Winboard Forum

Hi folks,
during the last Chess War tournament the following position was found in a match between Kiwi (black) and another engine (whose name I forgot):

[diag]3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 22 45[/diag]

FEN: 3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 22 45

Here Kiwi evaluates the position as +0.32 for black, while a few other strong engines I tried give it at least +0.53 for white... almost a full pawn delta. While I could tweak some parameters a little, there is no way I can turn this position score so much without some major changes. The problem is, I'm not a good player and I don't understand the position at all... can you tip me on what is contributing to a strong positive evaluation for white?

Hi Alessandro,
Did white win that game?
I can't see a white advantage looking at the position (I didn't try my engine because I can't right now). Black has doubled and isolated pawns, which might get a big penalty in some programs. Both white knights have safe (from enemy pawns) position (static bonus), but black has also a good knight, and can offer the other in exachage.
Without analizying (just a quick look), black looks perfectly fine.

I forgot: other static penalty for black is many pawns in black squares plus bishop in black squares.
But I stand that black is ok for a passive defense and draw.

Alessandro Scotti wrote:Hi folks,
during the last Chess War tournament the following position was found in a match between Kiwi (black) and another engine (whose name I forgot):

[diag]3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 22 45[/diag]

Here Kiwi evaluates the position as +0.32 for black, while a few other strong engines I tried give it at least +0.53 for white... almost a full pawn delta. While I could tweak some parameters a little, there is no way I can turn this position score so much without some major changes. The problem is, I'm not a good player and I don't understand the position at all... can you tip me on what is contributing to a strong positive evaluation for white?

You do not need to be a strong player to have positive evaluation for white.

I do not understand why kiwi evaluates it as advantage for black.
I see some reasons for a small advantage for white:

1)White pawns are more advanced than black's pawns
2)black has a double pawn
3)black has no pawn in the 7th rank near the king when white has pawn at h2.

Movei evaluates it as less than 0.53 for white but it is still advantage for white.

Uri

Jos? Carlos wrote:Hi Alessandro,
Did white win that game?

Hi Jose,
yes White proceeded to win the game, which I have now found and copied below. Kiwi was outsearched by at least one ply at every move but looking at the game continuation it seems the position is not very promising for black anyway... Petir evaluated the above position at +0.69 for white and it proceeded quite convincingly to show it was right!

Code: Select all

[Event "ChessWar VI D 40m/20'"]
[Site "CHESSWAR"]
[Date "2004.12.27"]
[Round "11.5"]
[White "Petir 2.0"]
[Black "Kiwi 0.3c"]
[Result "1-0"]
[WhiteElo "2078"]
[BlackElo "1841"]
[ECO "E97"]
[EventDate "2004.??.??"]
[PlyCount "186"]

1.e4 g6 2.c4 Bg7 3.d4 d6 4.Nc3 Nf6 5.Nf3 O-O 6.Be2 e5 7.O-O Nc6 8.d5
Ne7 9.b4 Nh5 10.Re1 a5 11.bxa5 Rxa5 12.Rb1 b6 13.Qc2 Bd7 14.g3 Nf6
15.Bf1 Qc8 16.Bd2 Ra7 17.Nb5 Ra8 18.Bg2 Bg4 19.Bb4 Ne8 20.Nd2 Ra6 21.
Kh1 Qd7 22.f3 Bh3 23.Bxh3 Qxh3 24.Re2 h5 25.Rg1 h4 26.g4 Bh6 27.g5
Bg7 28.Qb3 Qd7 29.a3 f6 30.Qe3 Kh7 31.Rg4 h3 32.Qg1 fxg5 33.Qg3 Nf6
34.Qxh3+ Nh5 35.Rg3 Qc8 36.Kg1 Qd8 37.Re1 Ng8 38.Kh1 Nh6 39.Rgg1 Rf4
40.Qg2 Nf7 41.Nf1 Rh4 42.Ne3 Nf4 43.Qd2 Bf6 44.Ng4 Be7 45.Qe3 Qc8 46.
Rg3 Nh5 47.Rg2 Qd7 48.Qb3 Nf4 49.Rg3 Nh6 50.Nf2 Bf6 51.a4 Nf7 52.Ng4
Bg7 53.Ra1 Ne2 54.Rg2 Nd4 55.Qd1 c6 56.dxc6 Qxc6 57.Qd3 Ra8 58.Ra3
Kg8 59.Ne3 Rh3 60.Nd5 Kh7 61.Qf1 Rh4 62.Be1 Rh6 63.Rb2 Rb8 64.Nxd4
exd4 65.a5 Ne5 66.Rxb6 Qc8 67.Rab3 Nd7 68.Rxb8 Nxb8 69.Bg3 Nc6 70.Ra3
Be5 71.Bxe5 dxe5 72.a6 Qb8 73.Qc1 Kg7 74.Ra1 Qb3 75.Qxg5 Qxf3+ 76.Qg2
Qxg2+ 77.Kxg2 Rh8 78.a7 Ra8 79.Ra6 Rxa7 80.Rxc6 Ra2+ 81.Kg3 Re2 82.
Rc7+ Kh6 83.h4 g5 84.Rc6+ Kg7 85.hxg5 Rxe4 86.Nf6 Re3+ 87.Kg4 d3 88.
g6 d2 89.Nh5+ Kh6 90.g7+ Kh7 91.Rd6 Re1 92.Rxd2 Rg1+ 93.Kf5 Kh6 1-0


Uri Blass wrote:You do not need to be a strong player to have positive evaluation for white.

I do not understand why kiwi evaluates it as advantage for black.
I see some reasons for a small advantage for white:

1)White pawns are more advanced than black's pawns
2)black has a double pawn
3)black has no pawn in the 7th rank near the king when white has pawn at h2.

Movei evaluates it as less than 0.53 for white but it is still advantage for white.

Uri

Hi Uri,
it is possible that some parameters related to pawn evaluation are too large. Here is how Kiwi evaluates the pawn structure:

Code: Select all

********** Pawn evaluation **********
Black pawn e5 piece/square: 10..10
Black pawn e5 score: 10..10
Black pawn g5 piece/square: 0..5
The black pawn g5 is doubled
The black pawn g5 is isolated and part of a column
Black pawn g5 score: -17..-25
Black pawn b6 piece/square: 0..2
Black pawn b6 score: 0..2
Black pawn d6 piece/square: 5..2
Black pawn d6 score: 5..2
Black pawn g6 piece/square: 0..2
The black pawn g6 is doubled
The black pawn g6 is isolated and part of a column
Black pawn g6 score: -17..-28
Black pawn c7 piece/square: 2..0
Black pawn c7 score: 2..0

White pawn h2 piece/square: 2..0
The white pawn h2 is isolated (isolani=0)
...and placed on a open file
White pawn h2 score: -23..-15
White pawn a3 piece/square: 0..2
The white pawn a3 is isolated (isolani=1)
...and placed on a open file
White pawn a3 score: -30..-23
White pawn f3 piece/square: 2..2
The white pawn f3 is backward and blocked by an enemy pawn!
White pawn f3 score: -18..-48
White pawn c4 piece/square: 5..5
White pawn c4 score: 5..5
White pawn e4 piece/square: 10..10
White pawn e4 score: 10..10
White pawn d5 piece/square: 10..10
White pawn d5 score: 10..10
Black pawn score: -17..-39
White pawn score: -46..-61
Pawn score: -28


It seems too much importance is given to white's a and h isolated pawns, and also possibly to the value of central pawns. (BTW the first value is used with all pieces on the board and the latter with no pieces at all, the actual value is interpolated between the two according to the current stage). So the final score is -0.28, i.e. an advantage for Black.
I know I'll have to modify some values now... any suggestions where to start?

Hi Alssandro,

I guess that g6 g5 should be evaluated as bigger weakness than a3 h2
because this pawns are not only isolated but also doubled pawns.

I also think that white pawn at d5 should be evaluated as slightly more than white pawn at e4(or black pawn at e5) because it is more advanced pawn.

Uri

Hi Alessandro,

Smirf has a completly different way of evaluation, but it is voting +0.41 for Black:

07:23.6 (12.01) +0.41 Sf7-h6
03:57.5 (11.01) +0.38 Sf7-h6
02:05.6 (10.03) +0.27 Sf7-h6
01:33.0 (10.01) +0.27 Dd8-d7 Sb5-c3 c7-c6 Lb4:d6 Sf7:d6 d5:c6 Dd7:c6 De3:f4 e5:f4 Sg4-f6+
00:23.7 (09.01) +0.18 Dd8-d7 Sb5-c3
00:09.8 (08.02) +0.21 Dd8-d7 Sb5-c3
00:08.5 (08.01) +0.06 Dd8-c8 Tg1-g3 Ta6-a4 Tg3-g1
00:04.6 (07.22) +0.25 Dd8-c8 Tg1-g3 Ta6-a4 Tg3-g1 Ta4:b4 a3:b4
00:02.7 (07.01) +0.25 Dd8-d7 Sb5-c3
00:02.4 (07.13) +0.25 Dd8-d7 Sg4-f2 Ta6-a4 Lb4:d6 c7:d6 De3:f4 e5:f4 Sb5:d6
00:01.7 (07.03) +0.03 Sf7-h6
00:01.5 (07.02) +0.02 Ta6-a4 Sb5-c3 Ta4-a8 Sg4-f2 Sf4-e2 Te1:e2
00:01.3 (07.01) +0.02 Sf4-h3 Tg1-g2 Ta6-a4 Sg4-f2 Sh3-f4 De3:f4 e5:f4 Sb5:d6
00:00.8 (06.03) +0.00 Sf4-h3 Tg1-g2 Ta6-a4 Sg4-f2
00:00.0 (05.01) +0.66 Ta6-a4 Lb4:d6 c7:d6 De3:f4 e5:f4 Sg4-f6+ Le7:f6
00:00.0 (04.01) +0.27 Sf7-h6

Regards, Reinhard.

Shredder 7.04 evaluates it as slightly better for black 0.27 and Fruit 2.0 as slightly better for white 0.35.

I think in that case Fruit is right!

edit:

upps I'm not sure, black has some potential for a king attack. In a quick shootout between Fruit and Shredder 8 black soon build up a deadly king attack:


[Event "40 Z?ge in 5 min; 40 Z?ge in 5 min; Rest in 5 min"]
[Site "Engine Match"]
[Date "2004.12.31"]
[Round "1"]
[White "Fruit 2.0"]
[Black "Shredder 8"]
[Result "0-1"]
[SetUp "1"]
[FEN "3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - -"]

1... Qd7 {4:43m} 2. Qd2 {+0.37/11 13s} Nh6 {+0.38/12 6s}
3. Ne3 {+0.29/11 16s} Ra8 {+0.44/14 17s} 4. Qc2 {+0.25/11
15s} Kg7 {+0.52/12 6s} 5. Rd1 {+0.20/10 13s} Qh3 {+0.77/12
7s} 6. Qf2 {+0.23/10 10s} Rh8 {+0.71/13 6s} 7. Rd2
{+0.27/10 6s} Qd7 {+0.74/12 4s} 8. Qf1 {+0.13/10 9s} Rh3
{+0.72/11 7s} 9. Bc3 {+0.09/10 10s} Ng8 {+0.81/12 7s}
10. Ng4 {+0.06/10 5s} Nf6 {+0.86/12 8s} 11. a4 {0.00/10 7s}
R8h4 {+0.81/11 5s} 12. a5 {-0.17/10 4s} Nxg4 {+2.31/12 7s}
13. Rxg4 {-0.90/11 4s} Rxg4 {+2.76/13 11s} 14. fxg4
{-1.02/10 1s} c6 {+3.01/14 17s} 15. dxc6 {-1.34/11 4s} Qxc6
{+3.28/11 1s} 16. Rd5 {-2.33/11 15s} Re3 {+3.54/12 4s}
17. Na7 {-2.63/11 6s} Qa4 {+4.42/12 10s} 18. Rxe5 {-2.93/10
4s} dxe5 {+7.53/12 5s} 19. Bxe5+ {-5.26/11 6s} Kg8
{+9.51/12 9s} 0-1


regards Joachim

Ich w?rde mal sagen, schwarz steht besser. Shredder 8 entfaltet jedenfalls auch gegen sich selber einen t?dlichen K?nigsangriff:

1... Ra8 {+0.30/16 1:05m} 2. Rg3 {-0.31/15 1:08m} Qd7 {+0.43/15 30s} 3. Bd2 {-0.43/14 55s} Kg7 {+0.41/15 1:30m} 4. Bc3 {-0.36/15 40s} Nh5 {+0.48/15 1:01m} 5. Rg2 {-0.60/15 55s} Rh8 {+0.60/15 1:07m} 6. a4 {-0.37/14 43s} Nf4 {+0.36/14 28s} 7. Rd2 {-0.61/15 1:21m} Rh3 {+0.68/15 33s} 8. Rg1 {-0.58/15 53s} R8h4 {+0.55/15 41s} 9. Qf2 {-0.60/16 42s} Nh6 {+0.82/16 43s} 10. Nxh6 {-1.24/16 2:38m} Rxh6 {+1.20/17 35s} 11. Qf1 {-1.20/16 21s} R6h4 {+1.20/16 59s} 12. Rc2 {-1.44/15 1:02m} g4 {+1.21/15 1:01m} 13. fxg4 {-1.22/14 30s} Rxg4 {+1.22/15 27s} 14. Rf2
{-1.22/15 50s} Kh7 {+1.30/14 26s} 15. a5 {-1.55/14 44s} Bh4
{+2.47/13 31s} 16. Ra2 {-2.47/12 30s} Rxg1+ {+5.27/13 31s}
17. Qxg1 {-6.10/15 36s} Rf3 {+6.85/16 1:02m})

Gru? Joachim

To Uri:

You do not need to be a strong player to have positive evaluation for white.

I do not understand why kiwi evaluates it as advantage for black.
I see some reasons for a small advantage for white:

1)White pawns are more advanced than black's pawns
2)black has a double pawn
3)black has no pawn in the 7th rank near the king when white has pawn at h2.

Movei evaluates it as less than 0.53 for white but it is still advantage for white.

Uri

IMHO every strong player would favour Black a bit here because it has
some attacking prospects and I see nothing for White, than defending h2.
OTH I guess I am only a bit stronger than you, but at least I would
never trust programs at all here...

Guenther

Guenther Simon wrote:To Uri:

You do not need to be a strong player to have positive evaluation for white.

I do not understand why kiwi evaluates it as advantage for black.
I see some reasons for a small advantage for white:

1)White pawns are more advanced than black's pawns
2)black has a double pawn
3)black has no pawn in the 7th rank near the king when white has pawn at h2.

Movei evaluates it as less than 0.53 for white but it is still advantage for white.

Uri

IMHO every strong player would favour Black a bit here because it has
some attacking prospects and I see nothing for White, than defending h2.
OTH I guess I am only a bit stronger than you, but at least I would
never trust programs at all here...

Guenther

I understand human reasons to prefer black but
I see nothing that can be defined to help chess programs to understand it.

I was thinking about the question if there are good reason that a chess program will define it as better for black and I found no good reason.

The only reason for advantage for black is some better control of squares near the opponent king but this reason should not be enough to compensate for all the other reasons.

I did not answer about the question if I prefer white or black but what I expect chess programs to prefer by reasons that I know how to define.

Uri

Well it seems the position is not so clear after all, so I think I will leave the evaluation as it is for the moment. I found the comments from Uri and Guenter very insightful, and seeing also the Shredder game I think maybe Kiwi lost the game not because the position was so bad in itself, but because it was not able to play it correctly.

This two games prove nothing really, but you might find some insights from them.

[Event "Blitz:10'"]
[Site "?"]
[Date "2005.01.03"]
[Round "1"]
[White "Fruit 2.0"]
[Black "Shredder 8"]
[Result "0-1"]
[TimeControl "300+0"]
[SetUp "1"]
[FEN "3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 0 45 "]
[Termination "normal"]
[PlyCount "58"]
[WhiteType "program"]
[BlackType "program"]

Qd7 { Both last book move [%emt 0:00:13]} 46.Rec1 {[%emt 0:00:19]} Nh6 { [%emt
0:00:09]} 47.Nf2 {[%emt 0:00:18]} Ra8 {(Kh7-g7) [%emt 0:00:14]} 48.Rc2 {[%emt 0:00:18]}
Rh8 {(Ra8-f8) [%emt 0:00:14]} 49.Be1 {[%emt 0:00:20]} Kg7 { (Rh8-f8) [%emt
0:00:14]} 50.Nd3 {[%emt 0:00:14]} Rh3 {[%emt 0:00:11]} 51.Bg3 {[%emt 0:00:14]} Qd8
{(Nf4xd3) [%emt 0:00:13]} 52.Nb4 {(Nb5-a7) [%emt 0:00:12] } Qd7 {[%emt 0:00:15]} 53.Nc6
{(Nb4-a6) [%emt 0:00:25]} Bf6 { (g5-g4) [%emt 0:00:14]} 54.Nc3 {[%emt 0:00:16]} Nf7
{(Nh6-g8) [%emt 0:00:05]} 55.Rf2 {(Nc3-d1) [%emt 0:00:13]} Nh5 {(Nf7-d8) [%emt 0:00:14]}
56.Ne2 { [%emt 0:00:15]} g4 {[%emt 0:00:08]} 57.Rgg2 {(f3-f4) [%emt 0:00:08]} Nxg3+ {
[%emt 0:00:12]} 58.Nxg3 {[%emt 0:00:13]} Bg5 {[%emt 0:00:09]} 59.Qd3 { [%emt
0:00:21]} Bf4 {[%emt 0:00:14]} 60.fxg4 {[%emt 0:00:09]} Ng5 { [%emt 0:00:10]} 61.Rxf4
{[%emt 0:00:16]} exf4 {[%emt 0:00:02]} 62.Qd4+ { [%emt 0:00:08]} Kg8 {[%emt 0:00:04]}
63.Nf1 {[%emt 0:00:15]} f3 { [%emt 0:00:13]} 64.Rb2 {(Rg2-a2) [%emt 0:00:12]} Qxg4
{[%emt 0:00:08]} 65.Qf6 {[%emt 0:00:10]} Nxe4 {(Kg8-h7) [%emt 0:00:16]} 66.Qd8+ {[%emt
0:00:10]} Kg7 { (Kg8-h7) [%emt 0:00:23]} 67.Qxc7+ {[%emt 0:00:08]} Kh6 {[%emt 0:00:08]}
68.Qxb6 {[%emt 0:00:09]} f2 {[%emt 0:00:20]} 69.Rxf2 {[%emt 0:00:09]} Rxa3 { (Ne4xf2+)
[%emt 0:00:18]} 70.Rg2 {(Qb6-b2) [%emt 0:00:06]} Qf3 {[%emt 0:00:07] } 71.Ne5 {[%emt
0:00:27]} Qxf1+ {(Ne4-g3+) [%emt 0:00:11]} 72.Qg1 { [%emt 0:00:07]} Qxg1+ {[%emt
0:00:03]} 73.Rxg1 {[%emt 0:00:07]} dxe5 { adjud. [%emt 0:00:08]} {} 0-1


[Event "Blitz:10'"]
[Site "?"]
[Date "2005.01.03"]
[Round "2"]
[White "Shredder 8"]
[Black "Fruit 2.0"]
[Result "1-0"]
[TimeControl "300+0"]
[SetUp "1"]
[FEN "3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 0 45 "]
[Termination "normal"]
[PlyCount "50"]
[WhiteType "program"]
[BlackType "program"]

Qd7 { Both last book move [%emt 0:00:13]} 46.Rg3 {[%emt 0:00:14]} Nh6 { (Ra6-a8)
[%emt 0:00:19]} 47.Reg1 {(Ng4xh6) [%emt 0:00:12]} Nxg4 { (Ra6-a8) [%emt 0:00:12]}
48.R3xg4 {(f3xg4) [%emt 0:00:03]} Bf6 { (Rh4-h6) [%emt 0:00:14]} 49.Rxh4+ {(Bb4-e1)
[%emt 0:00:15]} gxh4 { [%emt 0:00:05]} 50.Rg4 {(Bb4-d2) [%emt 0:00:13]} Ra8 {[%emt
0:00:14]} 51.Bd2 {[%emt 0:00:14]} g5 {[%emt 0:00:14]} 52.Qf2 {(Qe3-b3) [%emt 0:00:13]}
Ra4 { (Kh7-h8) [%emt 0:00:27]} 53.Bxf4 {(Qf2-f1) [%emt 0:00:04]} gxf4 { [%emt
0:00:09]} 54.Qf1 {[%emt 0:00:11]} Qe7 {[%emt 0:00:11]} 55.Rg1 { [%emt 0:00:15]} Bg7
{[%emt 0:00:16]} 56.Qd3 {[%emt 0:00:13]} Qd7 { (Kh7-g8) [%emt 0:00:14]} 57.Rg4
{(Qd3-c2) [%emt 0:00:21]} Qe7 {[%emt 0:00:09]} 58.Qc2 {[%emt 0:00:20]} Ra8 {[%emt
0:00:11]} 59.Rg1 {[%emt 0:00:17]} h3 { [%emt 0:00:13]} 60.Qd1 {(Qc2-e2) [%emt 0:00:11]}
Kg8 {(Ra8-h8) [%emt 0:00:15]} 61.Qf1 {[%emt 0:00:11]} Qd7 {[%emt 0:00:12]} 62.Rg5
{[%emt 0:00:08]} Ra4 { [%emt 0:00:12]} 63.Rg4 {(Rg5-h5) [%emt 0:00:07]} Qd8 {(Qd7-f7)
[%emt 0:00:11]} 64.Qxh3 {[%emt 0:00:05]} Rxc4 {[%emt 0:00:11]} 65.Rg1 {[%emt
0:00:07]} Rc2 { (Qd8-f6) [%emt 0:00:07]} 66.Na7 {(Qh3-e6+) [%emt 0:00:10]} Qf6 { (Qd8-f8)
[%emt 0:00:11]} 67.Nc6 {[%emt 0:00:13]} Rc3 {(b6-b5) [%emt 0:00:20]} 68.Qd7 {[%emt
0:00:13]} Rxc6 {(Qf6-f7) [%emt 0:00:10]} 69.dxc6 { (Qd7xc6) [%emt 0:00:10]} Qf7 {adjud.
[%emt 0:00:18]} {} 1-0

Jaime Benito de Valle wrote:This two games prove nothing really, but you might find some insights from them.

I expected this Shredder is way too strong for fruit. In fact, I think that Shredder can beat fruit even if you handicap it with two pawns. Perhaps a better test is to pit List512 against ProDeo??

Or better still, post it on talkchess.com forums.

Pallav

Yes, Shredder is just stronger than Fruit, but I thought it'd be interesting to see it play against another strong engine, such as Hiarcs 9, so I chose a couple of random time controls:

10 min blitz: Hiarcs-Shredder 1/2 - 1/2
Shredder -Hiarcs 0 - 1

25 min blitz: Hiarcs-Shredder 0 - 1
Shredder -Hiarcs 0 - 1

White don't seem to be getting anywhere... at least with these two engines and these time controls. Again, these results are not conclusive, but at least interesting.

The PGNs:

[Event ", Blitz:10'"]
[Site ""]
[Date "2005.01.03"]
[Round "1"]
[White "Hiarcs 9"]
[Black "Shredder 8"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 0 45"]
[PlyCount "49"]

{W=10.6 ply; 237kN/s
B=13.2 ply; 439kN/s
} 45... Qd7 {
Both last book move [%emt 0:00:13]} 46. Rg3 {[%emt 0:00:30]} Ra8 {
(Nf7-h6) [%emt 0:00:11]} 47. Reg1 {[%emt 0:00:39]} Rh8 {(Nf7-h6) [%emt 0:00:11]
} 48. Qd2 {[%emt 0:00:43]} Nh6 {[%emt 0:00:07]} 49. a4 {(Bb4-c3) [%emt 0:00:22]
} Kg7 {[%emt 0:00:06]} 50. Ne3 {(Ng4xh6) [%emt 0:00:15]} Nh5 {[%emt 0:00:18]}
51. R3g2 {[%emt 0:01:19]} Qh3 {[%emt 0:00:20]} 52. Rf2 {[%emt 0:00:19]} Bd8 {
[%emt 0:00:11]} 53. a5 {(Ne3-f1) [%emt 0:00:12]} bxa5 {[%emt 0:00:13]} 54. Bxa5
{[%emt 0:00:13]} Ng3+ {[%emt 0:00:14]} 55. Rxg3 {[%emt 0:00:00]} Qxg3 {
[%emt 0:00:12]} 56. Bxc7 {[%emt 0:00:07]} Bxc7 {[%emt 0:00:18]} 57. Nxc7 {
[%emt 0:00:14]} Rb8 {[%emt 0:00:09]} 58. Ne6+ {[%emt 0:00:06]} Kh8 {
(Kg7-g8) [%emt 0:00:18]} 59. Qe2 {(Rf2-f1) [%emt 0:00:07]} g4 {
(Nh6-f7) [%emt 0:00:12]} 60. Nf1 {[%emt 0:00:06]} gxf3 {(Rb8-b1) [%emt 0:00:12]
} 61. Rxf3 {[%emt 0:00:03]} Qg4 {[%emt 0:00:16]} 62. Qf2 {[%emt 0:00:10]} Qxe4
{[%emt 0:00:07]} 63. Ng5 {[%emt 0:00:09]} Qd4 {[%emt 0:00:09]} 64. Rf8+ {
[%emt 0:00:07]} Rxf8 {[%emt 0:00:04]} 65. Qxf8+ {[%emt 0:00:05]} Ng8 {
[%emt 0:00:00]} 66. Nf7+ {[%emt 0:00:06]} Kh7 {[%emt 0:00:00]} 67. Ng5+ {
[%emt 0:00:08]} Kh8 {[%emt 0:00:00]} 68. Nf7+ {[%emt 0:00:08]} Kh7 {
[%emt 0:00:00]} 69. Ng5+ {[%emt 0:00:07]} Kh8 {Draw accepted [%emt 0:00:00]}
1/2-1/2

[Event ", Blitz:10'"]
[Site ""]
[Date "2005.01.03"]
[Round "2"]
[White "Shredder 8"]
[Black "Hiarcs 9"]
[Result "0-1"]
[SetUp "1"]
[FEN "3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 0 45"]
[PlyCount "49"]

{W=13.9 ply; 447kN/s
B=10.4 ply; 225kN/s
} 45... Qd7 {
Both last book move [%emt 0:00:13]} 46. Rg3 {[%emt 0:00:14]} Ra8 {
[%emt 0:00:29]} 47. Kg1 {(Re1-g1) [%emt 0:00:25]} Nh6 {[%emt 0:00:10]} 48. Nxh6
{(Ng4-f2) [%emt 0:00:38]} Kxh6 {[%emt 0:00:29]} 49. Rc1 {
(Re1-b1) [%emt 0:00:24]} Rf8 {(Kh6-g7) [%emt 0:00:15]} 50. Rc2 {
(Rc1-b1) [%emt 0:00:11]} Nh5 {(Kh6-g7) [%emt 0:00:10]} 51. Rgg2 {[%emt 0:00:19]
} Rh3 {[%emt 0:00:12]} 52. Qc1 {(Rg2-f2) [%emt 0:00:18]} Rfxf3 {[%emt 0:00:16]}
53. Be1 {[%emt 0:00:18]} Nf4 {[%emt 0:00:11]} 54. Bg3 {[%emt 0:00:14]} Nxg2 {
[%emt 0:00:13]} 55. Rxg2 {[%emt 0:00:07]} c6 {(Kh6-g7) [%emt 0:00:12]} 56. Nc3
{[%emt 0:00:08]} cxd5 {(c6-c5) [%emt 0:00:24]} 57. Nxd5 {[%emt 0:00:03]} Bd8 {
[%emt 0:00:05]} 58. Qd1 {(Bg3-f2) [%emt 0:00:12]} Rxa3 {(g5-g4) [%emt 0:00:05]}
59. Qf1 {[%emt 0:00:15]} Kg7 {[%emt 0:00:13]} 60. Rf2 {[%emt 0:00:10]} Rh8 {
(g5-g4) [%emt 0:00:08]} 61. Nb4 {(Qf1-g2) [%emt 0:00:13]} Re3 {[%emt 0:00:18]}
62. Nd5 {(Qf1-g2) [%emt 0:00:11]} Rxe4 {[%emt 0:00:11]} 63. Nc3 {[%emt 0:00:19]
} Re3 {(Re4-d4) [%emt 0:00:08]} 64. Nb5 {(Nc3-d5) [%emt 0:00:33]} Qb7 {
[%emt 0:00:07]} 65. Rg2 {[%emt 0:00:07]} Be7 {(Qb7-c6) [%emt 0:00:11]} 66. Bf2
{[%emt 0:00:10]} Rf8 {[%emt 0:00:11]} 67. Qb1 {[%emt 0:00:09]} Re4 {
(Re3-f3) [%emt 0:00:08]} 68. Qc2 {[%emt 0:00:11]} Qc6 {[%emt 0:00:06]} 69. Qb2
{(Nb5-a3) [%emt 0:00:35]} Rxc4 {(Qc6xc4) [%emt 0:00:08]} 0-1

[Event ", Blitz:25'"]
[Site ""]
[Date "2005.01.03"]
[Round "1"]
[White "Hiarcs 9"]
[Black "Shredder 8"]
[Result "0-1"]
[SetUp "1"]
[FEN "3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 0 45"]
[PlyCount "75"]

{W=11.5 ply; 264kN/s
B=15.9 ply; 512kN/s
} 45... Qd7 {
Both last book move [%emt 0:00:26]} 46. Ref1 {[%emt 0:01:38]} Nh6 {
[%emt 0:00:15]} 47. Qd2 {(Ng4-f2) [%emt 0:01:04]} Ra8 {[%emt 0:00:36]} 48. Qd1
{(Ng4-e3) [%emt 0:01:48]} Kg7 {(Nf4-h3) [%emt 0:00:29]} 49. Be1 {[%emt 0:01:19]
} Rh3 {(Nh6xg4) [%emt 0:00:34]} 50. Rg3 {[%emt 0:00:35]} Rh4 {
(Rh3-h5) [%emt 0:00:27]} 51. Rgg1 {(Ng4xh6) [%emt 0:00:37]} Rh3 {[%emt 0:00:45]
} 52. Rg3 {[%emt 0:00:36]} Rh5 {[%emt 0:00:32]} 53. a4 {[%emt 0:00:32]} Rh8 {
[%emt 0:00:25]} 54. Rf2 {[%emt 0:00:27]} Nxg4 {(Rh5-h4) [%emt 0:00:30]} 55.
fxg4 {[%emt 0:00:17]} Rh3 {[%emt 0:00:46]} 56. Rb2 {(Rg3xh3) [%emt 0:00:24]}
R3h6 {(c7-c6) [%emt 0:00:48]} 57. Na7 {[%emt 0:00:18]} Qe8 {
(Rh6-h4) [%emt 0:00:49]} 58. Nc6 {[%emt 0:00:30]} Qf8 {[%emt 0:00:32]} 59. Rf2
{[%emt 0:00:30]} R8h7 {(Be7-f6) [%emt 0:00:27]} 60. Rgf3 {
(a4-a5) [%emt 0:00:49]} Qh8 {[%emt 0:00:23]} 61. Qc2 {[%emt 0:00:35]} Bf6 {
[%emt 0:00:21]} 62. a5 {[%emt 0:00:54]} bxa5 {[%emt 0:00:20]} 63. Bxa5 {
[%emt 0:00:19]} Qc8 {[%emt 0:00:06]} 64. Rg3 {(Rf3-c3) [%emt 0:00:40]} Rh3 {
[%emt 0:00:26]} 65. Rxh3 {[%emt 0:00:47]} Rxh3 {[%emt 0:00:16]} 66. Qd1 {
[%emt 0:01:33]} Re3 {[%emt 0:00:26]} 67. Rc2 {[%emt 0:00:44]} Qh8 {
[%emt 0:00:05]} 68. Qf1 {[%emt 0:00:42]} Qh4 {[%emt 0:00:20]} 69. Rf2 {
(c4-c5) [%emt 0:01:12]} Qxg4 {[%emt 0:00:22]} 70. Bd2 {[%emt 0:00:59]} Rxe4 {
[%emt 0:00:19]} 71. Na5 {[%emt 0:00:38]} Qh4 {[%emt 0:00:42]} 72. c5 {
[%emt 0:00:38]} g4 {[%emt 0:00:53]} 73. Bxf4 {[%emt 0:00:57]} Rxf4 {
(e5xf4) [%emt 0:00:14]} 74. Rxf4 {[%emt 0:00:03]} exf4 {[%emt 0:00:16]} 75.
Qxf4 {(Na5-c4) [%emt 0:00:19]} Qe1+ {[%emt 0:00:14]} 76. Kg2 {[%emt 0:00:00]}
Qe2+ {[%emt 0:00:15]} 77. Kg1 {[%emt 0:00:13]} Qd1+ {(Bf6-e5) [%emt 0:00:12]}
78. Kg2 {[%emt 0:00:04]} Qxd5+ {[%emt 0:00:16]} 79. Kf1 {[%emt 0:00:07]} Qd1+ {
[%emt 0:00:18]} 80. Kg2 {[%emt 0:00:08]} Bd4 {[%emt 0:00:14]} 81. Qf1 {
[%emt 0:00:13]} Qd2+ {(Qd1-c2+) [%emt 0:00:11]} 82. Kg3 {[%emt 0:00:02]} Qxa5 {
[%emt 0:00:45]} 0-1

[Event ", Blitz:25'"]
[Site ""]
[Date "2005.01.03"]
[Round "2"]
[White "Shredder 8"]
[Black "Hiarcs 9"]
[Result "0-1"]
[SetUp "1"]
[FEN "3q4/2p1bn1k/rp1p2p1/1N1Pp1p1/1BP1PnNr/P3QP2/7P/4R1RK b - - 0 45"]
[PlyCount "123"]

{W=17.9 ply; 577kN/s
B=13.0 ply; 255kN/s
} 45... Qd7 {
Both last book move [%emt 0:00:26]} 46. Ref1 {[%emt 0:01:05]} Nh6 {
[%emt 0:01:02]} 47. Qd2 {[%emt 0:00:36]} Ra8 {[%emt 0:01:02]} 48. Ne3 {
(Qd2-d1) [%emt 0:00:32]} Rh8 {[%emt 0:00:54]} 49. Qb2 {(Rg1-g3) [%emt 0:00:24]}
Kg8 {(Kh7-g7) [%emt 0:00:46]} 50. Be1 {(Rg1-g3) [%emt 0:00:25]} Rxh2+ {
[%emt 0:00:32]} 51. Qxh2 {[%emt 0:00:25]} Nf7 {[%emt 0:00:53]} 52. Qxh8+ {
(Ne3-g4) [%emt 0:00:31]} Nxh8 {[%emt 0:00:58]} 53. Rg3 {[%emt 0:00:34]} Nf7 {
[%emt 0:00:37]} 54. Bd2 {(Ne3-g4) [%emt 0:01:57]} Nd8 {(Kg8-g7) [%emt 0:00:35]}
55. Rf2 {[%emt 0:00:23]} Nh5 {(Nd8-b7) [%emt 0:00:50]} 56. Rg4 {[%emt 0:00:18]}
Nb7 {(Nh5-f6) [%emt 0:00:25]} 57. Kg1 {(Bd2-b4) [%emt 0:00:35]} Nc5 {
(Kg8-f7) [%emt 0:00:26]} 58. Rh2 {[%emt 0:00:24]} Bd8 {(Kg8-f7) [%emt 0:00:36]}
59. Nc2 {(Bd2-b4) [%emt 0:00:26]} Nf4 {(Kg8-g7) [%emt 0:00:39]} 60. Rg3 {
[%emt 0:00:44]} Bf6 {(Nf4-h5) [%emt 0:00:17]} 61. Nb4 {(Nc2-e3) [%emt 0:00:18]}
Nb3 {[%emt 0:00:28]} 62. Nc6 {[%emt 0:00:18]} Nxd2 {[%emt 0:00:16]} 63. Rxd2 {
[%emt 0:00:10]} Qc8 {[%emt 0:00:29]} 64. Ra2 {(a3-a4) [%emt 0:00:16]} Kf7 {
(Qc8-a6) [%emt 0:00:29]} 65. a4 {(Rg3-g4) [%emt 0:00:21]} Qh8 {
(Qc8-a6) [%emt 0:00:27]} 66. Rh2 {[%emt 0:00:18]} Qa8 {[%emt 0:00:10]} 67. Ra2
{[%emt 0:00:20]} Qc8 {(Qa8-b7) [%emt 0:00:18]} 68. Rg4 {[%emt 0:00:19]} Kf8 {
(Qc8-a6) [%emt 0:00:13]} 69. Kh1 {(Ra2-h2) [%emt 0:00:22]} Qb7 {
(Qc8-a6) [%emt 0:00:25]} 70. Kg1 {[%emt 0:00:15]} Kg7 {[%emt 0:00:15]} 71. Rh2
{(Kg1-f2) [%emt 0:00:18]} Qc8 {(Qb7-a6) [%emt 0:00:36]} 72. Rg3 {
(Rh2-a2) [%emt 0:00:13]} Qa6 {(Qc8-b7) [%emt 0:00:11]} 73. Ra2 {[%emt 0:00:14]}
Qb7 {(Kg7-g8) [%emt 0:00:22]} 74. Rg4 {(Ra2-b2) [%emt 0:00:19]} Qc8 {
[%emt 0:00:14]} 75. Kh1 {[%emt 0:00:20]} Kg8 {(Qc8-b7) [%emt 0:00:21]} 76. Rg3
{(Kh1-g1) [%emt 0:00:12]} Qd7 {(Kg8-f7) [%emt 0:00:25]} 77. Kg1 {[%emt 0:00:20]
} Kf8 {(Qd7-c8) [%emt 0:00:17]} 78. a5 {(Ra2-h2) [%emt 0:00:25]} bxa5 {
[%emt 0:00:18]} 79. Nxa5 {[%emt 0:00:12]} Qc8 {(Nf4-h3+) [%emt 0:00:16]} 80.
Nc6 {[%emt 0:00:25]} Qb7 {[%emt 0:00:12]} 81. Kh2 {[%emt 0:00:14]} Kg8 {
(Kf8-g7) [%emt 0:00:09]} 82. Rg1 {(Rg3-g4) [%emt 0:00:13]} Qb6 {
(Bf6-g7) [%emt 0:00:07]} 83. Rf1 {[%emt 0:00:18]} Qe3 {(g5-g4) [%emt 0:00:08]}
84. Nxc7 {[%emt 0:00:13]} Qd3 {[%emt 0:00:14]} 85. Raf2 {[%emt 0:00:16]} g4 {
[%emt 0:00:14]} 86. Kg3 {[%emt 0:00:21]} Ne2+ {[%emt 0:00:07]} 87. Kxg4 {
[%emt 0:00:33]} Qd2 {(Qd3-e3) [%emt 0:00:05]} 88. Kh3 {[%emt 0:00:13]} Bh4 {
[%emt 0:00:10]} 89. Ne6 {[%emt 0:00:13]} Bxf2 {[%emt 0:00:07]} 90. Rxf2 {
[%emt 0:00:14]} Qe1 {[%emt 0:00:08]} 91. Rxe2 {[%emt 0:00:15]} Qxe2 {
[%emt 0:00:06]} 92. Kg3 {[%emt 0:00:11]} Qxc4 {[%emt 0:00:06]} 93. Kf2 {
(Nc6-d8) [%emt 0:00:13]} Kf7 {(Qc4-c2+) [%emt 0:00:09]} 94. Ng5+ {
(Ne6-d8+) [%emt 0:00:12]} Kf6 {[%emt 0:00:07]} 95. Nh7+ {[%emt 0:00:13]} Kg7 {
[%emt 0:00:03]} 96. Ng5 {[%emt 0:00:18]} Qc1 {(Qc4-c2+) [%emt 0:00:18]} 97. Nh3
{[%emt 0:00:16]} Qh1 {(Kg7-f6) [%emt 0:00:08]} 98. Kg3 {[%emt 0:00:07]} Kf6 {
[%emt 0:00:17]} 99. Nd8 {[%emt 0:00:09]} g5 {[%emt 0:00:10]} 100. Ne6 {
[%emt 0:00:06]} g4 {(Qh1-e1+) [%emt 0:00:09]} 101. Kxg4 {[%emt 0:00:07]} Qg2+ {
[%emt 0:00:07]} 102. Kh4 {[%emt 0:00:06]} Qxf3 {[%emt 0:00:09]} 103. Nhg5 {
[%emt 0:00:15]} Qh1+ {[%emt 0:00:06]} 104. Kg3 {[%emt 0:02:34]} Qe1+ {
[%emt 0:00:06]} 105. Kg2 {(Kg3-g4) [%emt 0:00:10]} Qe2+ {[%emt 0:00:08]} 106.
Kg3 {[%emt 0:00:05]} Kg6 {[%emt 0:00:06]} 0-1

Pallav Nawani wrote:

Jaime Benito de Valle wrote:This two games prove nothing really, but you might find some insights from them.

I expected this Shredder is way too strong for fruit. In fact, I think that Shredder can beat fruit even if you handicap it with two pawns. Perhaps a better test is to pit List512 against ProDeo??

Or better still, post it on talkchess.com forums.

Pallav

I disagree here
2 pawns are clearly more than the difference between fruit and shredder.

The difference between Shredder and Fruit is less than 200 elo based on rating lists that I saw.

Fruit may have problem in understanding this type of position but I do not like bad remarks about fruit because of one position(in a different position Fruit may even beat Shredder with both sides).

Uri

Uri Blass wrote:

Pallav Nawani wrote:
I expected this Shredder is way too strong for fruit. In fact, I think that Shredder can beat fruit even if you handicap it with two pawns. Perhaps a better test is to pit List512 against ProDeo??

Or better still, post it on talkchess.com forums.

Pallav

I disagree here
2 pawns are clearly more than the difference between fruit and shredder.

The difference between Shredder and Fruit is less than 200 elo based on rating lists that I saw.

Fruit may have problem in understanding this type of position but I do not like bad remarks about fruit because of one position(in a different position Fruit may even beat Shredder with both sides).

Uri

I wasn't intending to bad mouth Fruit, I just wanted to point out that shredder was too strong for Fruit, so even if blacks position was slightly better, shredder would still win with white, and vice versa, and thus the results would prove nothing. The two pawns was just an example.

However, as you say, 2 pawns is too much. According to master opinion, 200 rating point difference is equivalent to 1 pawn advantage:
http://mywebpages.comcast.net/danheisman/Articles/evaluation_of_material_imbalance.htm

Pallav