Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka Human vs "My System"
- - By diskamyl (**) Date 2008-08-10 20:06
I've begun studying My System recently, and am using Rybka Human for some positions, assuming it's the best engine available for that. (This topic isn't trying to suggest otherwise.) But I'll post the contradictions between the two evaluations, as they come up.

in this position, after  1. e4 e5 2. d4 exd4 3. c3 Nf6 4. e5 Ne4 5. Bd3 Nc5 6. cxd4 Nxd3+ 7. Qxd3
rnbqkb1r/pppp1ppp/8/4P3/3P4/3Q4/PP3PPP/RNB1K1NR b KQkq - 0 7

Nimzowitsch evaluates the position as better for white, tells he's up 4 tempi now (and he is), but Rybka Human evaluates the position as equal, and very slightly better for black.

Why is that, isn't white clearly far ahead in developement?
Parent - - By Vempele (Silver) Date 2008-08-10 20:08
Rybka Human overvalues static factors compared to Default and Dynamic.
Parent - - By Permanent Brain (*****) Date 2008-08-10 20:38
The following position is from an experiment I made in 2002:

rnbqkbnr/pppppppp/8/8/3PP3/2NB1N2/PPPBQPPP/R3K2R b KQkq - 0 7

How do the Rybka 3 versions evaluate this position?

For comparison, Analysis by Rybka 2.3.2a 32-bit (D945 3.4 GHz):

7...Nc6 8.0-0-0 d6 9.d5 Nb4 10.Bc4
  +/-  (1.39)   Depth: 5   00:00:00
  +-  (1.61)   Depth: 9   00:00:00  50kN
7...d6 8.h3 Nc6 9.0-0 Nf6 10.a3 h5 11.Bc4 e6 12.Rad1 Na5
  +/-  (1.39)   Depth: 9   00:00:02  147kN
  +-  (1.41)   Depth: 14   00:01:53  7345kN
Parent - - By sarciness (***) Date 2008-08-10 20:47 Edited 2008-08-10 20:49
I did a similar exmeriment with R2. Here's R3 (32bit)'s analysis:

[+1.29]  d=14  1...e6 2.h3 Ne7 3.d5 g6 4.O–O–O Bg7 5.dxe6 dxe6 6.Kb1 c5 7.e5 O–O (0:06.28)
[+1.29]  d=13  1...e6 2.h3 Ne7 3.d5 g6 4.O–O–O Bg7 5.dxe6 fxe6 6.Kb1 Nbc6 7.h4 (0:05.16)
[+1.34]  d=12  1...e6 2.d5 Nf6 3.dxe6 dxe6 4.e5 Nd5 5.Nxd5 Qxd5 6.Be4 Qc5 7.O–O–O (0:01.19)
[+1.43]  d=11  1...e6 2.a3 Ne7 3.d5 d6 4.dxe6 fxe6 5.O–O Nbc6 (0:00.32)
[+1.44]  d=10  1...e6 2.O–O h6 3.h3 d6 4.a4 Ne7 (0:00.15)
[±/+1.64]  d=10  1...Nc6 2.O–O g6 3.e5 Bg7 4.d5 Nb4 5.d6 (0:00.08)
[±/+1.62]  d=9  1...Nc6 2.O–O d6 3.d5 Nb4 4.e5 a6 (0:00.01)
[+1.56]  d=8  1...Nc6 2.O–O d6 3.Bc4 Na5 (0:00.00)
[+1.52]  d=7  1...Nc6 2.O–O d6 3.Bc4 (0:00.00)
[+1.52]  d=6  1...Nc6 2.O–O d6 3.Bc4 (0:00.00)
[+1.46]  d=5  1...Nc6 2.O–O (0:00.00)
[+1.52]  d=4  1...Nc6 (0:00.00)
[+1.52]  d=3  1...Nc6 (0:00.00)
[+1.51]  d=2  1...Nc6 (0:00.00)
[±/+1.65]  d=2  1...Nf6 (0:00.00)
Parent - - By sarciness (***) Date 2008-08-10 20:59
Dynamic R3:

[+1.43]  d=13  1...e6 2.O–O d6 3.Be3 a6 4.Rad1 Nd7 5.e5 d5 6.Ng5 (0:05.27)
[+1.52]  d=12  1...e6 2.O–O Ne7 3.d5 Ng6 4.e5 Be7 5.h4 Nxh4 6.Nxh4 Bxh4 7.Qg4 g6 8.Bh6 (0:01.52)
[+1.49]  d=11  1...e6 2.h3 Ne7 3.O–O–O Ng6 4.h4 h5 5.d5 (0:01.09)
[+1.54]  d=10  1...e6 2.O–O Ne7 3.d5 Ng6 4.Be3 e5 (0:00.23)
[±/+1.72]  d=10  1...Nc6 2.O–O a6 3.a3 d6 (0:00.07)
[±/+1.63]  d=9  1...Nc6 2.O–O d6 3.d5 Nb4 4.e5 c6 (0:00.04)
[±/+1.69]  d=8  1...Nc6 2.O–O d6 3.d5 Nb4 (0:00.01)
[±/+1.68]  d=7  1...Nc6 2.O–O d6 3.d5 (0:00.01)
[±/+1.68]  d=6  1...Nc6 2.O–O d6 3.d5 (0:00.00)
[+1.60]  d=5  1...Nc6 2.O–O (0:00.00)
[±/+1.63]  d=4  1...Nc6 (0:00.00)
[±/+1.64]  d=3  1...Nc6 (0:00.00)
[+1.59]  d=2  1...Nc6 (0:00.00)
Parent - - By sarciness (***) Date 2008-08-10 21:06
Human R3:

[+1.24]  d=14  1...d6 2.e5 Nc6 3.Bc4 Nb4 4.Bxf7 Kxf7 5.Qc4 e6 (0:05.11)
[+1.27]  d=13  1...d6 2.O–O Bg4 3.h3 Bxf3 4.Qxf3 e6 5.Qg3 Nc6 6.d5 Nb4 7.dxe6 fxe6 8.Nb5 (0:01.57)
[+1.24]  d=12  1...d6 2.h3 e6 3.O–O Ne7 4.Be3 h6 5.a3 Nd7 (0:00.58)
[+1.34]  d=11  1...d6 2.h3 e6 3.O–O Ne7 4.a3 Ng6 5.Rad1 (0:00.10)
[+1.36]  d=10  1...d6 2.h3 e6 3.O–O Ne7 4.Rad1 h6 (0:00.07)
[+1.44]  d=10  1...Nc6 2.a3 d6 3.d5 Nb8 4.O–O Nf6 (0:00.04)
[+1.36]  d=9  1...Nc6 2.a3 d6 3.d5 Ne5 4.Nxe5 (0:00.03)
[+1.40]  d=9  1...d6 2.O–O Nc6 3.Bc4 a6 4.h3 Bd7 (0:00.02)
[+1.31]  d=8  1...d6 2.O–O Nc6 3.Bc4 a6 4.Rae1 (0:00.01)
[+1.33]  d=7  1...d6 2.O–O Nc6 3.Bc4 (0:00.01)
[+1.40]  d=7  1...Nc6 2.a3 d6 3.d5 (0:00.00)
[+1.30]  d=6  1...Nc6 2.O–O Nb4 (0:00.00)
[+1.33]  d=6  1...d6 2.O–O Nc6 3.Bc4 (0:00.00)
[+1.49]  d=6  1...e6 2.O–O h6 (0:00.00)
[+1.37]  d=5  1...e6 2.O–O (0:00.00)
[+1.39]  d=4  1...e6 (0:00.00)
[+1.34]  d=3  1...e6 (0:00.00)
[+1.35]  d=2  1...e6 (0:00.00)
[±/+1.61]  d=2  1...Nf6 (0:00.00)
Parent - - By Permanent Brain (*****) Date 2008-08-10 21:09 Edited 2008-08-10 21:14
Thanks! The eval differences are small but visible, and the variations quite different.

rnbqkbnr/pppppppp/8/8/3PP3/2NB1N2/PPPBQPPP/R3K2R b KQkq - 0 1

Combined excerpt: R3 Depth 13

  default: [+1.29]  d=13  1...e6 2.h3 Ne7 3.d5 g6 4.O–O–O Bg7 5.dxe6 fxe6 6.Kb1 Nbc6 7.h4 (0:05.16)
dynamic: [+1.43]  d=13  1...e6 2.O–O d6 3.Be3 a6 4.Rad1 Nd7 5.e5 d5 6.Ng5 (0:05.27)
   human:[+1.27]  d=13  1...d6 2.O–O Bg4 3.h3 Bxf3 4.Qxf3 e6 5.Qg3 Nc6 6.d5 Nb4 7.dxe6 fxe6 8.Nb5 (0:01.57)
Parent - By sarciness (***) Date 2008-08-10 21:19
I like castling queenside as suggested by default. Then you will be more likely to have a position where you're castled on opposite sides, where your development is likely to tell.
Parent - By diskamyl (**) Date 2008-08-10 20:48
All evaluate it clearly better for white with Rybka Human the least, Dynamic the most. they average around 1.50 pawns.
Parent - - By lkaufman (*****) Date 2008-08-11 04:50
I thought all these evals looked too low to my thinking; White is surely winning. But I tried them out with the Monte Carlo at both 5 and 8 ply, and both results came out about +285 Elo points, which according to my formula translates to an eval of +0.95, less than the eval shown by any of the three versions! This suggests that we could give this position as a handicap to a 2700 player like Milov; I don't believe this for a second though. Maybe the MC would show a more decisive advantage at a depth like 18 ply or so that would actually be reached in a real game, but this would take too long to run.
Parent - By Permanent Brain (*****) Date 2008-08-11 05:15 Edited 2008-08-11 05:32
In 2002, I tested a couple of engines in this position, and often Black would win. Among them were two of similiar strength, or within ~50 Elo, Shredder 5.32 and Chess Tiger 14.0. As it turned out, Shredder won with black against Tiger and Tiger won with black against Shredder! :-)

I looked up my article. The experiment contained two games Socrates vs. Yace. Conditions were 8m+3s on P3/700 for Yace and PII/333 for Socrates, with time compensation. I found a backup.

[Event "Experiment (B)"]
[Site "Wien"]
[Date "2002.??.??"]
[Round "?"]
[White "Socrates 3.0"]
[Black "Yace 0.99.56"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/3PP3/2NB1N2/PPPBQPPP/R3K2R b KQkq - 0 7"]
[PlyCount "187"]

7... g6 {1.49/9 18} 8. O-O {67} Bg7 {1.47/10 29} 9. Rad1 {61} Nh6 {1.30/9 16}
10. Bg5 {75} Nc6 {1.27/9 13} 11. d5 {78} Nb4 {1.42/9 18} 12. Bc4 {55} f6 {
1.53/8 14} 13. Bf4 {38} a5 {1.52/9 14} 14. Nb5 {47} d6 {1.31/9 14} 15. Qd2 {58}
g5 {1.24/9 13} 16. Be3 {57} Ng4 {1.45/8 13} 17. a3 {30} Nxe3 {1.26/9 13} 18.
fxe3 {9} Na6 {1.23/9 10} 19. Qxa5 {48} Bd7 {1.21/9 12} 20. Nfd4 {36} c6 {
1.18/8 6} 21. Qxd8+ {41} Kxd8 {1.20/10 8} 22. Nc3 {52} Nc5 {1.15/9 12} 23. Nf5
{29} Bf8 {1.08/9 11} 24. Kh1 {11} Na4 {0.96/10 11} 25. Nxa4 {21} Rxa4 {
0.99/10 8} 26. Bd3 {35} e6 {0.97/9 7} 27. Ng3 {43} Be7 {0.78/9 9} 28. Be2 {21}
b6 {0.74/8 8} 29. Kg1 {26} Kc7 {0.72/8 7} 30. Kf2 {7} Rha8 {0.66/8 7} 31. h3 {
34} c5 {0.51/8 8} 32. Bg4 {32} Bb5 {0.64/9 9} 33. Rh1 {8} e5 {0.70/9 8} 34. h4
{28} Rh8 {0.79/9 10} 35. hxg5 {20} fxg5 {0.70/9 6} 36. Nf5 {6} Bf8 {0.70/9 5}
37. Kf3 {9} h6 {0.70/9 6} 38. Bh5 {6} Rg8 {0.74/9 8} 39. Rd2 {18} Bd7 {0.72/8 6
} 40. g4 {21} Bg7 {0.74/10 7} 41. Bf7 {32} Bxf5 {0.73/10 7} 42. gxf5 {22} Rh8 {
0.84/10 7} 43. Bh5 {17} Rf8 {0.76/10 8} 44. Ra1 {19} Kd7 {0.59/9 5} 45. b3 {18}
Raa8 {0.69/10 5} 46. c4 {17} Bf6 {0.75/11 7} 47. Bg6 {20} Rh8 {0.66/10 7} 48.
Rh2 {13} Bg7 {0.81/10 5} 49. Bh5 {6} Ke7 {0.81/10 7} 50. a4 {
#Der Rest ist Trübsinn. 5} Kd7 {0.86/11 7} 51. Rf1 {11} Raf8 {0.86/11 5} 52.
Ke2 {17} Ke7 {0.89/11 5} 53. Kd2 {18} Rf6 {0.91/12 8} 54. Kd3 {24} Ra8 {
0.89/11 7} 55. Rg2 {12} Kd7 {0.88/11 4} 56. Rgf2 {11} Bf8 {0.89/12 6} 57. Rh2 {
18} Bg7 {0.89/12 6} 58. Rh3 {16} Rg8 {0.90/11 5} 59. Rhf3 {21} Ra8 {0.89/12 8}
60. R1f2 {16} Ke7 {0.89/11 4} 61. Rh3 {5} Kd8 {0.90/12 6} 62. Bg4 {16} Kd7 {
0.90/13 7} 63. Rhf3 {15} Rg8 {0.92/12 7} 64. Rf1 {10} Rh8 {0.90/12 4} 65. Rh3 {
11} Rhf8 {0.92/12 7} 66. Rh2 {12} Ra8 {0.90/11 4} 67. Rhf2 {10} Ke7 {0.90/13 7}
68. Bd1 {6} Rb8 {0.90/11 5} 69. Be2 {13} Ra8 {0.90/13 7} 70. Bg4 {8} Kd7 {
0.90/13 7} 71. Bd1 {10} Bf8 {0.90/11 4} 72. Bh5 {13} Bg7 {0.90/13 6} 73. Rf3 {
14} Bf8 {0.90/11 5} 74. Rh3 {9} Be7 {0.90/12 5} 75. Bg6 {12} Bf8 {0.90/12 4}
76. Rg3 {5} Bg7 {0.89/11 3} 77. Bh5 {6} Rff8 {0.90/11 4} 78. Rf2 {15} Rf6 {
0.90/12 4} 79. Rgf3 {10} Ke7 {0.90/12 3} 80. Rf1 {7} Rff8 {0.90/12 4} 81. R3f2
{4} Bf6 {0.90/12 6} 82. Bd1 {13} Rh8 {0.90/11 4} 83. Rh2 {11} Bg7 {0.90/12 5}
84. Bh5 {5} Rhf8 {0.90/12 6} 85. Be2 {12} Rf6 {0.90/11 3} 86. Rhf2 {9} Kd7 {
0.90/12 4} 87. Bh5 {13} Bf8 {0.90/11 5} 88. Bd1 {8} Be7 {0.90/13 5} 89. Bg4 {5}
Raf8 {0.90/13 6} 90. Ke2 {8} Rh8 {0.90/13 4} 91. Rh1 {7} Rg8 {0.91/13 6} 92.
Kf3 {7} Bf8 {0.90/11 3} 93. Rhf1 {8} Rh8 {0.89/12 3} 94. Bh5 {6} Rg8 {0.89/11 3
} 95. Ra2 {7} Be7 {0.90/11 5} 96. Rg2 {5} Rh8 {0.75/10 4} 97. Rh2 {8} Bd8 {
0.70/10 3} 98. Bg6 {6} Be7 {0.85/10 4} 99. Rg1 {4} Rff8 {0.65/11 4} 100. Rgh1 {
16} Rd8 {0.00/13 3} 1/2-1/2

[Event "Experiment (B)"]
[Site "Wien"]
[Date "2002.??.??"]
[Round "?"]
[White "Yace 0.99.56"]
[Black "Socrates 3.0"]
[Result "1-0"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/3PP3/2NB1N2/PPPBQPPP/R3K2R b KQkq - 0 7"]
[PlyCount "126"]

7... d6 {0} 8. e5 {1.54/9 33} Nc6 {116} 9. exd6 {1.49/8 15} cxd6 {80} 10. d5 {
1.43/9 25} Nb8 {43} 11. O-O {1.66/8 15} Nf6 {47} 12. Rae1 {1.61/8 17} Nbd7 {69}
13. Bf4 {1.80/8 11} Nb6 {38} 14. Ne4 {1.65/8 14} Nxe4 {35} 15. Qxe4 {1.53/9 14}
g6 {40} 16. Bg5 {1.71/7 7} h6 {31} 17. Bf6 {1.83/8 7} Rg8 {30} 18. Re2 {
1.59/8 13} Bf5 {56} 19. Bb5+ {1.52/8 13} Bd7 {34} 20. Nd4 {1.53/8 7} Bxb5 {83}
21. Nxb5 {1.40/9 8} a6 {6} 22. Nc3 {1.55/9 7} Nd7 {69} 23. Bd4 {1.50/9 7} Rc8 {
47} 24. Rfe1 {1.73/9 9} Rc4 {45} 25. Qd3 {1.58/8 12} Rc8 {18} 26. Ne4 {
1.82/8 10} Qc7 {36} 27. Qb3 {1.82/9 11} f5 {51} 28. Nf6+ {1.82/9 10} Nxf6 {22}
29. Bxf6 {1.74/9 8} Kf7 {33} 30. Bxe7 {1.96/9 11} Re8 {27} 31. Bxf8 {1.56/10 11
} Rxe2 {4} 32. Rxe2 {1.56/11 10} Rxf8 {4} 33. Qe3 {1.55/10 6} Kg7 {24} 34. Qd4+
{1.54/9 8} Kh7 {28} 35. c3 {1.56/9 5} Qd8 {20} 36. Re6 {1.60/9 6} Rf7 {20} 37.
h4 {1.76/9 7} b5 {30} 38. Qb4 {1.75/10 10} Rf6 {21} 39. Qa3 {1.59/11 9} Rxe6 {7
} 40. dxe6 {1.85/11 9} Qe7 {4} 41. Qxa6 {2.29/10 7} Qxe6 {4} 42. Qxb5 {
2.32/10 9} Qxa2 {6} 43. Qd7+ {2.37/10 7} Kg8 {6} 44. Qxd6 {2.38/11 9} Kf7 {4}
45. Qc7+ {2.46/10 8} Kf6 {13} 46. Qc6+ {2.48/10 8} Kf7 {3} 47. b4 {2.53/9 6}
Qe2 {22} 48. g3 {2.79/9 6} Qe4 {20} 49. Qxe4 {4.01/10 8} fxe4 {8} 50. c4 {
4.52/12 8} Ke6 {17} 51. b5 {5.05/11 8} Kd6 {14} 52. b6 {5.88/11 8} Kc6 {14} 53.
c5 {6.09/11 6} Kb7 {3} 54. Kf1 {6.61/11 7} Kc6 {13} 55. Ke2 {8.44/11 7} Kb7 {4}
56. Ke3 {11.21/10 7} Kc6 {3} 57. Kxe4 {12.00/10 7} Kb7 {8} 58. f4 {14.17/10 7}
Kc6 {15} 59. Ke5 {14.62/9 4} Kd7 {12} 60. Kf6 {14.97/9 6} Kc6 {23} 61. Kxg6 {
14.97/8 7} Kb7 {3} 62. f5 {19.40/9 4} Kc6 {12} 63. f6 {22.64/9 7} Kxc5 {10} 64.
b7 {22.70/10 6} Kc6 {10} 65. b8=Q {22.85/10 6} Kd5 {17} 66. Qb3+ {#5/9 4} Ke4 {
21} 67. f7 {#4/7 2} h5 {5} 68. f8=Q {#3/5 0} Kd4 {6} 69. Kxh5 {#2/4 0} Ke4 {6}
70. Qf4# {#1/4 0} 1-0
Parent - - By Arkansaw (***) Date 2008-08-12 05:31
What is the formula used for converting Monte Carlo results to evals?
Parent - By TGO (**) Date 2008-08-10 20:11
probably rybka thinks kinda like alekhine here. The idea behind alekhine defence is to let white advance the pawns and then attack his strecth center, it think this is similar.

My opinion is that black has to be careful, but can't really be much worse.
Parent - By Kappatoo (*****) Date 2008-08-10 20:18
Besides what has already been said, white has some nasty weaknesses on the white squares, and no white-squared bishop to defend them. I'm really not sure I share Nimzowitsch's judgment here.
Parent - - By sarciness (***) Date 2008-08-10 20:24
What is the title of the chapter you're quoting from?
Parent - By diskamyl (**) Date 2008-08-10 20:47
it's on page seven, on chapter "center and developement". Thanks for the replies, the light square weakness wasn't one I had thought about.
Parent - - By grolich (***) Date 2008-08-10 20:49
I've looked at the position for a while and I have to say I partially agree with Rybka.

You'll find many such cases of contradiction between Nimzowitsch's examples and Rybka.
What many people fail to realize is that you'll find many contradictions between his examples and modern GMs.

Don't believe me? just grab the book "Secrets of modern Chess strategy - advances since Nimzowitch" by  John Watson.

He actually brings a few examples in "My System" which were plain wrong. Not that nimzo didn't know how to play Chess, just that when trying to explain the basics to other players, many of whom are beginners, he tended to oversimplify matters, and in his examples, he tries to take the principles too far.

He states that you simply cannot evaluate an opening position based solely on tempi count.
That may sometimes work, but at other times may yield completely wrong answers.
Principles are only used as guidelines of what to look at first. They can't be used as an all knowing tool for selecting moves without calculating anything at all.

As for my own opinion of this position and Rybka's analysis of it:

In this position, for example, you claim white is ahead by 4 tempi? interesting claim, after
7...d6, how is white going to maintain that 4 tempi lead?

If he takes (8.exd6 Bxd6), he just lost 2 of these tempi:) Only two tempi ahead after his next move, and black, on an open board, has the two bishops, safe king after he castles imminently, and free piece development.

So white tries to maintain the lead by 7...d6 8.Nf3 instead
But after 8...dxe5 9.Nxe5, can white really be thought of as holding on to the same four tempi? the same piece made two moves, before other pieces were developed, didn't nimzowitch had something bad to say about that too?

Anyway, after 8.Nf3 Bd6, black not only has the option of taking that knight with the bishop at will, reducing the dreaded "tempi lead" and castling his king to safety, which would lead to an approximate equality I believe (with maybe white a bit better, I wouldn't know).

That's why it cannot be said that white has any real advantage in the original position.

However, black does not have to exchange, what if he just decides to leave the bishop and knight there?
Black's king will find safety  when he castles in the next move or 2, he has free piece development, and two bishops on an open board.

Which is more important, Two bishops, or a little bit of development? Hard to answer.
This is just a very difficult position to evaluate.

Another factor: after the exchange of central pawns, you have the case of the isolated queen pawn. May become a weakness if white can't find a useful way to use it, so just exchanging bishops may help black in other ways.

Bottom line:

You can't evaluate a Chess position based solely on a principle (or a number of them),
just as there can be a tactical shot that wins material that overturns the evaluation, there can be another form of tactic, more subtle in nature, that wins no material, but just changes the position enough to negate what the principles claimed a couple of moves ago. (Notice, I never went more than 3 moves in the analysis, it became more complicated almost instantly).

(personally, I believe the position is equal, and after the ...d6 line is played, whichever side a strong player would prefer is a matter of style).
Parent - - By diskamyl (**) Date 2008-08-10 21:12
Thanks for the deep evaluation, grolich, yours is probably better (and objectively deeper) than Nimzowitsch's.
Parent - - By grolich (***) Date 2008-08-10 22:18 Edited 2008-08-10 22:27
Sarcasm will get you nowhere (luckily, nowhere good OR bad).

However, always believing that if a very strong player gives an evaluation, it is much closer to the "objective truth" of the position, is a very bad way to improve.

"My system" is a great book, if used correctly. It is meant to help you with "how to think" and "what to think about". However, considering each and every positional evaluation in it as "correct" will not help your Chess greatly, nor is it true.

For one thing, the first part of the book is for real beginners, who have no idea how to look at the game, and for them, the general idea is far more important than anything else. Just get a bit of order into their games.
Some evaluations written there are considered outdated (a few GMs pointed out a couple of examples from "My system" during the last few years). That may be just because Nimzovitch didn't consider the positions deeply enough because they were meant only for that purpose.

Or because opening knowledge was still developing? hard to tell.
Moreover, what I mentioned about tempi counting is a known problem with Nimzo's book. It's a bad way to estimate any position by just counting tempi.

It gets worse. There is a part of the book which handles the concept of "overprotection", as Nimzowitch calls it.
Apparently, most modern GMs (GM John Watson alludes to the fact that it may be "ALL modern GMs", but I hate the "A" word) simply agree that Nimzovich was just wrong on that point. And yet he brings many positions and evaluations which are today considered wrong. How do you handle this discrepancy? Why was Nimzovich's Evaluation that "off"? the answer is, that as good as he was, he was still a human, and prone to dogmatism.

(By the way, the fact that you cannot evaluate an opening position based on tempi count was quoted originally from Alekhine, who ridiculed the "counting tempi" approach, and half jokingly gave an opening line where one side is 5 tempi ahead and in an awful position. I'll try to find that reference and post the position here).

If you only listen to GMs, and not to people whose opinions are "definitely not as objective or deep", I would recommend the book I mentioned in my previous post:
"Secrets of modern Chess strategy, advances since Nimzowitch" by GM John Watson.
Many examples of evaluations and ideas considered both from Nimzowitch and others of his time, and which are known to not hold true today are mentioned, including some from "My system".

Actually, I believe THIS was one of the positions John Watson was referring to as being wrong from "My system" (or another where there is a very large tempi lead. but it may very well be this one). Can't remembered what he suggested here.

However, what I would prefer, if you don't mind, is that you just offer an idea, or line, or move for white which challenges my suggestion.
That's a more interesting way for both of us to learn.

In my opinion, treating the great players and teachers of the world as deities who are "always right", or at least "always MORE right than lesser mortals", and whose opinions should be treated as correct and used to judge others is a sure way to stop or delay your own progress (in my opinion). Then you become a worshiper instead of a student. And deities make lousy teachers.

Thy were great, but they were also human, which inevitably means - made many mistakes, some of them serious ones.
Parent - - By diskamyl (**) Date 2008-08-10 22:53
grolich, I wasn't being sarcastic at all. I agree completely with your last post.  Actually the purpose of this thread is to challenge Nimzowitsch's ideas, not the opposite.
Parent - - By Permanent Brain (*****) Date 2008-08-10 23:15
We need to distuingish:

1. Ideas which may be good or bad, right or wrong.

2. Examples which may be good or bad, correct or incorrect.

For example, I think the concept of tempo is a good one, basically. We should develope fast, not waste time, (usually) not move a piece twice in the opening... Of course, it is just one concept, or principle for the opening and there are several others. It would be an exaggeration trying to base every move of every opening just on the concept of tempo.

If we would try to refute complete ideas, not just single example positions, then we will have the problem that we have no other arguments for the refutation of such ideas than other example positions. So, a discussion would arise which set of examples is more representative, those which support the idea, or those where the idea seems useless or fails.
Parent - By diskamyl (**) Date 2008-08-10 23:23
I guess that's right, "challenging ideas" is a bad statement.
Parent - By grolich (***) Date 2008-08-10 23:29
Ah, sorry, I misunderstood.

I was referring to some known issues about the book (As in the diagram you have given).

Most of nimzovitch's idea's (actually, probably almost all of his ideas, except just a couple) were a great contribution to Chess theory.
Other than the oversimplistic nature of the first part of the book (which actually makes some sense, considering the level it was aimed at), and the one little issue of overprotection in the second part (which is MUCH deeper, and on an entirely different level than the first part), you'd be hard pressed to challenge his ideas.

There will be a few examples which are wrong, no doubt.
and 1 or 2 strange ideas.

but I believe most of the ideas still hold true to this day.

If you want to find all incorrect examples, then that's another story, which may actually be a good idea. Who knows, you may come across things no one has found out to this day:)
Parent - - By Kappatoo (*****) Date 2008-08-11 08:57
I agree with you on most parts.
By the way, Watson is not a grandmaster. He is an IM, and not a particularly strong one.
Parent - By grolich (***) Date 2008-08-12 04:38
I stand corrected.
I mixed the titles with GM John Emms.

These two Johns have confused me before:)
Parent - By likesforests (**) Date 2008-08-12 18:10 Edited 2008-08-12 18:14
> If you only listen to GMs, and not to people whose opinions are "defihnitely not as objective or deep"... I believe THIS was one of the positions John Watson was referring to as being wrong from "My system"

Watson states his claim that Nimzowitsch was wrong about this position in chapter 1 or 2. Alas, it's IM Watson, so if one holds such beliefs they're missing out on many good books! Luckily I'm open to wisdom from GMs, IMs, NMs, and even fellow amateurs so I can enjoy both My System and Secrets of Modern Chess Strategy.  :)

This position doesn't refute the idea of tempo-counting so much as it's a bad (controversial) example, and perhaps because Nimzowitsch didn't value the bishop pair as highly as modern players (he didn't have the advantage of Kaufman's excellent statistical evaluation of material imbalances) that he didn't see that as a strong counter-balancing factor. He's a bit dogmatic, but he doesn't advocate following rules to the exclusion of all reason. As early as chapter one he chides the play of an amateur who follows his rules a bit too literally.

>  For one thing, the first part of the book is for real beginners, who have no idea how to look at the game

I agree that Part I is not complex. At the same time, I don't agree it's only for "real beginners" unless we define beginners as players below 2100. There's enough "meat" that I can spend hours contemplating each section. It's not enough to memorize, "Rooks belong on open files." We must know how to open them, how to control them, and how to evaluate how much that control is worth. If a file has no penetration points it may be useless.

> Alekhine, who ridiculed the "counting tempi" approach, and half jokingly gave an opening line where one side is 5 tempi ahead and in an awful position.

There are many positions where winning a queen loses the game, but winning a queen is still generally a good thing and thus one of the first lines we should investigate.

> "My system" is a great book, if used correctly. It is meant to help you with "how to think" and "what to think about". However, considering each and every positional evaluation in it as "correct" will not help your Chess greatly, nor is it true.

Absolutely! Subjecting his examples to critical analysis, as diskamyl is going, is a great way to make the knowledge truly ours. In most cases a top player like Nimzowitsch will have a better thought process than amateurs and we should be trying to learn from that. But of course modern opening theory, Rybka, other players, and sometimes our own minds can offer interesting alternatives and improvements.
Parent - - By diskamyl (**) Date 2008-08-10 21:09
(I just would like to remind I'm not favoring "my system" over rybka in this thread, probably rybka will be right in most of the contradictions).

the second contradiction: after  1. e4 e5 2. d4 exd4 3. c3 Nf6 4. e5 Ne4 5. Bd3 d5 6. cxd4
rnbqkb1r/ppp2ppp/8/3pP3/3Pn3/3B4/PP3PPP/RNBQK1NR b KQkq - 0 6

Nimzowitsch says 6...Bb4+ is illogical, becase 7.Bd2 would force an exchange with loss of tempi. However Rybka Human does play 6...Bb4+. after 7.Bd2 however, probably Nimzowitch didn't consider 7...Nxd2! 8.Nxd2 Nc6 and now Black has the two bishops, pressure on d2 knight, and according to R Human, it's even slightly better here.
Parent - By sarciness (***) Date 2008-08-10 21:24
My System was written in the 1930s. Almost all opening theory from that era has been superceded or improved on. Nimzowitsch lived in a time where many of the leading players thought that fianchettoing bishops was 'ugly'.
Parent - - By Permanent Brain (*****) Date 2008-08-10 21:45
I found your example in the german edition, which is a reprint of the edition from 1925/26. I checked how Rybka 2.3.2a handles this, at medium depths. At first, I let analyse in 2 variation mode after 6...Bb4+ and R232a preferred 7.Bd2 over 7.Nd2, only by a small margin. I went back one ply, and up to depth 16, she preferred 6...Nc6 over 6...Bb4+, but then:

Analysis by Rybka 2.3.2a 32-bit (at depth 16):

1. =  (-0.03): 6...Bb4+ 7.Bd2 Nxd2 8.Nxd2 Nc6 9.Ngf3 Bg4 10.Bb5 0-0 11.Bxc6 bxc6 12.0-0 Qd7 13.h3
2. =  (0.16): 6...Nc6 7.Ne2 Bb4+ 8.Nbc3 Bg4 9.Bxe4 dxe4 10.0-0 0-0 11.Qc2 Bxe2 12.Nxe2 Nxd4 13.Qxe4

Now, I entered the continuation Nimzowitsch gave: 6...Nc6 7.Nf3 Bg4, and did some interactive analysis. In fact, it seems that N.'s remark to 7...Bg4 "threatens pawn d4" is plain wrong, due to combinational circumstances... :-) R232a would probably again prefer 7...Bb4+ instead of Bg4. Also, after a while 7.Nf3 is discarded in favour of 7.Ne2:

Analysis by Rybka 2.3.2a 32-bit (after depth 14):

1. =  (0.00): 7.Ne2 Bb4+ 8.Nbc3 Bg4 9.Bxe4 dxe4 10.0-0 0-0 11.Nxe4 Bxe2 12.Qxe2 Qxd4 13.Bf4 Nxe5
2. =  (-0.03): 7.Nc3 Nxd4 8.Nf3 Nxf3+ 9.Qxf3 Nxc3 10.bxc3 Qd7 11.0-0 Bc5 12.Re1 0-0 13.Rb1 Qc6

Surprisingly, Black does not need to forget Bg4 completely because after 7.Ne2 Bg4, 8.f3 is refuted by 8...Bxf3! and if 9.gxf3 Qh4+ etc. But White can play 8.O-O and has the pawn fork threat at f3 now. Now it gets complicated. :-D

I remember The System as a very good, important book. Maybe some of N.'s examples were less well chosen and more doubtful than others.
Parent - By diskamyl (**) Date 2008-08-10 21:58
There are many tactical mistakes, right from the beginning. I'm not even bothering to post them.
Parent - By sarciness (***) Date 2008-08-10 21:22
Default Rybka 3 thinks black is slightly better:

[-0.28] d=15 1... d6 2. Nf3 dxe5 3. Nxe5 Bd6 4. O-O O-O 5. Nc4 Nc6 6. Be3 Nb4 7. Qd2 Be7 8.a3 Nd5 9. Nc3 Bf5 10. f3 (0:07.33)

I'm with Nimzowitsch!
Parent - - By Felix Kling (Gold) Date 2008-08-10 21:54
Nimzowitsch's my system is for sure a great book, but there are of course mistakes in it.

The example you give (it's at the beginning of the book I think) is one of them, it's surprising that Nimzowitsch shows that stupid rule of counting tempi in the opening, since in the rest of the book he shows deep understanding of chess. So it's quite ironic that he begins his book with ti.

Also the "overprotection" thing is strange, see the immortal overprotection game :) (I think it was created by Euwe or Kmoch:
[Event "Composition"]
[Site "Copenhagen"]
[Date "1927.??.??"]
[EventDate "?"]
[Round "?"]
[Result "1-0"]
[White "Aron Nimzowitsch"]
[Black "Systemsson"]
[ECO "C00"]
[WhiteElo "?"]
[BlackElo "?"]
[PlyCount "48"]

1.e4 {AN INGENIOUS EXAMPLE OF MY SYSTEM by Aaron Nimzowitsch -
Anderssen started the sacrificial style, Morphy and Gruenfeld
the pure attacking style, Steinitz the positional style,
Tarrasch the scientific style, Lasker the style of styles,
Capablanca the mechanical style, Alekhine a style as brilliant
as sunlight. But it is a generally known fact that originality
and modernism were introduced by me as my own personal
inventions and enthusiastically imitated (without being fully
understood) by the whole world of chess. For the ridiculously
small sum of ten marks, the reader can confirm all this in my
monumental work, My System, published by B. Kagan. Before my
time, chess was so naive and undistinguished! One or two
brutal opening moves, each one involving a vulgar, obvious
threat, a common, banal sacrifice, a painfully elementary,
bestially raw checkmate - such, more or less, was the course
of chess games before my heyday set in. Then I appeared on the
scene and the chess world paid heed. The hegemony of matter
was shattered at a stroke and the era of the spiritual
began. Under my creative guidance, the chessmen, hitherto
nothing but highwaymen, pirates and butcher boys, became
sensitive artists and subtle instruments of immeasurable
profundity. But why waste words !--accompany me, dear reader
to the dizzy heights of the following game.} e6 2.h4! { My
very oldest and latest thought in this opening. To the chess
addict nurtured on spineless convention, this move comes like
a slap in the face--but calm down, dear reader; after all, you
cannot be expected to understand such moves. (Forgive me - it
is not your fault, until now no one has opened your eyes and
ears.) Wait just a little while, and there will pass before
you a miracle of overprotection of more than earthly
beauty. (I assume that I rightly surmise that you are quite
familiar with my great theory of overprotection.)} 2...d5 {
Black of course has no suspicion of What is coming and
continues serenely in classical style. } 3.e5! {A move of
elemental delicacy. (We detest, as a matter of principle, such
words as "power" and "strength"; in the first place, such
banal expressions make us uncomfortable; and, in the second
place, we like even less the brutalizing tendency which such
words imply.) Wherein lies the beauty of 3.e5? Why is this
move so strong? The answer is as simple as it is
astonishing. The move is strong because it is weak! Weak, that
is, only in the traditional sense! In reality, that is to say,
it is not the move but the Pawn on e5 that is weak--a
tremendous difference! In former times, it is true, it was
customary to reject any move which created a weakness. Today,
thanks to me, this view is obsolete. For, look, my dear
reader, the fact that the Pawn on e5 is weak obliges White to
protect the Pawn more and more until at last the state of
overprotection arises as it were of itself. But, as we have
seen (cf. My System), overprotection is practically equivalent
to victory. Hence it follows automatically that the "weak"
move, 3.e5, is a certain road to triumph. The rest is more or
less a matter of technique. } 3...c5 { All according to a
famous precedent.} 4.d4 {Here it is quite clear that it is
more profitable for White first to provoke c5 and then play
d4, rather than the other way round, which is the customary
course. For, if White first plays d4, there follows c5 and
White's d-pawn is under attack. But my clever transposition of
moves changes the situation completely. For now Black's c-pawn
is suddenly attacked by White's d-pawn!} 4...cxd4 {What else
can Black do?} 5.h5! {All very clever, original and decisive!
Of course the ordinary run of people who envy my every spark
of genius but cannot follow my line of reasoning for even
three paces, outdo themselves in sneering at me with the
poison-dripping epithet, "bizarre." The text move creates
confusion in the whole Black army and prepares for the
annihilating invasion by the Queen 18 moves later.} 5...Qb6
{Naturally not 5...Nc6 6 Bb5! etc. Why should Black play the
French Defence only to allow the Ruy Lopez Bishop move after
all?!} 6.h6! {An avaricious dullard would never hit on this
deeply conceived Pawn sacrifice.} 6...Nxh6 {After 6...gxh6,
White has an even more comfortable game.} 7.Qh5! {The reason
for this becomes clear after next move.} 7...g6 {Black
threatens to begin a successful siege of the weakling at e5
with Bg7. But White forestalls this.} 8.Qh2! {To every
fair-minded observer, this move must come as a revelation! All
the previous manoeuvres now become clear! White has completed
his development brilliantly and proceeds to overprotect
e5. Against this, Black is helpless.} 8...Nf5 9.Bd3 {Note the
splendid cooperation of White's forces: while the e-pawn and
the King Bishop completely blockade Black's position, the
development of the overprotective forces takes place behind
the broad backs of these sturdy blockaders.} 9...Nc6 10.Nf3
{As a rule this is a routine move. But here it is strikingly
original and as such occupies a place in the treasury of my
intellectual property.} 10...h5 {Old stuff!} 11.b4 {A deep
trap, as will soon become apparent!} 11...Bg7 {How Black must
have rejoiced when he anticipated his formidable opponent in
the occupation of the long diagonal. But...} 12.Bf4! {
bitterly disappointed he must have been to realize that 11.b4
had only been a trap and Bb2 had not been intended at all. The
position of Black's Bishop at g7 is now quite
pointless. 11...Be7 would have been relatively better.}
12...Bd7 13.Nbd2 Rc8 {Black no longer has any good moves!}
14.Ke2! {Again, an extraordinarily deep move. White sees
through Black's plans, and in addition he prepares a
particularly powerful continuation of his strategy of
overprotection.} 14...Nxb4 {Just what White was waiting for.}
15.Ne1! {This was the point of his previous move! Black is now
forced to exchange off the attacking Bishop at d3. But, with
that, even White's King Knight enters the fray with fearful
effect at d3, while the square f3 becomes available to the
Queen Knight. Surely a grandiose piece of strategy. The fact
is that I'm a marvellous player, even if the whole chess world
bursts with envy.} 15...Nxd3 16.Nxd3! {Naturally not 16 cxd3?
which would have been quite inconsistent. The Pawn on c2 is
unimportant, and Black only wastes precious time by capturing
it.} 16...Rxc2 17.Rae1! {White continues his overprotection
without much ado.} 17...a5 {This counterattack has no
punch. Black would naturally like to get a passed Pawn plus a
Rook on the seventh, but it is too late for that.} 18.Kd1!
{Now the menaced Rook must scurry back, for capture on a2
would be much too dangerous.} 18...Rc6! {At last, Black gets
the right idea: overprotecting his Pawn at e6. But it is
already too late.} 19.Re2 Ke7 {Introduced into tournament play
by myself. See note to White's 14th move. The King
overprotects e6.} 20.Rhe1 Re8! 21.Nf3! {Completing the
overprotection of e5 and thus deciding the fate of the
game. Black has no defence. Note the aesthetic effect created
by White's position.} 21...Bf8 {Now Black threatens to
complete the overprotection of e6 by playing Ng7. But White
has prepared a brilliant combination.} 22.g4! {Much stronger
than the obvious Bg5+ etc.} 22...hxg4 23.Qh7! {Now one clearly
realizes the masterly understanding of position which went
into White's eighth move (Qh2!).} 23...gxf3 {Had Black
continued overprotecting by 23...Ng7 there would have followed
24.Bg5+ f6 25.Bxf6+ Kf7 26.Ng5 mate. Black's basic error was
that he started overprotecting much too late.} 24.Bg5# {One of
my best games! I am proud of it if only because Herr
Systemsson is one of the strongest Scandinavian players. The
game made an overwhelming impression on the players and
spectators as well as on my opponent. The game has become
famous in Denmark as "the immortal overprotection game."} 1-0 )
Parent - - By grolich (***) Date 2008-08-10 22:23
Ah, "the immortal overprotection game".

That would be by Kmoch.

I think he was too hard on Nimzo, even if Nimzo was a bit off about 1 of his ideas.
but it's extremely funny.
Parent - By Felix Kling (Gold) Date 2008-08-10 22:25
yes, of course it's funny :) When I read it first I couldn't stop laughing for a while and always when I look at it again it makes me smile :) He perfectly copied Nimzowitsch's style, an excellent parody.
Parent - By sarciness (***) Date 2008-08-13 00:55
Never seen this before, very funny!
Parent - By diskamyl (**) Date 2008-08-10 22:23
3) after 1. e4 Nc6 2. Nf3 Nf6 3. Nc3 d5 4. exd5 Nxd5 5. d4 Bf5 6. a3 g6 7. Bc4 Nb6 8.
Ba2 Bg7 9. Be3
, in
r2qk2r/ppp1ppbp/1nn3p1/5b2/3P4/P1N1BN2/BPP2PPP/R2QK2R b KQkq - 0 9

A.N. has played (and suggests) 9...e5 "!", with the remark "black has thus not played to restrain the d-pawn, but to kill it," where Rybka Human strongly disagrees with 10.d5! ( ...Ne7 11. O-O O-O 12. Re1 Qd6 ) and the pawn is not "killed". (the original game has continued 10.Qe2 0-0 11.dxe5 Bg4.) 
Parent - - By diskamyl (**) Date 2008-08-10 23:00
4) After 1. e4 e5 2. f4 d6 3. Nf3 Nc6 4. Bc4 Be6 5. Bxe6 fxe6 6. fxe5 dxe5, in
r2qkbnr/ppp3pp/2n1p3/4p3/4P3/5N2/PPPP2PP/RNBQK2R w KQkq - 0 7
Nimzowitsch suggests Black is better due to the open files for his rooks (the d and f files) and good developement, but Rybka human doesn't agree.

White's developement is not behind from black's, the f file could well be used for white's advantage as well, and now that the light square bishop is gone, white can safely play d3 and continue developement with natural moves like Be3, Nf6 or d2 and 0-0, where black has to deal with the doubled pawns.
Parent - By Felix Kling (Gold) Date 2008-08-11 14:34
Here Rybka may misevaluate the doubled pawn- same as in 1.e4 d6 2.d4 Nf6 3.Nc3 e5 4.dxe dxe 5.Qxd8 Kxd8 6.Bc4 Be6 7.Bxe6 and so on.

In this position the doubled pawns indeed favor black (controle the center!) and I think Nimzowitsch is right here.
Parent - By rivaldo (***) Date 2008-08-10 23:36
nimzowitsch is wrong: white isn't 4 tempi up. a move like e4-e5 in such a position is not desirable, because it weakens white's centre pawns and centre control. they are not flexible anymore. Qd1-d3 is also not a developing move, because the Q would be slightly better placed on d1. in addition to that, black has the bishop pair and is likely to get this in play in time.
conclusion: white only has the initial bonus of 1 move and is roughly 1 tempo up.
Parent - - By exigentsky (***) Date 2008-08-11 10:18 Edited 2008-08-11 10:42
All the Rybkas consider this position slightly better for Black. This seems strange. There aren't any significant positional problems with White's better developed position (bishop is opened) and he even traded a flank pawn for a center pawn. I suppose that Rybka has a high focus on the bishop pair and the information Black gains from White's less flexible setup.
Parent - By nine castles (**) Date 2008-08-12 19:27
I think developments in chess theory since Nimzo show that white's advantage here is strictly optical. As others have shown, after even a simple move like d6, white is losing his time advantage since the pawn which has moved twice will be traded, probably with another move that lets black develop another piece resulting in a 3 tempo loss; and white has no way to prevent this. IOW, white's position looks good, but after a few more moves have been played, it won't be so good anymore, and as you said black will have the bishop pair. Furthermore white's only bishop is a bad bishop and black has no weaknesses. Overall, not looking good for white.
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka Human vs "My System"

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill