Reply
 
LinkBack Thread Tools Display Modes
  #1   Report Post  
Old August 16th 07, 11:34 AM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Nov 2006
Posts: 364
Default Test position

Is there any chess program that, in the following test position,
avoids closing the position in this or the next move? Closing the
position with c5-c4 is very bad because white replies b3-b4 and then
black has no counterplay on the queen's flank. Although white's bishop
is closed in at a1, black also has a weakness at a3. White has the
king's flank to himself.

No human player above 1300 in ELO would play c5-c4. It's a strategical
blunder of immense proportions. That's why this position is
interesting to use for testing positional understanding in a chess
engine. Fritz10, of course, plays c5-c4. But how about Hiarcs11, Rybka,
and other engines? Please test! You can either import the following as
PGN, or copy and paste the FEN code.


Mats
-------------------------------------

[Event "*"]
[Site "?"]
[Date "2007.08.16"]
[White "Test position"]
[Black "-"]
[Result "*"]
[SetUp "1"]
[FEN "r1r2nk1/1bq1bppp/1p2p2n/2ppP3/3P4/pPP2N1P/P2Q1PP1/BBR1RNK1 b - - 0 1"]
*


  #2   Report Post  
Old August 16th 07, 01:23 PM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Jul 2006
Posts: 1,015
Default Test position

On Aug 16, 11:34 am, "M Winther" wrote:
Is there any chess program that, in the following test position,
avoids closing the position in this or the next move? Closing the
position with c5-c4 is very bad because white replies b3-b4


Almost all of them prefer one of Ng6 or Nf5. Even Fritz 8. I am very
surprised if FritzX plays c4 here.

engine. Fritz10, of course, plays c5-c4. But how about Hiarcs11, Rybka,
and other engines? Please test!


I tested at roughly 2 minutes a move with the following results:

Comet B6 ply10 Ng6, Rab8, c4 -0.15
Crafty19.19 ply 10 Ng6,Qd8,Qd7 0.12, Ba6 0.15 (c4 well down)
Fritz5.32 ply12 Ng6,Nf5 0.31, next 9 moves score 0.34 at ply13 prefers
c4
Fritz8 ply10 c4, ply11 Ng6, ply12 Ng6, Rd7,Rab8 -0.12, Bc6 -0.09,
c4,Qd8,Rcb8,Kh8 -0.06
HIARCS ply12 Ba6, Qd8, Rd8, Ng6, Qd7
Junior7 ply17 Ng6 0.14, c4 0.12
Rybka2.32 Nf5, c4, Ng6
Shredder10 ply12 Nf5 0.12, Ng6, Qd7 0.13, Qd8 0.14, c4 0.22
Turing ply6 Nf5

4 votes for Ng6, 3 votes for Nf5, 1 vote for c4 and 1 vote for Ba6 at
2 mins/move

PGN, or copy and paste the FEN code.
r1r2nk1/1bq1bppp/1p2p2n/2ppP3/3P4/pPP2N1P/P2Q1PP1/BBR1RNK1 b - - 0 1


Of the engines I tested Crafty19.19 and Shredder10 showed the most
disdain for 1. ... c4 because they could see that white could advance
a pawn and block the position. Engines that expect that any parity
material swapoff will occur take a long time to see that it isn't
whites best response - Fritz8 goes for Ng6 ply 12 and beyond.

It is amusing that the humble Turing engine agrees with Rybka on the
best move. I think I prefer Ng6 without having looked all that deeply.
Nf5 looks to me like it could help white develop an attack on the
kingside.

Regards,
Martin Brown

  #3   Report Post  
Old August 16th 07, 03:53 PM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Nov 2006
Posts: 364
Default Test position

Den 2007-08-16 14:23:32 skrev Martin Brown :

On Aug 16, 11:34 am, "M Winther" wrote:
Is there any chess program that, in the following test position,
avoids closing the position in this or the next move? Closing the
position with c5-c4 is very bad because white replies b3-b4


Almost all of them prefer one of Ng6 or Nf5. Even Fritz 8. I am very
surprised if FritzX plays c4 here.

engine. Fritz10, of course, plays c5-c4. But how about Hiarcs11, Rybka,
and other engines? Please test!


I tested at roughly 2 minutes a move with the following results:

Comet B6 ply10 Ng6, Rab8, c4 -0.15
Crafty19.19 ply 10 Ng6,Qd8,Qd7 0.12, Ba6 0.15 (c4 well down)
Fritz5.32 ply12 Ng6,Nf5 0.31, next 9 moves score 0.34 at ply13 prefers
c4
Fritz8 ply10 c4, ply11 Ng6, ply12 Ng6, Rd7,Rab8 -0.12, Bc6 -0.09,
c4,Qd8,Rcb8,Kh8 -0.06
HIARCS ply12 Ba6, Qd8, Rd8, Ng6, Qd7
Junior7 ply17 Ng6 0.14, c4 0.12
Rybka2.32 Nf5, c4, Ng6
Shredder10 ply12 Nf5 0.12, Ng6, Qd7 0.13, Qd8 0.14, c4 0.22
Turing ply6 Nf5

4 votes for Ng6, 3 votes for Nf5, 1 vote for c4 and 1 vote for Ba6 at
2 mins/move

PGN, or copy and paste the FEN code.
r1r2nk1/1bq1bppp/1p2p2n/2ppP3/3P4/pPP2N1P/P2Q1PP1/BBR1RNK1 b - - 0 1


Of the engines I tested Crafty19.19 and Shredder10 showed the most
disdain for 1. ... c4 because they could see that white could advance
a pawn and block the position. Engines that expect that any parity
material swapoff will occur take a long time to see that it isn't
whites best response - Fritz8 goes for Ng6 ply 12 and beyond.

It is amusing that the humble Turing engine agrees with Rybka on the
best move. I think I prefer Ng6 without having looked all that deeply.
Nf5 looks to me like it could help white develop an attack on the
kingside.

Regards,
Martin Brown




Thanks for the testing work. Fritz 10 prefers the move c5-c4 from the
beginning and never changes its mind, even to depth 20. It regards the
position as -0.94, that is, clearly better for black.

I tested an old DOS program, MChess Pro 7 (former World Champion,
and it regards only Ng6, from ply 8 and onwards. It evaluates the position
as 0.00.

I was glad to hear that Shredder and Crafty rejected c5-c4 out of
hand. That's what I would like to see. Hiarcs seems to reject the
move, too. But I was disappointed to see that Rybka at some level
considers it.

Two minutes on a powerful computer is very much. It's not certain that
the programs would use that much time in a tournament game. And they
would certainly not use that much time in a rapid game. During
analysis one often uses less time, too. Could you remake the test with
15 seconds also? It would clarify how big role the positional
evaluation plays here, rather than mere ply depth.

Moreover, if I'm not asking too much, could you check how the programs
evaluate the position after the following moves? 1...c5-c4 2.b3-b4. It's the
FEN below. The problem is, namely, that the programs might have rejected
c5-c4 because they evaluate b3xc4 as good for white. It's interesting to see
how they regard the closed position in which really only white can
win. Fritz 10 evaluates it as -1.27, clearly advantageous for black.

------------------

[Event "*"]
[Site "?"]
[Date "2007.08.16"]
[White "Test position 2"]
[Black "-"]
[Result "*"]
[SetUp "1"]
r1r2nk1/1bq1bppp/1p2p2n/3pP3/1PpP4/p1P2N1P/P2Q1PP1/BBR1RNK1 b - - 0 2
*


Mats



  #4   Report Post  
Old August 16th 07, 03:59 PM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Nov 2006
Posts: 364
Default Test position

Correction. This is the correct PGN:

[Event "*"]
[Site "?"]
[Date "2007.08.16"]
[White "Test position 2"]
[Black "-"]
[Result "*"]
[SetUp "1"]
[FEN "r1r2nk1/1bq1bppp/1p2p2n/3pP3/1PpP4/p1P2N1P/P2Q1PP1/BBR1RNK1 b - - 0 2"
*
  #5   Report Post  
Old August 16th 07, 07:08 PM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Aug 2007
Posts: 2
Default Test position

Here's what crafty 21.5 has to say about your position (first number is
depth, last number tells after how many seconds it has this principal
variation).

3 +0.04 1. ... Nf5 2. Ne3 Qd7 3. Nxf5 exf5 (0.01)

3 -0.15 1. ... Ba6 2. Ng3 Ng6 3. Bxg6 hxg6 (0.01)

3 -0.15 1. ... Ba6 2. Ng3 Ng6 3. Bxg6 hxg6 (0.02)

4 +0.00 1. ... Ba6 2. Ne3 Ng6 3. c4 (0.02)

4 +0.00 1. ... Ba6 2. Ne3 Ng6 3. c4 (0.03)

5 +0.05 1. ... Ba6 2. Ne3 cxd4 3. cxd4 Qd7 (0.03)

5 -0.09 1. ... Ng6 2. Ng3 Ba6 3. Kh1 Bb5 4. Bxg6 hxg6 (0.05)

5 -0.09 1. ... Ng6 2. Ng3 Ba6 3. Kh1 Bb5 4. Bxg6 hxg6 (0.06)

6 +0.04 1. ... Ng6 2. Ng3 Ba6 3. Bxg6 hxg6 4. Kh1 (0.07)

6 +0.04 1. ... Ng6 2. Ng3 Ba6 3. Bxg6 hxg6 4. Kh1 (0.12)

7 +0.01 1. ... Ng6 2. Ng3 Ba6 3. Kh1 Kh8 4. Bxg6 hxg6 (0.16)

7 +0.01 1. ... Ng6 2. Ng3 Ba6 3. Kh1 Kh8 4. Bxg6 hxg6 (0.22)

8 +0.00 1. ... Ng6 2. Ne3 Nh4 3. Nxh4 Bxh4 4. Bd3 Qe7 5. Kh1 (0.33)

8 +0.00 1. ... Ng6 2. Ne3 Nh4 3. Nxh4 Bxh4 4. Bd3 Qe7 5. Kh1 (0.53)

9 -0.05 1. ... Ng6 2. Ne3 Nh4 3. Nxh4 Bxh4 4. Bd3 Qe7 5. Qc2 f5 (0.68)

9 -0.05 1. ... Ng6 2. Ne3 Nh4 3. Nxh4 Bxh4 4. Bd3 Qe7 5. Qc2 f5 (0.83)

10 -0.01 1. ... Ng6 2. Ne3 c4 3. b4 Nh4 4. Nxh4 Bxh4 5. Bc2 Qe7 6. Kh1
(1.71)

10 -0.01 1. ... Ng6 2. Ne3 c4 3. b4 Nh4 4. Nxh4 Bxh4 5. Bc2 Qe7 6. Kh1
(2.83)

11 +0.01 1. ... Ng6 2. dxc5 Bxc5 3. Ng3 Ba6 4. Kh1 Be7 5. Rcd1 Qc5 6.
Nd4 Kh8 7. Bxg6 hxg6 (3.99)

11 +0.01 1. ... Ng6 2. dxc5 Bxc5 3. Ng3 Ba6 4. Kh1 Be7 5. Rcd1 Qc5 6.
Nd4 Kh8 7. Bxg6 hxg6 (5.19)

12 -0.04 1. ... Ng6 2. dxc5 bxc5 3. Ng3 c4 4. b4 Bc6 5. Bxg6 hxg6 6. Rb1
Nf5 7. Qf4 (8.31)

12 -0.04 1. ... Ng6 2. dxc5 bxc5 3. Ng3 c4 4. b4 Bc6 5. Bxg6 hxg6 6. Rb1
Nf5 7. Qf4 (9.72)

13 +0.01 1. ... Ng6 2. dxc5 bxc5 3. Ng3 Rab8 4. Bc2 {HT} (13.54)

13 +0.01 1. ... Ng6 2. dxc5 bxc5 3. Ng3 Rab8 4. Bc2 {HT} (19.82)

14 +0.02 1. ... Ng6 2. dxc5 bxc5 3. Ng3 c4 4. b4 {HT} (32.93)

14 +0.02 1. ... Ng6 2. dxc5 bxc5 3. Ng3 c4 4. b4 {HT} (52.82)

15 +0.04 1. ... Ng6 2. dxc5 bxc5 3. Ng3 Rcb8 4. Qe3 Bc6 5. Rcd1 c4 6.
bxc4 dxc4 7. Bxg6 hxg6 8. Qf4 Bd5 (81.68)

15 -0.11 1. ... c4 2. bxc4 dxc4 3. Be4 Bxe4 4. Rxe4 Nf5 5. Ne3 Nxe3 6.
Qxe3 Ng6 7. Rb1 Rd8 8. Ng5 Rd5 (148.46)

15 -0.11 1. ... c4 2. bxc4 dxc4 3. Be4 Bxe4 4. Rxe4 Nf5 5. Ne3 Nxe3 6.
Qxe3 Ng6 7. Rb1 Rd8 8. Ng5 Rd5 (156.22)

16 -0.10 1. ... c4 2. b4 Nf5 3. Ne3 Nxe3 4. Qxe3 Ng6 5. Ng5 Qd8 6. f4
Bc6 7. h4 Qd7 8. h5 Bxg5 9. fxg5 (207.44)

16 -0.10 1. ... c4 2. b4 Nf5 3. Ne3 Nxe3 4. Qxe3 Ng6 5. Ng5 Qd8 6. f4
Bc6 7. h4 Qd7 8. h5 Bxg5 9. fxg5 (239.54)

--- Here I play c4 and b4 ---

....

12 -0.02 2. ... Nf5 3. Bxf5 exf5 {HT} (0.31)

12 -0.02 2. ... Nf5 3. Bxf5 exf5 {HT} (0.50)

13 -0.02 2. ... Nf5 3. Bxf5 exf5 {HT} (0.51)

13 -0.02 2. ... Nf5 3. Bxf5 exf5 {HT} (2.55)

14 -0.10 2. ... Nf5 3. Bxf5 exf5 4. Ng3 g6 5. Ng5 Qd8 6. h4 Ne6 7. Nxe6
fxe6 8. h5 gxh5 9. Nxh5 (12.15)

14 -0.10 2. ... Nf5 3. Bxf5 exf5 4. Ng3 g6 5. Ng5 Qd8 6. h4 Ne6 7. Nxe6
fxe6 8. h5 gxh5 9. Nxh5 (12.97)

15 +0.07 2. ... Nf5 3. g3 Nh6 4. h4 Bc6 5. h5 Ba4 6. Ne3 Nd7 7. g4 Qd8
8. Re2 Ra6 9. Kh1 Raa8 (52.99)

15 +0.07 2. ... Nf5 3. g3 Nh6 4. h4 Bc6 5. h5 Ba4 6. Ne3 Nd7 7. g4 Qd8
8. Re2 Ra6 9. Kh1 Raa8 (123.70)

16 +0.08 2. ... Nf5 3. g3 g6 4. h4 Qc6 5. Ne3 Nxe3 6. Qxe3 Nd7 7. Ng5 h6
8. Nxf7 Kxf7 9. Qxh6 Ke8 10. Bxg6+ Kd8 11. Qf4 (168.74)

16 +0.04 2. ... Qd8 3. g3 Bc6 4. Ne3 Nd7 5. Re2 Ba4 6. h4 Bb5 7. h5 Bg5
8. Rce1 Bxe3 9. Qxe3 Nf5 10. Bxf5 exf5 (195.98)

16 +0.04 2. ... Qd8 3. g3 Bc6 4. Ne3 Nd7 5. Re2 Ba4 6. h4 Bb5 7. h5 Bg5
8. Rce1 Bxe3 9. Qxe3 Nf5 10. Bxf5 exf5 (221.50)


  #6   Report Post  
Old August 16th 07, 07:35 PM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Aug 2007
Posts: 2
Default Test position

Crafty 21.5 does recover from c4 in the intial position, however.
Unfortunately I want to restart X now, so I'll leave it like this.

17 -0.02 1. ... c4 2. b4 Nf5 3. Ne3 Nxe3 4. Qxe3 Ng6 5. Ng5 Qd7 6. h4
{HT} (336.51)

17 -0.02 1. ... c4 2. b4 Nf5 3. Ne3 Nxe3 4. Qxe3 Ng6 5. Ng5 Qd7 6. h4
{HT} (536.05)

18 +0.06 1. ... c4 2. b4 Nf5 3. Ne3 Nxe3 4. Qxe3 Ng6 5. Ng5 Qd8 6. h4
Nxh4 7. Nxh7 Nf5 8. Qh3 Bg5 9. Rcd1 g6 10. Nxg5 Qxg5 (759.74)

18 +0.01 1. ... Ng6 2. dxc5 bxc5 3. Ng3 Rcb8 4. Qe3 c4 5. b4 Bc6 6. Bc2
Ba4 7. Bxa4 Rxa4 8. Rb1 Ra6 9. Ng5 Rab6 10. Kh1 (1226.21)
  #7   Report Post  
Old August 17th 07, 12:36 AM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: May 2006
Posts: 1,105
Default Test position


"M Winther" wrote in message news[email protected]
Is there any chess program that, in the following test position,
avoids closing the position in this or the next move? Closing the
position with c5-c4 is very bad because white replies b3-b4 and then
black has no counterplay on the queen's flank. Although white's bishop
is closed in at a1, black also has a weakness at a3. White has the
king's flank to himself.

No human player above 1300 in ELO would play c5-c4.


This is certainly not true.1300's (and higher) make far more
straightforward blunders than this.

It's a strategical
blunder of immense proportions.


It would be educational for this player (somewhat 1300) to see actual analysis
supporting the "White has the king's flank to himself" claim. Or an example of
the "counterplay" when the Queenside isn't blocked.

Computers aren't trying to mimic human preference, but to win. Do you
beat these engines with Black before c4, or with White after c4 and b4?
If so, it would be interesting to post your games.

That's why this position is
interesting to use for testing positional understanding in a chess
engine. Fritz10, of course, plays c5-c4. But how about Hiarcs11, Rybka,
and other engines? Please test! You can either import the following as
PGN, or copy and paste the FEN code.


Mats
-------------------------------------

[Event "*"]
[Site "?"]
[Date "2007.08.16"]
[White "Test position"]
[Black "-"]
[Result "*"]
[SetUp "1"]
[FEN "r1r2nk1/1bq1bppp/1p2p2n/2ppP3/3P4/pPP2N1P/P2Q1PP1/BBR1RNK1 b - - 0 1"]
*






  #8   Report Post  
Old August 17th 07, 06:02 AM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Nov 2006
Posts: 364
Default Test position

Den 2007-08-17 01:36:45 skrev David Kane :


"M Winther" wrote in message news[email protected]
Is there any chess program that, in the following test position,
avoids closing the position in this or the next move? Closing the
position with c5-c4 is very bad because white replies b3-b4 and then
black has no counterplay on the queen's flank. Although white's bishop
is closed in at a1, black also has a weakness at a3. White has the
king's flank to himself.

No human player above 1300 in ELO would play c5-c4.


This is certainly not true.1300's (and higher) make far more
straightforward blunders than this.

It's a strategical
blunder of immense proportions.


It would be educational for this player (somewhat 1300) to see actual analysis
supporting the "White has the king's flank to himself" claim. Or an example of
the "counterplay" when the Queenside isn't blocked.

Computers aren't trying to mimic human preference, but to win. Do you
beat these engines with Black before c4, or with White after c4 and b4?
If so, it would be interesting to post your games.


David, "computers" generally seem to agree with me that Ng6 is better
than c5-c4, so I have no beef with them. However, as I expected, the
engines seem to overvalue c5-c4, i.e., it's not that bad.

Obviously c5-c4 is bad, I don't have to prove that in a superhuman way by
beating engines with grandmaster strength. It's bad because it cuts down
the strategical possibilities for black to a minimum. White, however, have
much better opportunities. Before c5-c4 black has a good position. After,
his strategical prospects are scarce.

Mats
  #9   Report Post  
Old August 17th 07, 06:05 AM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Nov 2006
Posts: 364
Default Test position

Den 2007-08-16 20:08:49 skrev JeeBee :

Here's what crafty 21.5 has to say about your position (first number is
depth, last number tells after how many seconds it has this principal
variation).

3 +0.04 1. ... Nf5 2. Ne3 Qd7 3. Nxf5 exf5 (0.01)

3 -0.15 1. ... Ba6 2. Ng3 Ng6 3. Bxg6 hxg6 (0.01)

3 -0.15 1. ... Ba6 2. Ng3 Ng6 3. Bxg6 hxg6 (0.02)

4 +0.00 1. ... Ba6 2. Ne3 Ng6 3. c4 (0.02)

4 +0.00 1. ... Ba6 2. Ne3 Ng6 3. c4 (0.03)

5 +0.05 1. ... Ba6 2. Ne3 cxd4 3. cxd4 Qd7 (0.03)

5 -0.09 1. ... Ng6 2. Ng3 Ba6 3. Kh1 Bb5 4. Bxg6 hxg6 (0.05)

5 -0.09 1. ... Ng6 2. Ng3 Ba6 3. Kh1 Bb5 4. Bxg6 hxg6 (0.06)

6 +0.04 1. ... Ng6 2. Ng3 Ba6 3. Bxg6 hxg6 4. Kh1 (0.07)

6 +0.04 1. ... Ng6 2. Ng3 Ba6 3. Bxg6 hxg6 4. Kh1 (0.12)

7 +0.01 1. ... Ng6 2. Ng3 Ba6 3. Kh1 Kh8 4. Bxg6 hxg6 (0.16)

7 +0.01 1. ... Ng6 2. Ng3 Ba6 3. Kh1 Kh8 4. Bxg6 hxg6 (0.22)

8 +0.00 1. ... Ng6 2. Ne3 Nh4 3. Nxh4 Bxh4 4. Bd3 Qe7 5. Kh1 (0.33)

8 +0.00 1. ... Ng6 2. Ne3 Nh4 3. Nxh4 Bxh4 4. Bd3 Qe7 5. Kh1 (0.53)

9 -0.05 1. ... Ng6 2. Ne3 Nh4 3. Nxh4 Bxh4 4. Bd3 Qe7 5. Qc2 f5 (0.68)

9 -0.05 1. ... Ng6 2. Ne3 Nh4 3. Nxh4 Bxh4 4. Bd3 Qe7 5. Qc2 f5 (0.83)

10 -0.01 1. ... Ng6 2. Ne3 c4 3. b4 Nh4 4. Nxh4 Bxh4 5. Bc2 Qe7 6. Kh1
(1.71)

10 -0.01 1. ... Ng6 2. Ne3 c4 3. b4 Nh4 4. Nxh4 Bxh4 5. Bc2 Qe7 6. Kh1
(2.83)

11 +0.01 1. ... Ng6 2. dxc5 Bxc5 3. Ng3 Ba6 4. Kh1 Be7 5. Rcd1 Qc5 6.
Nd4 Kh8 7. Bxg6 hxg6 (3.99)

11 +0.01 1. ... Ng6 2. dxc5 Bxc5 3. Ng3 Ba6 4. Kh1 Be7 5. Rcd1 Qc5 6.
Nd4 Kh8 7. Bxg6 hxg6 (5.19)

12 -0.04 1. ... Ng6 2. dxc5 bxc5 3. Ng3 c4 4. b4 Bc6 5. Bxg6 hxg6 6. Rb1
Nf5 7. Qf4 (8.31)

12 -0.04 1. ... Ng6 2. dxc5 bxc5 3. Ng3 c4 4. b4 Bc6 5. Bxg6 hxg6 6. Rb1
Nf5 7. Qf4 (9.72)

13 +0.01 1. ... Ng6 2. dxc5 bxc5 3. Ng3 Rab8 4. Bc2 {HT} (13.54)

13 +0.01 1. ... Ng6 2. dxc5 bxc5 3. Ng3 Rab8 4. Bc2 {HT} (19.82)

14 +0.02 1. ... Ng6 2. dxc5 bxc5 3. Ng3 c4 4. b4 {HT} (32.93)

14 +0.02 1. ... Ng6 2. dxc5 bxc5 3. Ng3 c4 4. b4 {HT} (52.82)

15 +0.04 1. ... Ng6 2. dxc5 bxc5 3. Ng3 Rcb8 4. Qe3 Bc6 5. Rcd1 c4 6.
bxc4 dxc4 7. Bxg6 hxg6 8. Qf4 Bd5 (81.68)

15 -0.11 1. ... c4 2. bxc4 dxc4 3. Be4 Bxe4 4. Rxe4 Nf5 5. Ne3 Nxe3 6.
Qxe3 Ng6 7. Rb1 Rd8 8. Ng5 Rd5 (148.46)

15 -0.11 1. ... c4 2. bxc4 dxc4 3. Be4 Bxe4 4. Rxe4 Nf5 5. Ne3 Nxe3 6.
Qxe3 Ng6 7. Rb1 Rd8 8. Ng5 Rd5 (156.22)

16 -0.10 1. ... c4 2. b4 Nf5 3. Ne3 Nxe3 4. Qxe3 Ng6 5. Ng5 Qd8 6. f4
Bc6 7. h4 Qd7 8. h5 Bxg5 9. fxg5 (207.44)

16 -0.10 1. ... c4 2. b4 Nf5 3. Ne3 Nxe3 4. Qxe3 Ng6 5. Ng5 Qd8 6. f4
Bc6 7. h4 Qd7 8. h5 Bxg5 9. fxg5 (239.54)

--- Here I play c4 and b4 ---

...

12 -0.02 2. ... Nf5 3. Bxf5 exf5 {HT} (0.31)

12 -0.02 2. ... Nf5 3. Bxf5 exf5 {HT} (0.50)

13 -0.02 2. ... Nf5 3. Bxf5 exf5 {HT} (0.51)

13 -0.02 2. ... Nf5 3. Bxf5 exf5 {HT} (2.55)

14 -0.10 2. ... Nf5 3. Bxf5 exf5 4. Ng3 g6 5. Ng5 Qd8 6. h4 Ne6 7. Nxe6
fxe6 8. h5 gxh5 9. Nxh5 (12.15)

14 -0.10 2. ... Nf5 3. Bxf5 exf5 4. Ng3 g6 5. Ng5 Qd8 6. h4 Ne6 7. Nxe6
fxe6 8. h5 gxh5 9. Nxh5 (12.97)

15 +0.07 2. ... Nf5 3. g3 Nh6 4. h4 Bc6 5. h5 Ba4 6. Ne3 Nd7 7. g4 Qd8
8. Re2 Ra6 9. Kh1 Raa8 (52.99)

15 +0.07 2. ... Nf5 3. g3 Nh6 4. h4 Bc6 5. h5 Ba4 6. Ne3 Nd7 7. g4 Qd8
8. Re2 Ra6 9. Kh1 Raa8 (123.70)

16 +0.08 2. ... Nf5 3. g3 g6 4. h4 Qc6 5. Ne3 Nxe3 6. Qxe3 Nd7 7. Ng5 h6
8. Nxf7 Kxf7 9. Qxh6 Ke8 10. Bxg6+ Kd8 11. Qf4 (168.74)

16 +0.04 2. ... Qd8 3. g3 Bc6 4. Ne3 Nd7 5. Re2 Ba4 6. h4 Bb5 7. h5 Bg5
8. Rce1 Bxe3 9. Qxe3 Nf5 10. Bxf5 exf5 (195.98)

16 +0.04 2. ... Qd8 3. g3 Bc6 4. Ne3 Nd7 5. Re2 Ba4 6. h4 Bb5 7. h5 Bg5
8. Rce1 Bxe3 9. Qxe3 Nf5 10. Bxf5 exf5 (221.50)



He regards it as equal then. Just as I expected, today's engines, not even Rybka, have
an effective way of thinking strategically. This weakness could be utilized by human
grandmasters.

Mats
  #10   Report Post  
Old August 17th 07, 07:28 AM posted to rec.games.chess.computer
external usenet poster
 
First recorded activity by ChessBanter: Jun 2007
Posts: 21
Default Test position

M Winther a écrit :
Den 2007-08-17 01:36:45 skrev David Kane :


"M Winther" wrote in message
news[email protected]
Is there any chess program that, in the following test position,
avoids closing the position in this or the next move? Closing the
position with c5-c4 is very bad because white replies b3-b4 and then
black has no counterplay on the queen's flank. Although white's bishop
is closed in at a1, black also has a weakness at a3. White has the
king's flank to himself.

No human player above 1300 in ELO would play c5-c4.


This is certainly not true.1300's (and higher) make far more
straightforward blunders than this.

It's a strategical
blunder of immense proportions.


It would be educational for this player (somewhat 1300) to see actual
analysis
supporting the "White has the king's flank to himself" claim. Or an
example of
the "counterplay" when the Queenside isn't blocked.

Computers aren't trying to mimic human preference, but to win. Do you
beat these engines with Black before c4, or with White after c4 and b4?
If so, it would be interesting to post your games.


David, "computers" generally seem to agree with me that Ng6 is better
than c5-c4, so I have no beef with them. However, as I expected, the
engines seem to overvalue c5-c4, i.e., it's not that bad.

Obviously c5-c4 is bad, I don't have to prove that in a superhuman way by
beating engines with grandmaster strength. It's bad because it cuts down
the strategical possibilities for black to a minimum. White, however, have
much better opportunities. Before c5-c4 black has a good position. After,
his strategical prospects are scarce.

Mats

Well, that is not that sure.
I tried a simple test. I set up the position after c4 and b4, and let
Fritz10 play black and rybka2.3.2 play white at tournament time control
(40 moves in 2 hours).
Well, evaluation by Fritz doesn't change and, believe it or not, Fritz
ends up winning the game by very slowly opening the game on the
queenside and letting white bishop trapped on a1.
What is an "evident mistake by a 1300 player" might not be a mistake for
a computer chess program, which has all the necessary patience and
computing power to exploit that kind of position...
The game is available he

http://www.chess-lovers.org/fritz10/test.htm

Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Test Positions Martin Brown rec.games.chess.analysis (Chess Analysis) 0 May 9th 07 04:20 PM
Test Positions Martin Brown rec.games.chess.computer (Computer Chess) 0 May 9th 07 04:20 PM
Bobby Fischer has been reinstated in the USCF [email protected] rec.games.chess.misc (Chess General) 89 September 4th 06 08:02 PM
ChessCafe blackmailing USCF? [email protected] rec.games.chess.politics (Chess Politics) 165 October 7th 05 08:05 PM
ChessCafe blackmailing USCF? [email protected] rec.games.chess.misc (Chess General) 117 October 7th 05 06:04 PM


All times are GMT +1. The time now is 08:42 PM.

Powered by vBulletin® Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
Copyright ©2004-2019 ChessBanter.
The comments are property of their posters.
 

About Us

"It's about Chess"

 

Copyright © 2017