Pawn structures often determine on which side of the board each player tries to attack, but as World Chess’s columnist explains, there are quite a few exceptions.

In many openings, particularly those with fixed central pawn structures, one player often controls one side of the board and the other player dominates the other. For example, when White plays e4-e5 in the French Defense he will usually strive for play on the kingside because his e-pawn gives him control of more space there, while Black will aim for counterplay on the queenside. Something similar occurs in the King’s Indian Defense. Consider the following position:

Classical King's Indian vs. White plays g4
? | ? | Round ? | 13 Jan 2017 | ECO: E99 | *
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. Ne1 Nd7 10. f3 f5 [#]
11. g4 [#]

The usual plan for Black is to throw everything, including the kitchen sink, at White’s king in an all-out attack, while White aims to open lines on the queenside (typically with c4-c5) and to outflank Black before Black’s kingside pressure becomes too much. There are plenty of thrilling, spectacular games that follow that template, but sometimes White flips the script by playing 11.g4. It isn’t White’s main option, but it is an important alternative that Black needs to know and understand. So Black’s standard plan in the Classical King’s Indian isn’t a rule, it is at most a rule of thumb.

Another example is in the Carlsbad pawn structure (White pawns on d4 and e3, Black pawns on d5 and c6, where the White c-pawn has been traded for the Black e-pawn, or the mirror situation in which White’s pawns are on d4 and c3 and Black pawns on d5 and e6). This can arise from multiple openings, including the Exchange Variation of the Caro-Kann Defense. Here’s a representative position from that opening:

Benjamin, Joel vs. Christiansen, Larry Mark
USA-ch/zt | South Bend | Round 12 | 1981.??.?? | ECO: B13 | 1-0
1. e4 c6 2. d4 d5 3. exd5 cxd5 4. Bd3 Nf6 5. c3 Nc6 6. Bf4 Bg4 7. Qb3 Qd7 8. Nd2 e6 9. Ngf3 Bxf3 10. Nxf3 Bd6 11. Bxd6 Qxd6 12. O-O O-O 13. Rae1 Rab8 14. Ne5 b5 [#]

White has a beautifully posted knight on e5, which he will soon reinforce with f4. After that his heavy pieces will migrate to the kingside, hoping to deliver a mating attack. Black, who has just played …b5, is trying to start a minority attack with …b5-b4. He hopes to break up and weaken White’s queenside, and to exploit it and break through before White’s attack succeeds. The following games illustrating the basic template.

Benjamin, Joel vs. Christiansen, Larry Mark
USA-ch/zt | South Bend | Round 12 | 1981.??.?? | ECO: B13 | 1-0
1. e4 c6 2. d4 d5 3. exd5 cxd5 4. Bd3 Nf6 5. c3 Nc6 6. Bf4 Bg4 7. Qb3 Qd7 8. Nd2 e6 9. Ngf3 Bxf3 10. Nxf3 Bd6 11. Bxd6 Qxd6 12. O-O O-O 13. Rae1 Rab8 14. Ne5 b5 [#]
15. a3 a5 16. f4 b4 17. axb4 axb4 18. Qd1 bxc3 19. bxc3 g6 20. Re3 The chances are equal but not in a static, boring way. White's c-pawn is a serious liability that cannot be safely liquidated by playing c4, as that would doom his d-pawn. So White needs to make something happen on the kingside before his positional problems present permanent problems.
20... Nd7
20... Ne7  )
20... Rfc8  )
21. Qe1 Ncxe5
21... Rb7  )
21... Rb3  )
22. fxe5 Qe7 23. Ref3 f5 24. exf6 Rxf6 25. Rxf6 Nxf6 26. Rf3
26. Qe3  )
26... Re8
26... e5!  )
27. h3 Nd7 28. Bb5 Rf8
28... Qd6  )
29. Re3 Rf6 30. Qg3 Nf8 31. Kh2 Qf7 32. Be2 Rf4 33. Bg4 Rf6 34. Re2 g5 35. Qe5 h6 White's advantage is pretty serious, and the weakness of his c-pawn is purely theoretical.
36. Rb2 Rf4 37. Re2 Rf6 38. Re1 Kg7 39. Ra1 Kg6 40. Ra8?
40. Ra6  )
40... h5? This further weakening of Black's kingside proves fatal.
40... Rf1  )
41. Be2 Rf2 42. Kg1 Kh6 43. h4 Rf5 44. Qh8+ Nh7 45. Bd3 gxh4 46. Bxf5 Qxf5 And White converted his material advantage without any serious trouble.
47. Ra7 Qb1+ 48. Kh2 Qg6 49. Qe5 Nf6 50. Qf4+ Qg5 51. Qxg5+ Kxg5 52. Rg7+ Kf5 53. Kh3 Ne4 54. Kxh4 Nxc3 55. Kxh5 Ne2 56. g4+ Ke4 57. g5 Kxd4 58. Rf7 Ng3+ 59. Kh4 Ne4 60. g6 Nd6 61. Rf8
Zhao, Jun vs. Laznicka, Viktor
Ho Chi Minh City HD Bank op-A 6th | Ho Chi Minh City | Round 4 | 10 Mar 2016 | ECO: B13 | 1-0
1. e4 c6 2. d4 d5 3. exd5 cxd5 4. Bd3 Nc6 5. c3 Nf6 6. Bf4 Bg4 7. Qb3 Qc8 8. Nd2 e6 9. Ngf3 Be7 10. O-O O-O 11. Qc2 Bh5 12. Ne5 Qe8 13. Rae1 Rc8 14. Qb1 Bg6 15. Nxg6 hxg6 16. Nf3 Again White shuffles everything to the kingside. The knight is in touch with both e5 and g5, the rooks are on the e- and f-files, and White will build an attack as best he can. Black isn't as well set up for queenside counterplay as he was in the previous game - or in the next game we'll see - and tries to put out the kingside fires instead.
16... a6 17. Qd1 Qd8 18. h4 Re8 19. g3 Nd7 20. Re2 Nf8 21. Rfe1 Bd6 22. Bc1! Black is cramped, so it makes sense for White to avoid the swap. Further, White's bishop on c1 may be able to participate in the kingside festivities, so for both reasons it makes sense to keep it.
22... Qc7 23. Ng5 Ne7 24. f4 A reasonable decision, though not the only good one available to White. White's Darth Vader-like grip on e5 suffocates Black in the center; the only question for White is whether this bit of central overkill is worth the pain it causes the bishop on c1.
24. Kg2 followed by Rh1 is also very promising.  )
24... Nf5 25. Kg2 Nh6 26. Rh1 f6 Risky.
27. Nf3 Qf7?! 28. Rf2! Bb8?! 29. Ne5! Bxe5
29... fxe5 30. fxe5 Qd7 31. h5 g5 32. Bxg5 Bc7 33. Qd2 Bd8 34. Bxh6 gxh6 35. Qxh6 Qg7 36. Qf4 Black's pieces are mostly cut off from meaningful contact with the kingside, and h5-h6 will only make it worse. White is winning.  )
30. fxe5 f5 31. Bxh6 gxh6 32. g4 Black's kingside is about to collapse, and with no queenside counterplay on the horizon the rest is predictable.
32... h5 33. gxh5 Re7 34. Kf1 gxh5 35. Rg1+ Kh8 36. Rfg2 Nh7 37. Rg6 b5 Way too late.
38. a3 Qe8 39. Qxh5 a5 40. Qh6 Rcc7 41. Ke2 b4 See the previous comment. The net effect here is that it gives White a second extra pawn - a passer.
42. cxb4 axb4 43. axb4 Qc8 44. Kd2 Qe8 45. h5 Qc8 46. b5 Qe8 47. b6 Yes, it's a genuinely useful passed pawn!
47... Rb7 48. Ba6! Taking advantage of the overloaded rook.
48... Rxb6 49. Rg7 Forcing a speedy mate.
49... Rxb2+ 50. Kd3
50. Ke1 Rb1+ 51. Kf2 Rb2+ 52. Be2 was the most expeditious finish.  )
50... Rb3+ 51. Ke2 Rb2+ 52. Kf3 Rb3+ 53. Kf4
Howell, David vs. So, Wesley
Biel GM 43rd | Biel | Round 1 | 19 Jul 2010 | ECO: B13 | 0-1
1. e4 c6 2. d4 d5 3. exd5 cxd5 4. Bd3 Nc6 5. c3 Nf6 6. Bf4 Bg4 7. Qb3 Qc8 8. Nd2 e6 9. Ngf3 Be7 Almost 1200 games have reached this point, and now David Howell comes up with something brand new.
10. Kf1!? Clearing e1 for one rook - an idea we've seen before - while keeping the other rook free for different adventures. It isn't that bad, but it should be acknowledged that no one has followed his lead in the intervening six and a half ears.
10... Bh5 11. Re1 a6 The counterplay begins.
12. Qc2 b5 13. b4 A very consequential decision. White stops the minority attack plan dead in its tracks, but it comes at a cost. White is weakening at least two squares on the half-open c-file, c3 and c4. Black hurries to take advantage.
13... Bg6 Black would like to use the c4 square, so it makes sense to trade off one of the pieces that can defend it. This also makes e4 a potential outpost for a Black knight.
14. Bxg6 hxg6 15. Qd3 a5! 16. a3
16. Qxb5? axb4  )
16... axb4 17. axb4 Qb7 18. Nb3 Hoping to plug up one of the two open queenside files.
18... O-O
18... Ne4  )
19. h4 Ne4 20. Nfd2 Ra3 21. Rb1 Rfa8 22. Kg1 Nd8 23. Rh3 Qc6 Black has the initiative here, but objectively White is still okay.
24. Na5
24. Nxe4!? dxe4 25. Qe3 Qd5 26. Nc5  )
24... R8xa5! 25. bxa5 Nxc3 26. Rb3 Ra1+ 27. Nb1 b4 28. Bd2 Na2 White is still hanging on - barely.
29. Re3? This makes sense, hoping to bring the rook back to e1 (castling would have been much simpler!). Unfortunately for White, there's a fatal tactical problem.
29. Rb2! was forced, and kept White in the game.  )
29... Nc1! 30. Bxc1 Qxc1+ 31. Kh2?
31. Qf1  )
31... Nc6 32. g3??
32. Re2 Nxa5 33. Rc2 Qh6 34. Rbb2 Nc4 35. Rb3 g5  )
32... Nxa5 Black is winning a full rook from here, so White called it a day.

White owns the kingside, Black the queenside, and whoever’s attack strikes first wins. Everything is clear, right?

Perhaps not. The following game from the World Rapid Championship featured a seemingly completely different opening – the London System – that reached a Carlsbad structure. Instead of following the well-established plan, White changed the strategy – and won because Black failed to follow suit.

Carlsen, Magnus vs. Bok, Benjamin
World Rapid 2016 | Doha QAT | Round 6.1 | 27 Dec 2016 | ECO: A45 | 1-0
1. d4 Nf6 2. Bf4 d5 3. e3 c5 4. Nc3 cxd4 5. exd4 a6 6. Bd3 Nc6 7. Nce2 Bg4 8. c3 e6 From a very different opening, we suddenly see something that looks an awful lot like an Exchange Caro-Kann.
9. Qd2 Bh5 10. Bg5
10. Ng3 Bg6 11. Nf3 Be7 12. O-O O-O 13. Bxg6 hxg6 14. Rfe1 Rc8 15. Qe2 Nd7 16. Nf1 Na5 17. Ne3 Re8 18. Ng4 Nc4 19. Nge5 Ndxe5 20. Nxe5 Nxe5 21. Bxe5 Bd6 22. Qg4 b5 23. a3 a5 24. h4 b4 25. axb4 axb4 26. Bxd6 Qxd6 27. Re3 bxc3 28. bxc3 Rc4 29. Rh3 e5 1/2-1/2 (29) Archangelsky,M (2270)-Van Delft,M (2404) Hoogeveen 2006  )
10... Bg6 11. Nf4 Bxd3 12. Nxd3 Bd6 13. Nf3 Black is completely fine in this middlegame, and his decision to liquidate to an ending doesn't harm his position a bit.
13... Ne4 14. Bxd8 Nxd2 15. Kxd2 Rxd8
15... Kxd8 looks sensible too. Black's king doesn't need to castle but should get out of the rooks' way. It may turn out that the rook should stay on a8, however, or go to c8, so this recapture may be the most time-efficient of Black's options.  )
16. Rhe1 O-O 17. a4!? Up to now we may have thought about this position under the influence of the previous three games, but now things go upside down. White starts to play on the queenside, and it turns out that at least one very good plan for Black here is to go ...Rfe8, ...f6, and ...e5. Black drifts for a little while, and by the time he comes to Carlsen has taken over.
17... Rc8
17... Rfe8! 18. a5 Not necessarily best.
18... f6 /=+  )
18. a5! Rfe8 19. Nfe5 Bxe5
19... Nxe5 was better, as Black's bishop can keep watch over both c5 and e5.  )
20. Nxe5 Nxe5
20... f6  )
21. Rxe5 f6 22. Re3 Kf7 23. Ra4!? The rook would like to snake its way to b4 and then maybe b6, so Black tries to nip this plan in the bud.
23. f4 is natural (and good), but Carlsen has a completely different idea.  )
23... Rc4?! 24. Rxc4 dxc4 25. b3!! It looks as if nothing is going on or should be going on, but Carlsen finds a way to make something happen in a drawish-looking ending.
25... Rc8?!
25... cxb3 26. c4 Rc8 27. Kc3 This is dangerous for Black, but with a precise series of moves it seems he can survive.
...  b2! 28. Re1 Rc6! 29. c5! e5! 30. Rb1 exd4+ 31. Kxd4 Rc7! 32. Rxb2 Rd7+!  )
26. Re1! cxb3 27. Rb1 Rc4?! Active play is generally one's best bet in a rook ending, but this is a case where it will backfire.
28. Rxb3 Ra4? 29. Rxb7+ Kg6 30. Rb6 Rxa5 31. Rxe6 Black's biggest problem isn't his one pawn deficit; it's that White has two connected passers.
31... Ra2+ 32. Kd3! This wasn't necessary, but it highlights the power of White's passers.
32... Rxf2 33. c4 Rxg2 34. c5 Kf5 35. Re1 Rb2 36. c6
36. c6 Rb8 37. c7 Rc8 38. Re7 followed by d5-d6-d7, and Black is kaput.  )


Dennis Monokroussos is a FIDE master who has written about chess on his blog “The Chess Mind,” since 2005. He has been teaching chess for almost 20 years and for the last 10 years has been making instructional chess videos, which can be found at Between 1995 and 2006, he taught philosophy, including a four-year stint at the University of Notre Dame.