Home > GM Repertoire > GM 16 French – Killing the King’s Indian Attack

GM 16 French – Killing the King’s Indian Attack

Here’s another preview of Grandmaster Repertoire 16 by Emanuel Berg, which is published today. It consists of a few lines I pulled from different parts of the relevant chapter, so this should not be considered an excerpt from the book, which is a lot more detailed.

I decided to show you a glimpse of Emanuel’s recommendation against the King’s Indian Attack, which is always popular at club level. I have focused on a couple of lines involving the author’s own games. Throughout the book and indeed most of the series, he has recommended lines that he himself plays, and this chapter is no exception.

1.e4 e6 2.d3 d5 3.Nd2 c5 4.Ngf3

4.g3 is an alternative move order, with which White may try to avoid Black’s recommended system. The point is that, after 4…Nc6 5.Bg2, Black is unable to develop the bishop on d6 due to the hanging d5-pawn.

Therefore Emanuel prefers 4…Bd6, when 5.Bg2 Ne7 6.Ngf3 Nbc6 reaches the desired set-up. 5.Qg4!? is an interesting way to deviate, but Emanuel looked at it carefully and found an improvement for Black, which will be revealed in the book.

4…Nc6 5.g3 Bd6 6.Bg2 Nge7 7.0–0 0–0 8.Re1 Qc7

This set-up a great practical choice against the KIA. Black prevents the standard plan of e4-e5, and seeks to gain space. If White is not careful, he may easily find himself in a passive position resembling a reversed King’s Indian Defence gone wrong.

9.c3 Bd7

The main line continues:


The following is mentioned in a note: 10.Qc2 Rac8 11.a3 (11.d4? loses a pawn after 11…cxd4 12.cxd4 Nxd4!) 11…b5 12.Nf1 d4 13.cxd4 cxd4 14.Bd2 e5 Black had a stable position and a small plus in Atarov – Berg, Internet 2004. The idea of meeting Nf1 with …d4, gaining space and restricting the knight, is a recurring theme in the chapter.

10…f6 11.a3


Emanuel also gives some brief analysis of 11…a5 and 11…d4!?, so the reader can choose the type of position he likes best.


This is a typical move for the KIA, and exactly the type of move that your opponents might make when they are not sure what else to do. Another example is Nd1-f1-h2. One of the nice things about Black’s chosen set-up is that these standard manoeuvres tend not to work well.

12.b4 is the main line, in which Emanuel found an important novelty for Black. Again, details can be found in the book.

12…b5 13.h5 d4 14.c4 a6 15.Nh4 e5

Black already held a clear advantage in Kinnmark – Berg, Sweden 1999. The note in the book ends here, but I will give a few more moves of the game.

16.b3 Rb8 17.Rb1 Be6 18.cxb5 Rxb5

18…axb5! is stronger.

19.Bf3 Na5 20.Bg4 Qd7 21.Bxe6+ Qxe6 22.Qf3

Black went on to win after 22…Rfb8, but 22…Nxb3! would have been a simpler way to win a pawn for nothing. Emanuel was only rated 2348 at the time, but when you see that the game is dated 1999, you get an idea of his experience and how well he knows the topic under examination.


Categories: GM Repertoire Tags:
  1. March 25th, 2015 at 15:32 | #1

    I have a book which calls this Black setup against the KIA the “Karpov System”, since the whole piece configuration is prophylaxis against White’s e4-e5 advance.

  2. Ray
    March 25th, 2015 at 18:12 | #2

    Interesting recommendation – I guess we’re spoilt for choice, since I also liked the recommendation in Playing the French. Hard times for KIA afficionados 🙂

  3. Alice
    March 25th, 2015 at 20:04 | #3

    It’s very nice to have new complete French repertoire same way as incredible deep King’s indian books from Kotronias. However time to time I have some feeling that authors something overlooked during their hard work on the books. For example I mean Berg chapter 23 in Winawer, where he missed 15.Ng5 (which was lately analysed in Quality newsletter) which changed evaluation to +/= as minimum, now I found at the end of Mar del Plata II book small improvement in the main-main-main line on page 272 – Champion-Van Unen corr. 2009, when author write: since after 42..Qf4 … White won’t be able to block the … black pawn, but in fact after 43.Qxf4 exf4 position is mate in 105 moves 🙂 Of course it is not simple win game this way OTB… 🙂

  4. Stigma
    March 25th, 2015 at 23:50 | #4

    This setup scores well for Black and I’m not aware of any theoretical problem with it, but it’s a bit funny that it’s basically the same piece setup as White’s “Seirawan system” against the KID (Bd3/Nge2). In this French/KIA line Black often plays …e6-e5 eventually, so isn’t he in a sense playing the Seirawan system a tempo or two down? 1 for being Black, 1 for playing …e5 in two moves, and 1 for playing …Qc7 and …Bd7, which are not necessarily the very best squares for the queen and bishop in that structure. But then of course White has usually played Re1, when the rook might well have been better on f1 in a closed KID position.

  5. garryk
    March 26th, 2015 at 13:09 | #5

    Which newsletter are you referring to? I can’t find that analysis in the newsletter file. Question for QC team – it’s possibile to update the file with all the newsletters up to date and/or provide separate files for each newsletter until today? Thanks!

  6. Alice
    March 26th, 2015 at 13:18 | #6

    I mixed things little bit. It was mentioned on ChessPublishing (and not Quality newsletter), game Atabayev-Huzman Yerevan 2014

  7. John Shaw
    March 26th, 2015 at 13:35 | #7


    Thanks, we will certainly check out that mate-in-105 moment. It may be next week before we get a free moment to do that.

  8. John Shaw
    March 26th, 2015 at 13:41 | #8


    Updating the “Updates & Newsletters” link on our main website will certainly happen – next week at latest.

  9. Alice
    March 26th, 2015 at 13:52 | #9

    @John Shaw
    Good luck, I don’t know where is the problem, but this was just one mouse click to check 7-man TB position which appeared on the board..?! I was only surprised how quickly you can get from opening to TB position.

  10. John Shaw
    March 26th, 2015 at 14:24 | #10


    I think my tablebase only goes up to 6 men, so no one-click option for me yet on my machine. But we will get there. And even if it is mate-in-105 what about the 50-move rule!?

  11. Alice
    March 26th, 2015 at 14:37 | #11

    @John Shaw
    There were still 3 black pawns on the board which could be taken during this 105 moves sequence to pass over 50moves rule. I don’t think that it is crucial for something like this excellent book, I mentioned this only as curiosity.

    My idea was based on some experience, what I get in the last months, that sometimes is missed something important in the main lines (where one side, usually black, doesn’t have possibility play different way) and what close whole line (but for example, in mentioned game black has 35..Ra8 with drawing chances)

  12. Ray
    March 26th, 2015 at 19:14 | #12

    It all sounds a little bit scary to me – 43 moves of theory straight into a theoertical endgame win 🙂

  13. Nikos Ntirlis
    March 26th, 2015 at 20:08 | #13

    The amazing thing is that with the 6-men TBases both SF and Komodo insist that this is a draw regardless of the time they are allowed to search the position. Amazing…

  14. wok64
    March 26th, 2015 at 23:19 | #14

    Can’t White just play 4.Qe2 or 5.Qe2 to prevent this Setup or am I missing something? And of course you will need an own System for 1.e4 e6 2.Qe2

  15. Alice
    March 27th, 2015 at 07:41 | #15

    There is mentioned game:

    [Event “ICCF Olympiad 18 Preliminaries – Sectio”]
    [Site “?”]
    [Date “2009”]
    [Round “?”]
    [White “Champion,Chris”]
    [Black “Unen,J.”]
    [Result “1/2-1/2”]

    1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5
    Ne7 9. b4 Nh5 10. Re1 f5 11. Ng5 Nf6 12. f3 c6 13. Be3 Bh6 14. h4 cxd5 15.
    cxd5 Bd7 16. Rc1 a6 17. b5 Bxb5 18. Bxb5 axb5 19. Nxb5 fxe4 20. fxe4 Nxe4 21.
    Nxe4 Bxe3+ 22. Rxe3 Qb6 23. Qb3 Nf5 24. Rcc3 Nxe3 25. Rxe3 Rxa2 26. Nbxd6 Ra1+
    27. Kh2 Qd8 28. Rh3 Rff1 29. h5 Rh1+ 30. Kg3 Rxh3+ 31. gxh3 Qf8 32. Qe3 b5 33.
    hxg6 Qg7 34. Qf2 Qxg6+ 35. Kh2 b4 36. Nf5 Kh8 37. d6 Qf7 38. d7 Qxd7 39. Nfd6
    Rh1+ 40. Kxh1 Qxh3+ 41. Kg1 Qg4+ 42. Kf1 {there game ended in a draw} Qf4 43. Qxf4 exf4 {White mates in
    105:} 44. Ke2 f3+ 45. Kd3 h5 46. Nf7+ Kg7 47. Nfg5 Kg6 48. Nh3 Kf5 49. Nef2
    Kf6 50. Kc4 Ke5 51. Kxb4 Kd4 52. Kb3 h4 53. Kb2 Kc4 54. Kc2 {Lomonosov tables}
    {White mates in 94:} 54… Kb4 55. Kd2 Kc4 56. Ke3 Kc5 57. Ng4 Kd5 58. Kxf3
    Kd4 59. Kf4 Kd5 60. Kf5 Kc5 61. Ke5 Kc4 62. Ngf2 Kc5 63. Nd3+ Kc4 64. Ke4 Kc3
    {Lomonosov tables}{White mates in 84:} 65. Ne5 Kb3 66. Kd5 Kb4 67. Kc6 Kb3 68.
    Kc5 Kc3 69. Nf7 Kb3 70. Nfg5 Kc3 71. Ne4+ Kd3 72. Kd5 Ke2 73. Nef2 Ke1 74. Kc5
    Kd2 75. Kc4 {Lomonosov tables}{White mates in 73:} 75… Ke3 76. Kc3 Ke2 77.
    Kd4 Kf3 78. Kd3 Kg2 79. Ke4 Kg3 80. Ke3 Kh2 81. Kf4 Kg2 82. Kg4 Kf1 83. Kf3
    Ke1 84. Ke3 Kf1 85. Ne4 Ke1 {Lomonosov tables}{White mates in 63:} 86. Nd2 Kd1
    87. Kd3 Ke1 88. Nc4 Kf1 89. Ne3+ Ke1 90. Ng4 Kd1 91. Ng5 Kc1 92. Ne4 Kd1 93.
    Nh2 Kc1 94. Nf3 Kb2 95. Nc5 Kc1 96. Na4 {Lomonosov tables}{White mates in 52:}
    96… Kb1 97. Kd2 Ka2 98. Kc3 Kb1 99. Nb6 Kc1 100. Nd5 Kb1 101. Nf4 h3 102.
    Nh2 Ka1 103. Kb3 Kb1 104. Nd3 Ka1 105. Kc2 Ka2 106. Nb2 Ka1 {Lomonosov tables}
    {White mates in 42:} 107. Nc4 Ka2 108. Kc3 Kb1 109. Kd2 Ka1 110. Kc1 Ka2 111.
    Kc2 Ka1 112. Kb3 Kb1 113. Nd2+ Kc1 114. Kc3 Kd1 115. Nb3 Ke1 116. Kd4 Ke2 117.
    Ke4 {Lomonosov tables}{White mates in 31:} 117… Ke1 118. Ke3 Kd1 119. Kd3
    Ke1 120. Nd4 Kd1 121. Ne2 Ke1 122. Nc3 Kf2 123. Kd2 Kg2 124. Ke2 Kg3 125. Ke3
    Kh4 126. Kf4 Kh5 127. Kf5 Kh6 {Lomonosov tables}{White mates in 21:} 128. Kf6
    Kh5 129. Ne4 Kh4 130. Kf5 Kh5 131. Ng3+ Kh4 132. Ngf1 Kh5 133. Ne3 Kh6 134.
    Kf6 Kh7 135. Nf5 Kg8 136. Ke7 Kh7 137. Kf7 Kh8 138. Kg6 {Lomonosov tables}
    {White mates in 10:} 138… Kg8 139. Ng7 Kf8 140. Kf6 Kg8 141. Ne6 Kh7 142.
    Kg5 Kg8 143. Kg6 Kh8 144. Kf7 Kh7 145. Ng4 h2 146. Nf8+ Kh8 147. Nf6 h1=R 148.
    Ng6# {Lomonosov tables} 1/2-1/2

    Mate sequence doesn’t break 50-moves rule

  16. Nikos Ntirlis
    March 27th, 2015 at 11:11 | #16

    I am deeply comfused….. Both Komodo and SF with the 6-piece syzygy give the game as drawn after 77…Kd2 deviating from the above variation

  17. Michel Barbaut
    April 1st, 2015 at 17:54 | #17

    I’ve just received Berg’s vol.3 : high classs analysis as well as objective ones ! Another must buy !

  18. Ray
    April 1st, 2015 at 19:30 | #18

    @Michel Barbaut
    Great to hear – this one’s on my list as well (which is building up a serious backlog at the moment)

  19. Thomas
    April 2nd, 2015 at 06:42 | #19

    I have to confirm that. I had a look at some of the lines only but those looked really- really – well elaborated. And black is NOT winning in all lines, he takes a very objective view. Outstanding!

  20. John Upper
    June 26th, 2015 at 17:38 | #20

    @Nikos Ntirlis
    Here’s the analysis, starting after 77….Kd2.

    I don’t have the Syzygy TBs, it might be interesting to compare the two side by side.
    You might be able to do this by downloading the free Lomonosov Android app. Link via:

    [Event “Lomonosov TableBases”]
    [Site “tb7.chessok.com”]
    [Date “????.??.??”]
    [Round “?”]
    [White “?”]
    [Black “?”]
    [Result “1-0”]
    [SetUp “1”]
    [FEN “8/8/8/8/3K3p/7N/3k1N2/8 w – -“]

    1. Ng4 Kc2 2. Kc4 Kb2 3. Nh2 Ka3 4. Nf1 Kb2 5. Nf2 Ka3 6. Kb5 Kb3 7. Ne4 Ka3 8. Nc5 Kb2 9. Kc4 Kc2 10. Nh2 Kd2 11. Kd4 Kc2 12. Nf3 h3 13. Nh2 Kd2 14. Nb3+ Kc2 15. Kc4 Kb2 16. Nd4 Ka2 17. Kc3 Ka3 18. Nb3 Ka2 19. Nc5 Ka3 20. Kc4 Ka2 21. Kd3 Kb2 22. Kd2 Ka2 23. Kc2 Ka3 24. Kc3 Ka2 25. Nd3 Kb1 26. Kb3 Ka1 27. Kc2 Ka2 28. Nb2 Ka1 29. Nc4 Ka2 30. Kc3 Kb1 31. Kd2 Ka1 32. Kc1 Ka2 33. Kc2 Ka1 34. Kb3 Kb1 35. Nd2+ Kc1 36. Kc3 Kd1 37. Nb3 Ke1 38. Kd4 Ke2 39. Ke4 Ke1 40. Ke3 Kd1 41. Kd3 Ke1 42. Nd4 Kd1 43. Ne2 Ke1 44. Nc3 Kf2 45. Kd2 Kg2 46. Ke2 Kg3 47. Ke3 Kh4 48. Kf4 Kh5 49. Kf5 Kh6 50. Kf6 Kh5 51. Ne4 Kh4 52. Kf5 Kh5 53. Ng3+ Kh4 54. Ngf1 Kh5 55. Ne3 Kh6 56. Kf6 Kh7 57. Nf5 Kg8 58. Ke7 Kh7 59. Kf7 Kh8 60. Kg6 Kg8 61. Ng7 Kf8 62. Kf6 Kg8 63. Ne6 Kh7 64. Kg5 Kg8 65. Kg6 Kh8 66. Kf7 Kh7 67. Ng4 h2 68. Nf8+ Kh8 69. Nf6 h1=R 70. Ng6# 1-0

  1. No trackbacks yet.

 Limit your comments to