ply when movei follows the pv.


2)I see that I made an error when I reported 40/40 and it was 40 minutes/60
moves(I thought to play 40/40 but set the time control wrong)

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.