I do not understand the diagrammed position. Do you?

[fen ""]
[title "A. Grischuk vs. V. Kramnik, FIDE Candidates 2018"]
[startply "8"]

1. c4 e6 2. Nc3 d5 3. d4 Nf6 4. Nf3 c5 5. e3 dxc4 6. Bxc4 a6 7. Bb3 b5 8. e4 cxd4 9. Nxd4 Bb7 10. e5 Ne4 11. O-O Nxc3 12. bxc3 Nc6 13. a4 Bc5 14. axb5 axb5 15. Rxa8 Bxa8 16. Nxb5 O-O 17. Qxd8 Rxd8 18. Bf4 h6 19. h4 Ne7 20. Rd1 Rb8 21. Bc4 Bc6 22. Nd4 Be4 23. h5 Nd5 24. Bxd5 Bxd5 25. Be3 Rc8 26. Re1 Bc4 27. Nf3 Ba3 28. Bd4 Bd3 29. Nd2 Bb2 30. Nf1 Rc4 31. Re3 Bf5 32. Ng3 Bh7 33. Kh2 Bc1 34. Re2 Bf4 35. Kh3 Ra4 36. Kg4 Bc1 37. Kf3 Bd3 38. Re1 Bb2 39. Ke3 Bc2 40. Kd2 Bb3 41. Rb1 Ra2 42. Kd3 Ba4 43. Ne4 Ba3 44. Ke3 Bc2 45. Re1 Bb3 46. g4 Bd5 47. Kd3 Be7 48. Rb1 Ra8 49. f4 Bh4 50. Rh1 Bd8 51. Rb1 Bc7 52. Ke3 Kh7 53. Nd6 Kg8 54. Rb5 Bc6 55. Rb4 Bd8 56. Bb6 Bh4 57. Bd4 Bg2 58. Rb2 Bc6 59. Rb6 Bg2 60. Rb2 Bc6 61. Rb3 Bg3 62. c4 Bh2 63. f5 exf5 64. Nxf5 Kh7 65. Bb2 Re8 66. Nd6 Re7 67. Rb8 f6 68. Rc8 Bh1 69. Nf5 Rb7 70. exf6 Rb3+ 71. Bc3 gxf6 72. Kd2 Bf4+ 73. Kc2 Rb8 74. Rxb8 Bxb8 75. Ne7 Be4+ 76. Kd2 Bf4+ 77. Ke2 Kg7 78. Nf5+ Kf7 79. Bd2 Be5 80. Bxh6 Ke6 81. Be3 Bxf5 82. gxf5+ Kxf5 83. h6 Kg6 84. c5 f5 85. Kf3 Kf7 86. Bf4 Bd4 87. c6 Ke7 88. c7 Kd7 89. h7 Kc8 90. Ke2 Kd7 91. Kd3 1-0

What is the idea behind 4... c5? Why form a hostile bloc of two black against two white pawns on the c and d files? Once the bloc is formed, how does one decide whether

  • to capture toward the c file (as Black soon does),
  • to capture toward the d file (which, inscrutably, neither player wishes to do), or
  • not to capture at all?

This kind of position is opaque to me. I do not understand it. I am aware of no principles by which one can treat it. I perceive no relevant pattern in a bloc of four pawns. Please advise.

I would gladly read an answer at any level but would find most useful an answer pitched at about Elo 1350 or USCF class C.

(Source for the game: Chess.com.)

2 Answers 2


Black's idea is to try to neutralize White's center pressure by making the situation appear equal. One idea for both sides is to take the c-pawn after their opponent has moved the King's Bishop. For example, if White plays Bd3, Black may play ...dxc4, causing White to lose a tempo by having to move the Bishop again.

This mutual dilemma makes development tricky, as both sides need to develop the King's Bishop in order to castle. What often happens is that White resolves the tension with his c-pawn and Black's d-pawn, and then both sides continue with normal development. The main line from the diagrammed position (after 4...c5) is as follows:

  1. cxd5 (resolving the tension) exd5 6. g3 Nc6 7. Bg2 Be7 8. 0-0 0-0

At this point, Black's c-pawn and White's d-pawn tend to be exchanged soon. This results in a position where Black has an isolated pawn, but in return gets slightly more active pieces.

One of the main goals (for Black) of the Tarrasch is to reach such an isolated d-pawn position, since in return for an isolated pawn Black gets good activity.

In the game you posted, Kramnik decided to take on c4 before White moved his King's Bishop. If he wanted to play a pure Tarrasch setup, he would have waited with ...Nc6 or ...a6. The reason he took on c4 immediately is that he wanted to transpose to a QGA setup.

  • You seem to mix Tarrasch (without Nf3/Nf6) and semi-Tarrasch (as in the question). In any case, your main line is wrong. In case of 5.cxd5, Kramnik would have taken with the Nf6, the main line of the semi-Tarrasch being 5...Nxd5 6.e4 Nxc3 7.bxc3 cxd4 8.cxd4 Bb4+. In case of 5...exd5?, instead of transposing to a Tarrasch with 6.g3?!, White can gain a huge advantage with 6.Bg5!
    – Evargalo
    Mar 21, 2018 at 9:44
  • @Evargalo Not necessarily a huge advantage if Black plays 6...Be6. Also, I forgot about the Semi-Tarrasch so thanks for pointing it out. But I think his question was mainly focused on the Tarrasch pawn structure, not on move-orders. Mar 21, 2018 at 14:06

What is the idea behind 4... c5?

Basically the same ideas as behind playing c4 in the Queen's gambit:

  • challenging the center,
  • potentially exchanging the c-pawn for a valuable central pawn,
  • opening the c file for a rook,
  • giving the knight a square on c6 (which is often a better square for the knight if the pawn has moved to c5 already)
  • potentially threatening to leave white with an isolated queen pawn (see below)

Why form a hostile bloc of two black against two white pawns on the c and d files?

Why not? "Hostility" is not something you should be afraid of when playing chess.

Once the bloc is formed, how does one decide whether ...

There is no strict rule here and various captures and no-captures have been tried and tested by good players. A typical motive you need to be aware of in these positions is the isolated queen pawn (IQP), i.e. a player might be left with a pawn on the d-file and without any pawns on the c and e files which could support it.

An IQP leads to somewhat asymmetric play, with one player trying to block and put pressure on this potentially weak pawn, while the other player (the one with the IQP) is trying to take advantage of his better piece play that is a result of the pawn structure and also has ideas related to advancing the IQP.

If there is no IQP on either side, the positions resulting from this opening are often very symmetric, potentially boring, and it is difficult to find a good plan in them.

So I'd say the decision on how to capture is basically determined by what kind of game you want:

  1. with an IQP; asymmetric, unbalanced game with clear plans for both sides
  2. without an IQP; symmetric, equal position, with unclear plans
  • 1
    That makes sense. So, if I understand, in the diagrammed game, Black initiates the exchange after White has declined an opportunity to initiate it, because White wants active pieces whereas Black prefers a sound pawn structure toward the endgame. Neither side wants symmetry. Tarrasch's four-pawn bloc is less an attempt to dictate a particular outcome than it is to allow one's opponent to commit to an asymmetry of the opponent's choosing. I too prefer asymmetry but had heretofore tended toward hypermodern means to achieve asymmetry as Black. Tarrasch affords a classical alternative.
    – thb
    Mar 20, 2018 at 20:16
  • @thb Basically yes, though, as in any chess opening, there are many subtleties here, which are way beyond the average chess player. Still I believe that the IQP issue is very topical in these lines. Personally I also prefer the hypermoderns for their asymmetry. Mar 21, 2018 at 6:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.