Two Knights Defence- 4.Nc3, 4.0-0, 4.d3

White's most aggressive tries against the Two Knights Defence are 4.Ng5 and 4.d4 (the latter transposes to the Scotch Gambit after 4...exd4) but in practice White sometimes prefers a quieter approach, simply defending the e4-pawn from attack.  The most important is 4.d3, but White can also try 4.Nc3, or leave the pawn hanging with 4.0-0.

4.Nc3 and 4.0-0

4.Nc3 should be met by the "fork trick" 4...Nxe4, intending 5.Nxe4 d5, attacking the bishop on c4 and the knight on e4.  I used to have the impression that this gave Black the advantage, reinforced by White's tendency to respond sub-optimally in my own games with it, but in fact the position is merely equal if White plays 6.Bd3, intending 6...dxe4 7.Bxe4, whereupon 7...Ne7 (intending ...f7-f5 hitting out at the bishop) may be more reliable than the standard "book move" 7...Bd6.

Instead White can transpose into a line of the Boden-Kieseritzky Gambit by playing 5.0-0, inviting Black to play 5...Nxc3 6.dxc3 f6 and play to hold onto the extra pawn.  White gets some hacking chances on the kingside (the most challenging plan probably being a quick Nf3-h4, threatening Qh5+, followed by pushing the f-pawn) and Black has to be careful, but it should not really provide enough compensation for the pawn.  A safer option is 6...Be7, returning the pawn, although many games with this line tend to end in draws.   4.0-0 will usually come to the same thing after 4...Nxe4 5.Nc3, as White doesn't have any particularly good alternatives (e.g. not 5.Re1 d5).  Two Knights Defence players should check out the notes to Game 3 (Fabri-Ashton, Blackpool 2014).  

4.d3

This line is often neglected in coverage of the Two Knights Defence because the wilder lines stemming from 4.Ng5 and 4.d4 are more fun to analyse, but in practice many White players at all levels prefer the quiet 4.d3, defending the pawn on e4.  White can develop in "Giuoco Pianissimo" style or try to steer the game into channels similar to the Closed variation of the Ruy Lopez (1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 etc.)  Since this approach from White is not really in keeping with the spirit of this site (Dr. Dave Regis at Exeter Chess Club calls this type of approach the "Old Stodge") I have focused mainly on active responses from Black's point of view, rather than giving an encyclopaedic coverage.

The most obvious way to force an active game is 4...d5, but unfortunately for Black it is rather dubious because after 5.exd5 Nxd5 6.0-0, the e5-pawn becomes too weak.  Black can sacrifice it and generate complications with 6...Bc5 7.Re1 0-0 8.Nxe5 Qh4, which might work at rapid time controls, but after 9.Qf3 and even 9.Nf3!?, inviting Black to take on f2, White can get quite a large advantage.

If Black wishes to play ...d7-d5, then a more reliable method of doing so is to strive for rapid kingside castling, getting the king off the e-file, and then play ...d7-d5.  Black can play 4...Bc5 transposing to the Giuoco Piano, with the aim of 5.0-0 0-0 6.c3 d5, which gives active play, though I think White can get a slight edge, and 6...d5 doesn't work very well if White tries 6.Nbd2 instead.  An alternative is 5.Nc3, against which Black cannot get in ...d7-d5, and after 5...d6 White can consider 6.Bg5, the Canal Variation, with the idea that White is doing quite well after 6...h6 7.Bxf6 Qxf6 8.Nd5 Qd8 9.c3 intending d3-d4, but 6...Na5 is an effective response for Black. 

Black appears to get more scope for active play with the apparently-restrained 4...Be7, as recommended by John Emms in Play the Open Games as Black.  Again, the main idea is to get in ...d7-d5, which often works better with the bishop placed on e7 rather than on c5, such as after 5.0-0 0-0 6.c3 d5, which appears to equalise fully here.  6...d5 is also playable against 6.Nbd2 and 6.Bb3, though is less likely to give full equality, and Black can offer a Ruy Lopez Marshall Gambit-style sacrifice of the e5-pawn in some lines, aiming to generate kingside play to compensate.  Against 6.Re1, Black cannot get in ...d7-d5 without it involving an unsound sacrifice of the e5-pawn, but can get good play with an ambitious plan involving ...Kh8, ...Ng8 and ...f7-f5, which is often followed up by a ...Rf8xf3 exchange sacrifice, taking advantage of the fact that White's Rf1-e1 has reduced the defence of the f2-pawn.  White can try 5.Nc3 instead, after which Black can consider a plan with ...d6 and ...Nf6-d7 followed by an eventual ...f7-f5.

Another idea for Black is to delay kingside castling and push the kingside pawns forward towards White's king.  I have given a few examples of this in the notes to the illustrative games.  This works quite well if White plays passively, but be warned, it is dubious against many of White's more ambitious approaches, for Black's king can become vulnerable in the centre and if Black castles queenside then White can get a strong attack with a2-a4 and b2-b4-b5.

Games and analysis

The coverage is divided up as follows:
1.  4...Bc5 5.c3 (or 5.0-0) after which I mainly focus on plans with ...d7-d5.
2.  4...Bc5 5.Nc3 (I mainly focus on the Canal Variation, 5...d6 6.Bg5)
3.  4...Be7 5.Nc3 (I also examine 5.c3, which allows Black to get in ...d7-d5 effectively, e.g. 5...0-0 6.0-0 d5).
4.  4...Be7 5.0-0 0-0 6.Re1 (here we see the plan with ...Kh8, ...Ng8 and ...f5 in action.)
5.  4...Be7 5.0-0 0-0 6.Bb3 (I think 6...d5 is Black's most reliable way of getting an active game here, though 6...d6 with the idea of steering play into Closed Ruy Lopez style channels may objectively be superior).

[Event "Capablanca Memorial Elite 49th"] [Site "Havana"] [Date "2014.05.12"] [Round "5"] [White "Ivanchuk, Vassily"] [Black "Vallejo Pons, Francisco"] [Result "0-1"] [ECO "C54"] [WhiteElo "2753"] [BlackElo "2700"] [PlyCount "106"] [EventDate "2014.05.08"] [EventType "tourn"] [EventRounds "10"] [EventCountry "CUB"] [EventCategory "19"] [SourceDate "2014.05.21"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 (3... Bc5 4. c3 Nf6 5. d3 {was the move-order used to get to the position at move 5.}) 4. d3 Bc5 {A sensible response which transposes into the closed lines of the Giuoco Piano.} 5. c3 {The most ambitious response, preparing either b2-b4 or an eventual d3-d4. Black's plan with ...d7-d5 will not work, however, if White plays 5.Nc3 as in the next illustrative example.} (5. O-O {is also fully playable.} O-O {(for 5...d6 see the note to 5.c3 d6)} 6. Nbd2 {This move discourages ...d5 because after exchanging pawns on d5 White then gets a nice outpost for the knight on e4.} ( 6. c3 d5 {transposes into the game.}) (6. Re1 {discourages ...d7-d5 but can be met by} Ng4 7. Re2 Nd4 8. Nxd4 Bxd4 (8... exd4 {is a good alternative, with the idea} 9. h3 Ne5 {and Black is well-placed to get in ...Kh8 and ...f7-f5}) 9. h3 Nf6 10. Nd2 d6 11. Nf3 {with equal chances, Brkic,A (2606)-Filippov,A (2606) Sibenik 2011}) 6... d6 (6... d5 {is dubious here, for example,} 7. exd5 Nxd5 8. Re1 Bg4 9. h3 Bh5 10. Ne4 {gives White the advantage.}) 7. c3 a6 { (This is played to give the c5-bishop an escape hatch on a7.)} 8. Bb3 Kh8 { With the long-term idea of playing ...f7-f5 after the f6-knight moves out of the way.} 9. Nc4 Ba7 10. Bg5 h6 11. Bh4 Ne7 (11... g5 12. Bg3 {has scored 100% for Black in my database, although objectively White is probably marginally better.} Ne7 13. Bc2 Nh5 14. d4 {(Hedman,E (2391)-Nabaty,T (2523) Pardubice 2010) and I suggest simply 14...Nxg3.}) 12. Bxf6 gxf6 13. Nh4 Rg8 {and White has the better pawn structure but Black has some counterattacking chances down the g-file, Petrosian,T (2602)-Mamedyarov,S (2721) Novi Sad 2009}) 5... O-O { Black gets the king off the e-file and keeps open the possibility of playing .. .d7-d5 in one move, rather than ...d7-d6 and then ...d6-d5 later on.} (5... d6 {is also playable, but it is less flexible as Black can no longer play ... d7-d5 in one move. I have experimented with approaches involving queenside castling in this line, which may be playable at club level but they are objectively quite dubious, for example,} 6. O-O a6 {(this is played to give the c5-bishop an escape hatch on a7 if White plays b2-b4 or Nb1-d2-b3)} 7. Bb3 Bg4 (7... Ba7 8. Nbd2 O-O {is the standard continuation, e.g.} 9. h3 h6 10. Re1 Be6 11. Nf1 {is reminiscent of the Closed Ruy Lopez, with White intending a slow kingside attack with Nf1-g3, Radjabov,T (2788)-Naiditsch,A (2712) Istanbul 2012}) 8. Nbd2 Qe7 9. Re1 O-O-O 10. h3 Bh5 11. a4 g5 12. Bd5 g4 13. hxg4 Nxg4 14. Rf1 {with some advantage for White, since following Black's flurry of kingside activity, White intends b2-b4 and a potent queenside pawn storm.}) 6. O-O ({Another idea for White is immediate queenside expansion:} 6. b4 Bb6 7. a4 a6 {Diagram [#] is the usual sequence that follows, and then} 8. Qb3 (8. O-O {can again be met by} d5 9. exd5 Nxd5 {and Black is comfortable here.}) 8... d6 9. O-O Qe7 10. Nbd2 {offers equal chances, Naby,S (2340)-Kamel, M (2243) Cairo 2014}) (6. Bb3 {can also be met effectively by} d5 {and there have been numerous tests of this line at grandmaster level. Black can get away with striving to open the centre.} 7. Qe2 Re8 8. Bg5 dxe4 9. dxe4 h6 10. Bh4 Be7 11. Nbd2 Nh5 {gives equal chances, Carlsen,M (2837)-Bacrot,E (2713) Biel 2012}) 6... d5 {A refinement over the 4...d5 idea seen in the previous game, which turned out to be too hasty. With Black having castled, Black can now get in this central thrust without leaving the e5-pawn too weak.} 7. exd5 { With the bishop on c4 attacked, White has no good alternative to taking on d5.} Nxd5 8. a4 (8. Re1 {is also playable, with the idea of attacking the e5-pawn, a plan which proved effective after the immediate 4...d5.} Bg4 ({The rarely-played} 8... Be6 9. Nbd2 Bd6 {may be more reliable.}) 9. h3 Bh5 10. Nbd2 (10. g4 Bg6 11. Nxe5 Nxe5 12. Rxe5 c6 {is reminiscent of the Marshall Gambit in the Ruy Lopez for Black, and is quite a favourable version since White has weakened the kingside.}) 10... Nb6 (10... Bb6 11. Ne4) 11. b4 Bd6 12. Bb3 {and White may have a slight advantage, Eliseev,U (2541)-Najer,E (2634) Moscow 2013} ) (8. b4 {allows White to go after the e5-pawn, but at the cost of development. } Bb6 (8... Be7 {is more popular, e.g.} 9. b5 Na5 10. Nxe5 Nxc4 11. dxc4 Nb6 12. Ba3 Bxa3 13. Qxd8 Rxd8 14. Nxa3 Na4 {and Black has enough compensation for the pawn, Ruan,L (2480)-Kosteniuk,A (2507) Antakya 2010}) 9. b5 Na5 10. Nxe5 Be6 {and Black has reasonable compensation for the pawn as White is well behind in development, Kralevski,A (2136)-Georgescu,T (2493) Albena 2014}) 8... a6 9. a5 {I am not convinced by this move- personally I would prefer to leave the pawn on a4 to maintain the possibility of playing b2-b4 and then b4-b5, supported by the a4-pawn.} (9. Nbd2 {is a sensible alternative with the idea of Ne4. For example,} Nb6 10. Ba2 Bg4 (10... Qxd3 11. a5 Nd7 12. Bb1 {gives White reasonable compensation for the pawn, as Black's kingside is set to come under fire.}) 11. h3 Bh5 12. a5 Nd7 13. Ne4 $14 {Demchenko,A (2613) -Tomashevsky,E (2695) Vladivostok 2014}) 9... Nf6 10. Nbd2 h6 {Diagram [#] Chances are equal in this position. Black has a significant presence in the centre, but always has to take care of the possibility of the e5-pawn becoming weak.} 11. Qe2 Re8 12. Ne4 Bf8 {A standard plan for Black, protecting the g7-pawn and opening the e-file for the rook on e8.} 13. Bb3 Be6 14. Ba4 Nd7 15. b4 f5 {Black starts a kingside pawn storm.} 16. Ned2 g5 17. Bb2 (17. h3 {comes into consideration. It makes it easier for Black to prise open the kingside but then Black's king ends up exposed, as well as White's. For example,} Qf6 18. Bb2 g4 $6 (18... Rad8 {gives equal chances.}) 19. hxg4 fxg4 20. Ne4 Qg6 21. Nfd2) 17... Bg7 18. Nc4 Bxc4 19. dxc4 g4 {Diagram [#]} 20. Rad1 $6 {This amounts to a piece sacrifice but I am not convinced that White gets enough for the piece.} (20. Nd2 e4 21. Rfd1 {and I think chances are still equal.}) 20... gxf3 21. Qxf3 Rf8 22. Bc1 Rf7 {This defends the knight on d7 and thus allows Black to free the position. It may entail giving up an exchange on f7, but being a piece ahead, Black can afford to return some material and keep some advantage.} 23. c5 Qf6 {Black is happy to let an exchange drop, leaving Black with two pieces and attacking chances for a rook.} 24. Bb3 Nf8 25. g4 (25. Bxf7+ Qxf7 {intends ...Qg6, ...Ne6 and/or ...f4.}) 25... f4 26. Qe4 Ne6 {Now Black is clearly better since Black maintains an initiative and White can no longer take the rook on f7. Black went on to win with the extra piece.} 27. f3 Kh8 28. Rde1 Re8 29. Kh1 Rfe7 30. Qf5 Qh4 31. Bxe6 Rxe6 32. Bxf4 Rf6 33. Bg3 Rxf5 34. Bxh4 Rf4 35. Re4 Ref8 36. Kg2 Kg8 37. Bg3 Rxe4 38. fxe4 Rxf1 39. Kxf1 Bf6 40. h4 Kf7 41. Ke2 Ne7 42. Kf3 Ke6 43. Bf2 Ng6 44. g5 hxg5 45. h5 Ne7 46. Be3 c6 47. Kg4 Ng8 48. Bd2 Nh6+ 49. Kg3 Be7 50. Bc1 Kf6 51. Bd2 Bf8 52. Bc1 Ng8 53. Bd2 Bh6 0-1 [Event "Baltic Club-chT email"] [Site "ICCF email"] [Date "2005.02.15"] [Round "?"] [White "Lapeginas, Zilvinas"] [Black "Frey Beckman, Kenneth"] [Result "1/2-1/2"] [ECO "C50"] [WhiteElo "2389"] [BlackElo "2700"] [PlyCount "45"] [EventDate "2005.??.??"] [EventType "tourn (corr)"] [EventRounds "9"] [Source "ChessBase"] [SourceDate "2008.06.24"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 (3... Bc5 4. Nc3 Nf6 5. d3 {was the rather unusual move-order used in this game.}) 4. d3 Bc5 5. Nc3 {This is the main alternative to 5.c3. The main advantage of this move is that it covers the d5-square and thus prevents the plan of castling and playing ...d5 that Black uses in this game, but the downside is that it doesn't give White as much scope for the "Closed Ruy Lopez" ideas of Bc4-b3-c2 and c2-c3 intending d3-d4.} d6 (5... O-O 6. Bg5 {is an improved version of the game continuation for White, since Black's e5-pawn is not strongly defended and so Black cannot play 6... Na5 here without losing a pawn for next to nothing.} h6 (6... d6 7. Nd5 {puts Black's kingside under dangerous pressure.}) 7. Bxf6 Qxf6 8. Nd5 Qd8 9. c3 { gives White chances of an edge, preparing to advance in the centre with d3-d4.} ) 6. Bg5 {Diagram [#] This is known as the Canal Variation (not to be confused with Canal's line in the 4.d4 line of the Two Knights Defence).} (6. Be3 Bb6 { (Exchanging on e3 opens the f-file for White's f1-rook after castling kingside, and strengthens White's centre.)} 7. O-O {Diagram [#] This is quite a common Giuoco Pianissimo position. One recent game saw Black attempt a double-edged kingside pawn advance:} h6 8. h3 g5 $5 (8... O-O 9. Qd2 Be6 10. Bb3 Qd7 {is a more typical continuation, which is quite drawish because of the completely symmetrical position.}) 9. Bxb6 axb6 10. d4 g4 11. hxg4 Bxg4 12. dxe5 Nxe5 13. Be2 Bxf3 14. Bxf3 Rg8 {with counterplay down the g-file, although White enjoys the better pawn structure, Thorgeirsson,S (2223)-McShane,L (2683) Reykjavik 2011}) 6... Na5 {This is not the most popular response but it is probably best. } (6... h6 7. Bxf6 Qxf6 8. Nd5 Qd8 9. c3 {intending d3-d4 is the usual continuation, but I think White has chances of a slight edge here. For example,} Ne7 10. Ne3 O-O 11. O-O Bb6 12. d4 Ng6 13. a4 c6 14. Re1 {Piscopo,P (2409)-Anand,V (2785) Porto Vecchio 2014}) 7. Bb3 (7. Bxf6 Qxf6 8. Nd5 Qd8 {is acceptable for Black, who is set to force White back with ...c7-c6 thanks to the knight having vacated the c6-square. One complicated line then runs} 9. b4 Nxc4 10. bxc5 c6 11. dxc4 cxd5 12. Qxd5 Qa5+ {with equal chances, Rabiega,R (2455)-Lukacs,P (2485) Budapest 1993}) 7... Nxb3 (7... c6 {will most likely transpose into the game continuation following} 8. d4 Nxb3 9. axb3 exd4 10. Nxd4) 8. axb3 c6 {This is preferable to the 8...Be6 given by Tim Harding in one of his Kibitzer articles.} (8... Be6 9. Na4 {gives White a slight edge as White is set to get the bishop-pair, if} Bd4 (9... Bb4+ 10. c3 Ba5 11. b4 {is also slightly better for White, e.g.} Bb6 12. Nxb6 axb6 13. Bxf6 gxf6 14. O-O) 10. c3) 9. d4 (9. Na4 Bb4+ 10. c3 Ba5 {and the bishop on a5 can drop back to c7, and so White does not get the bishop-pair.}) 9... exd4 10. Nxd4 h6 11. Bh4 {Diagram [#] This position offers equal chances.} O-O {The more aggressive idea of queenside castling is virtually ruled out here because of White's half-open a-file.} 12. O-O Re8 (12... g5 {is also playable, breaking the pin but weakening the black kingside.} 13. Bg3 Re8 14. Re1 d5 15. e5 Ne4 {gives equal chances, Morozevich,A (2575)-Georgiev,K (2615) Tilburg 1994}) 13. f3 a5 14. Kh1 Bd7 15. f4 b5 16. Qd3 Bxd4 17. Qxd4 b4 18. Nb1 c5 19. Qd3 ({Not} 19. Qxd6 $2 Nxe4 {counterattacking against the white queen.} 20. Bxd8 Nxd6 {leaves Black with the better chances due to the threat of ...Re8-e2.}) 19... a4 20. bxa4 Rxa4 21. Rxa4 Bxa4 22. Nd2 Bc6 23. Re1 {The players agreed to a draw here but at club level I would expect there to be plenty of scope for play in this position.} 1/2-1/2 [Event "GRE-chT 40th"] [Site "Achaea"] [Date "2012.07.02"] [Round "4.8"] [White "Managadze, Nikoloz"] [Black "Mastrovasilis, Athanasios"] [Result "1-0"] [ECO "C55"] [WhiteElo "2472"] [BlackElo "2537"] [PlyCount "69"] [EventDate "2012.06.30"] [EventType "team-swiss"] [EventRounds "7"] [EventCountry "GRE"] [Source "ChessBase"] [SourceDate "2012.08.24"] [WhiteTeam "Von Heraklion Contact"] [BlackTeam "Florina"] [WhiteTeamCountry "GRE"] [BlackTeamCountry "GRE"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d3 Be7 {Diagram [#] Recommended by John Emms in his book Play the Open Games as Black. This restrained development of the bishop gives Black less potential counterplay against f2 in the short-term, but the bishop is not prone to harassment from b2-b4 and a2-a4-a5 and the approaches with ...d7-d5 are often more effective. If White attempts to prevent ...d7-d5 then play often heads into Closed Ruy Lopez-esque channels, but Black has an interesting independent plan involving engineering an f-pawn push, as used in this game.} 5. Nc3 {White continues in Giuoco Pianissimo style, covering the d5-square and making it harder for Black to get in ... d7-d5. However, with the black king's bishop placed on e7 rather than c5, White cannot effectively follow this up with Bc1-g5 as there is no pin on the f6-knight.} (5. c3 {is less effective than it is against 4...Bc5, because a subsequent b2-b4 or d3-d4 will not hit a bishop on c5. Black can aim to strike out in the centre with ...d7-d5.} O-O (5... d6 {is less flexible, as it prevents Black from playing ...d7-d5 in one move. Black has again occasionally experimented with kingside pawn advances, with the aim of eventually castling to the queenside, but again these are unconvincing, e.g.} 6. O-O Bg4 7. h3 Bh5 8. Bb3 Qd7 9. Be3 g5 (9... O-O-O 10. Nbd2 {followed by Bb3-a4 and b2-b4 is quite tricky for Black to deal with.}) 10. Bxg5 Rg8 11. Bxf6 Bxf6 12. Kh2 {with advantage for White, Tiviakov,S (2686)-Bacrot,E (2705) Ajaccio 2008}) 6. O-O (6. Bb3 {has the idea that after} d5 {White doesn't have to capture on d5 immediately, but Black has a reasonable game following} 7. Nbd2 Re8 8. O-O dxe4 9. dxe4 Bc5 10. Qe2 Qe7 {, Tiviakov,S (2654)-Kobese,W (2343) Port Elizabeth 2013}) 6... d5 7. exd5 Nxd5 8. Re1 (8. h3 {can be tried instead, preventing ...Bc8-g4, but Black is doing fine in this position, for example,} Nb6 9. Bb5 Nb8 10. d4 ({Not} 10. Nxe5 $2 Qd5) 10... e4 11. Nfd2 (11. Ne5 c6) 11... c6 12. Be2 f5 {and if anything I prefer Black's chances, Gunnarsson,J (2430)-Adams,M (2734) Kallithea 2008}) 8... Bg4 9. h3 Bh5 10. Nbd2 Nf4 (10... Nb6 {is also good}) 11. Nf1 Na5 12. Bxf4 exf4 13. Bb5 a6 14. Ba4 { (Yudasin,L (2575)-Dreev,A (2605) Lvov 1990) and now John Emms suggests} c5) 5... d6 {Preparing to bring the c8-bishop out to g4 or e6 and threatening ... Nc6-a5, securing the bishop-pair.} (5... O-O {is also playable, but scores slightly less well for Black in the database. It will normally transpose to the main line with 5...d6, since Black cannot easily engineer ...d7-d5. In the case of 5.Nc3, unlike most other responses, it seems that ...d6 is more flexible than castling.}) 6. a3 {This is played to give the c4-biahop an escape hatch on a2 if it is attacked.} (6. a4 {is another way to generate an escape hatch on a2 for the c4-bishop.} O-O {is usually played in response} ( 6... Bg4 7. h3 Bxf3 8. Qxf3 Nd4 9. Qd1 c6 {is a reasonable alternative.}) 7. h3 Be6 8. O-O (8. Bxe6 fxe6 {opens up the f-file for the f8-rook.}) 8... Qd7 (8... Nxe4 {is a good way to stir up some complications, e.g.} 9. Nxe4 d5 10. Neg5 dxc4 11. Nxe6 fxe6 12. dxc4 e4) 9. Re1 a6 10. Bd2 Rae8 11. Rb1 {gave an equal game in Short,N (2690)-Ganguly,S (2655) Sestao 2010}) (6. h3 {is most often played (although it is not often used at grandmaster level), preventing ...Bg4, but it scores quite poorly for White. Since White's h2-h3 is rather slow and presents Black with a potential target on the kingside, in this line Black can afford to leave the king in the centre for a while and push the kingside pawns forward:} Na5 {This grabs the bishop-pair, since White has no way of moving the bishop out of the way of the knight's attack.} 7. Bb3 (7. Be3 Nxc4 8. dxc4 c6 9. Qe2 Qc7 10. Nh2 Be6 11. Ng4 Nd7 12. O-O h5 13. Nh2 f6 {Renner,C (2405) -Sanikidze,T (2512) Nuremberg 2011}) 7... Nxb3 8. axb3 c6 9. Bg5 h6 10. Bh4 Qb6 11. Qd2 g5 12. Bg3 Nh5 {with good kingside play, Zharokov,K (2175)-Kazakov,K (2322) Pavlodar 2010}) (6. O-O {can again be met effectively by the grab of the bishop-pair with} Na5 {, for example,} 7. Qe2 c6 8. a4 Nxc4 9. dxc4 Qc7 10. Rd1 O-O 11. h3 h6 12. b3 Be6 13. Bb2 Nh5 14. Qe3 Nf4 {with some play developing on the kingside, Hrabusa,M (2287)-Zeberski,J (2400) Czechia 2010}) 6... O-O (6... Bg4 7. h3 Bxf3 (7... Bh5 8. Be3 {is slightly better for White, Bohm,H (2410)-Timman,J (2585) Leeuwarden 1978}) 8. Qxf3 Nd4 9. Qd1 c6 { intending ...d6-d5 is a reasonable alternative for Black, although White now has the bishop-pair. Garcia Palermo,C (2475)-Valsecchi,A (2336) Sarre 2009}) 7. O-O Nd7 {This is played with the intention of bringing the knight around to b6 and may envisage a subsequent ...f7-f5 advance. Mark Hebden has also tried out this move a couple of times in the late 1980s.} ({Again} 7... Bg4 8. h3 Bxf3 9. Qxf3 Nd4 10. Qd1 c6 {comes into consideration.}) (7... Be6 {is another decent idea, encouraging a bishop trade on e6 which would open the f-file for the f8-rook, but I think White can get around this idea by playing Nc3-d5, e.g. } 8. Re1 Qd7 9. h3 h6 10. Nd5 Bd8 {(otherwise White gets the bishop-pair after Nd5xe7)} 11. Ne3 Ne7 12. d4 {with a slight edge for White, Bauer,C (2634) -Fressinet,L (2654) Ajaccio 2007}) 8. Nd5 {This has invariably been played in the high-level games that I can find.} (8. Be3 {looks sensible.} Nb6 9. Bb3 Bg4 10. Nd5 Nd4 {is one possible continuation, leading to some complications.}) 8... Nb6 9. Nxb6 (9. Nxe7+ Qxe7 10. Bg5 {is a good alternative for White, bagging the bishop-pair.} Qe8 11. Ba2 Be6 12. Bxe6 ({Or} 12. Be3 Bxa2 13. Rxa2 f5) 12... fxe6 13. a4 Qg6 {gives chances for both sides, with White having the advantage on the queenside and Black seeking counterplay down the f-file, Hauchard,A (2315)-Hebden,M (2455) Cappelle la Grande 1989}) 9... axb6 10. h3 { Diagram [#]} Kh8 {The start of an ambitious plan, intending ...f7-f5 hitting out in the centre and introducing the possibility of a kingside offensive.} 11. d4 (11. Be3 {is probably superior, with the idea} f5 12. exf5 Bxf5 13. Bd5 { with equal chances.}) 11... f5 12. dxe5 fxe4 13. exd6 exf3 (13... cxd6 {with the idea of ...d5, ...Qe8 and ...Qg6 probably gives Black better winning chances.}) 14. dxe7 Qxe7 15. Re1 {Diagram [#]} Qh4 $6 {A pawn sacrifice which does not provide Black with enough compensation.} (15... Qf6 16. Qxf3 Qxf3 17. gxf3 Bxh3 {was best, with equal chances in the endgame, as Black's better pawn structure is offset by the white bishop-pair following} 18. Kh2) 16. Qxf3 Be6 17. Re4 Qd8 18. Qg3 (18. Qd3 {is significantly better for White, whereas the move played in the game allows Black to get counterplay.}) 18... Bxc4 19. Rxc4 Qd1+ 20. Kh2 Qe2 21. Rg4 Ne5 $6 (21... Rxf2 22. Qxc7 Rg8 {is probably better. Black has quite a powerful attack on the seventh rank.}) 22. Rf4 Ng6 23. Rxf8+ Rxf8 24. Be3 {Now White is winning. Black cannot afford to take on c2 because of Rc1 and Rxc7.} c5 25. Qd6 Rf6 26. Qb8+ Nf8 27. b4 cxb4 28. axb4 Qc4 29. Rd1 Qxb4 30. Qxb7 h6 31. Rd4 Qc3 32. Rg4 Ne6 33. Bd4 Qxd4 34. Rxd4 Nxd4 35. Qa8+ 1-0 [Event "Bundesliga 0708"] [Site "Germany"] [Date "2008.04.20"] [Round "15.4"] [White "Kuczynski, Robert"] [Black "Naiditsch, Arkadij"] [Result "0-1"] [ECO "C55"] [WhiteElo "2483"] [BlackElo "2639"] [PlyCount "48"] [EventDate "2007.10.20"] [EventType "tourn"] [EventRounds "15"] [EventCountry "GER"] [Source "ChessBase"] [SourceDate "2008.05.28"] [WhiteTeam "Erfurt"] [BlackTeam "Baden-Baden"] [WhiteTeamCountry "GER"] [BlackTeamCountry "GER"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d3 Be7 5. O-O {White castles out of any possible trouble in the centre and may consider Rf1-e1.} (5. c3 O-O {prepares to play ...d7-d5, the same plan as seen in the 5...Bc5 variation.} (5... d6 { means that Black cannot get in ...d5 without losing a tempo. Again Black has occasionally tried aggressive kingside pawn advances in this line with the aim of eventually castling to the queenside, but they are not convincing, e.g.} 6. O-O Bg4 7. h3 Bh5 8. Bb3 Qd7 9. Be3 g5 (9... O-O-O 10. Nbd2 {followed by 11.Ba4 }) 10. Bxg5 Rg8 11. Bxf6 Bxf6 12. Kh2 $14 {Tiviakov,S (2686)-Bacrot,E (2705) Ajaccio 2008}) 6. O-O d5 7. exd5 Nxd5 8. Re1 Bg4 9. h3 Bh5 10. Nbd2 Nf4 {is one typical line, giving equal chances.} (10... Nb6 {is a good alternative- John Emms.}) 11. Nf1 Na5 12. Bxf4 exf4 13. Bb5 a6 14. Ba4 {with an equal game, Yudasin,L (2575)-Dreev,A (2605) Lvov 1990}) 5... O-O {Black removes the king from the e-file and prepares to play ...d7-d5.} ({Again, the idea of leaving the black king in the centre for a while and pushing the kingside pawns may be playable, but dubious:} 5... d6 6. c3 Bg4 (6... O-O {is normal, which probably suffices for equality, e.g.} 7. Bb3 Be6 8. Bc2 d5 9. Qe2 dxe4 10. dxe4 Nd7 11. Rd1 Qe8 12. Nbd2 {Tiviakov,S (2673)-Mirzoev,A (2560) Nakhchivan 2012}) 7. Nbd2 Qd7 8. Bb3 g5 9. Nc4 Rg8 10. Ne3 (10. Ba4 {threatening Na5 is the main problem with Black's idea.}) 10... Be6 11. Nd5 g4 12. Ng5 Bxd5 13. exd5 Na5 {with a sharp and unclear position, Lach,A (2160)-Naiditsch,A (2712) Warsaw 2011}) 6. Re1 {For 6.Bb3, discouraging ...d7-d5 due to not having yet played c2-c3, see the next game (although 6...d5 is still playable against 6.Bb3). The idea of 6.Re1 is that Black cannot play ...d7-d5 without it involving a dubious pawn sacrifice.} (6. c3 {is well met by} d5 {transposing to the line 5.c3 0-0 6.0-0 d5 discussed in the notes to the previous game.} 7. exd5 Nxd5 8. Re1 Bg4 9. h3 Bh5 10. Nbd2 {is the typical continuation and now Black chooses between 10... Nb6 and 10...Nf4, with reasonable chances.}) (6. Nbd2 {is a reasonable alternative for White, which also discourages ...d7-d5, though White does not benefit as strongly from the potential outpost for the knight on e4 as in the 4...Bc5 variation because Nd2-e4 does not attack a bishop on c5.} d5 (6... d6 { tends to lead to Closed Ruy Lopez type positions, albeit a slightly improved version for Black:} 7. c3 Na5 ({Aiming for ...f7-f5 with} 7... Kh8 8. Re1 Ng8 { doesn't appear to work well for Black here, e.g.} 9. d4 f5 10. dxe5 dxe5 11. exf5 Bxf5 12. Qe2 Bg4 13. Bxg8 Rxg8 14. h3 {with advantage for White, Emms,J (2505)-Trent,L (2387) London 2005}) 8. Bb5 a6 9. Ba4 c5 10. Re1 Qc7 {with equal chances, Vajda,L (2557)-Mastrovasilis,D (2607) Brasov 2011}) 7. exd5 Nxd5 8. Re1 {scores very well for White in my database.} Bg4 9. h3 Bh5 10. g4 Bg6 11. Nxe5 Nxe5 12. Rxe5 {has been tested in a handful of games, with a strong plus score for White, but none of these games continued with} c6 {defending the knight on d5 (which surprises me, as it is the standard reaction in the Marshall Gambit in the Ruy Lopez). With White's kingside weakened, Black's chances are better than in the 6.Re1 d5 line examined earlier. For example,} 13. Qf3 Rc8 (13... Bd6 14. Rxd5 cxd5 15. Bxd5 {is an important resource for White.}) 14. Nf1 Bd6 15. Re2 f5 16. Ne3 fxg4 17. Qxg4 Kh8 {and Black is developing reasonable attacking chances on the kingside.}) 6... d6 (6... d5 7. exd5 Nxd5 8. Nxe5 Nxe5 9. Rxe5 c6 {is an inferior relative of the Ruy Lopez Marshall Gambit for Black, since White has played d2-d3 instead of c2-c3, which makes it easier for White to bring pieces over to defend the white king. The line has been attempted occasionally at high levels, but without much success, e.g.} 10. Re1 Bd6 11. Nc3 Nxc3 12. bxc3 Qh4 13. g3 Qf6 14. d4 {with advantage for White, Dominguez Perez,L (2719)-Kosteniuk,A (2517) Moscow 2009}) 7. c3 (7. a4 {was noted in John Emms's Play the Open Games as Black as White's main try for an advantage, giving the c4-bishop an escape hatch on a2 and discouraging ...b7-b5.} Kh8 (7... Na5 8. Ba2 c5 9. c3 {is probably a little better for White.}) 8. a5 a6 (8... Ng8 9. a6 {is troublesome for Black- John Emms.}) 9. c3 Ng8 10. d4 f5 11. dxe5 fxe4 12. exd6 (12. Rxe4 {is stronger, challenging Black to try the exchange sacrifice} Rxf3 13. gxf3 Nxe5) 12... Bxd6 13. Ng5 Nf6 {with good play for Black, Kovalev,V (2583)-Mastrovasilis,D (2589) Warsaw 2013}) 7... Kh8 {Black intends ...Ng8 followed by ...f5, with the aim of generating a kingside offensive. In most cases this plan works well, but there are a couple of White responses that render it less effective, which should be noted.} (7... Na5 {is the most popular continuation, which tends to be reminiscent of the Closed Ruy Lopez.} 8. Bb5 a6 9. Ba4 c5 10. Nbd2 { transposes into the note to 6.Nbd2 d6 7.c3.}) (7... h6 {intending ...Nh7 is another reasonable plan, e.g.} 8. Nbd2 Nh7 9. h3 Ng5 10. Nf1 {with an equal game, Lavrik,D (2427)-Lund,S (2408) Vladimir 2007}) 8. h3 {This move is quite slow and allows Black to go ahead with the plan of ...Ng8 and ...f5, but it does prevent the pinning ...Bc8-g4.} (8. Nbd2 {is the response against which Black has to be careful, as going ahead with the ...f7-f5 plan doesn't work as well because the knight on d2 covers the important f3 and e4-squares.} Ng8 ( 8... Bg4 {is probably best, taking advantage of the omission of h2-h3, and then } 9. h3 Bh5 10. Nf1 Qd7 {is a plausible continuation, with equal chances.}) 9. d4 f5 10. dxe5 dxe5 (10... fxe4 11. Nxe4) 11. Bd5 $16) (8. Bb3 {is similar:} Bg4 (8... Ng8 9. d4 f5 {doesn't work very well for Black, e.g.} 10. exf5 Bxf5 11. dxe5 dxe5 12. Qxd8 Raxd8 13. Nxe5 Nxe5 14. Rxe5) 9. h3 Bh5 10. Nbd2 Qd7) 8... Ng8 9. d4 f5 {Diagram [#] The key strike, preparing to open up the f-file. } 10. dxe5 fxe4 11. Rxe4 Rxf3 {This exchange sacrifice is probably Black's strongest move in this position.} (11... Bf5 12. Re1 {is the main alternative, although White may end up slightly better after a queen trade on d8, e.g.} dxe5 13. Qxd8 Raxd8 14. Nbd2 {and Black will have some difficulty defending the isolated e-pawn.}) 12. gxf3 (12. Qxf3 {runs into the fork} d5 {and if} 13. Bxd5 Qxd5) ({If} 12. Bxg8 Rf8) 12... Bf5 13. Re1 (13. e6 {is the computer's preference, though Black still gets some kingside play, e.g.} Qe8 (13... Bxe4 $6 14. fxe4 {allows the white queen to cover the kingside.}) 14. Rg4 Nf6 (14... Bxg4 $6 15. fxg4 {is again too materialistic.}) 15. Qe2 Ne5 {with good play.}) 13... Nxe5 14. Bf1 Qe8 15. f4 ({Perhaps} 15. Bf4 {bringing a piece over to the kingside to help defend the king.}) 15... Qg6+ 16. Kh2 Bg4 (16... Bh4 {is a good alternative, attacking f2, though it may lead to a draw by repetition following} 17. Be3 (17. fxe5 $2 Bxf2 {leaves Black with too many kingside threats.}) 17... Bc2 18. Qe2 Bd3 19. Qd1) 17. Be2 Bxe2 18. Rxe2 Nf3+ 19. Kh1 { Diagram [#] Chances are approximately equal in this position, but White has to be careful due to the exposed king.} Qh5 20. Kg2 $2 (20. Qf1 {was the best way to defend.}) 20... Rf8 21. Qd3 ({No better is} 21. Nd2 Nh4+ 22. Kh1 Rxf4) 21... Rf6 22. Qxf3 Rg6+ 23. Qg4 Rxg4+ 24. hxg4 Qd5+ 0-1 [Event "Bazna Kings 4th"] [Site "Medias"] [Date "2010.06.22"] [Round "8"] [White "Radjabov, Teimour"] [Black "Ponomariov, Ruslan"] [Result "1/2-1/2"] [ECO "C55"] [WhiteElo "2740"] [BlackElo "2733"] [PlyCount "59"] [EventDate "2010.06.14"] [EventType "tourn"] [EventRounds "10"] [EventCountry "ROU"] [EventCategory "20"] [Source "ChessBase"] [SourceDate "2010.07.15"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d3 Be7 5. O-O O-O 6. Bb3 {Diagram [#] White drops the bishop away from c4, and discourages ...d7-d5 due to the fact that White has not yet played c2-c3 (although as John Emms notes, 6...d5 is "just about playable"). I take a look at both ...d5 and ...d6.} d5 (6... d6 {Black opts to be restrained with the d-pawn, but I think in this variation it is hard to see a good alternative to acquiescing to Closed Ruy Lopez type positions.} 7. c3 Kh8 {Black continues with the ambitious plan of striving for ...f7-f5, but this is less convincing if White plays for a quick d3-d4.} ({ More common is for Black to continue in a similar vein to the Ruy Lopez Closed Variation, with} 7... Na5 8. Bc2 c5 9. Nbd2 Nc6 10. Re1 Qc7 {with equal chances, as given by John Emms, Areshchenko,A (2647)-Graf,A (2620) Torrelavega 2007}) 8. Nbd2 Ng8 9. d4 (9. Re1 {can also be considered, e.g.} f5 10. h3 f4 11. d4 {with a slight advantage for White due to the strong centre although Black can generate attacking chances with ...Qe8 and ...Qg6, and went on to win, Olesen,P (2065)-Gofshtein,L (2542) Le Touquet 2007}) 9... Bg4 (9... f5 10. dxe5 dxe5 11. Nc4 {is better for White, Michiels,B (2465)-Claverie,C (2345) Metz 2010}) 10. h3 Bh5 {(Jonkman,H (2425)-Milov,V (2683) Port Erin 2004) and White can consider} 11. d5 Nb8 12. Nc4 $14) 7. exd5 {Exchanging pawns on d5 and preparing to attack the vulnerable e5-pawn.} ({If} 7. Nbd2 Bg4 8. c3 d4 9. h3 Bh5 {with equal chances for both sides, one game featured an ambitious piece sacrifice on the kingside with} 10. Qe2 b5 11. g4 Nxg4 (11... Bg6 {is reasonable enough.}) 12. hxg4 Bxg4 13. Bd5 Rb8 {, Mamikonian,T (2283)-Pap,M (2500) Rethymno 2012}) 7... Nxd5 8. h3 {White prevents the pin with ...Bg4, with the aim of pressuring the e5-pawn and making it difficult for Black to organise an effective way of sacrificing it for a kingside attack.} (8. Re1 Bg4 9. h3 Bh5 10. g4 (10. c3 {can be met by} Kh8 {intending ...f7-f5.}) 10... Bg6 11. Nxe5 Nxe5 12. Rxe5 c6 {gives Black reasonable compensation for the pawn, in the style of the Ruy Lopez Marshall Gambit, Tiviakov,S (2623)-Buhmann,R (2573) Vila Nova de Gaia 2010}) 8... a5 {The idea is to kick the white bishop on b3 away from the a2-g8 diagonal to enable Black to defend the e5-pawn with . ..f7-f6.} (8... Bf6 {defending e5 is the main alternative, but it is rather passive and White tends to get some advantage against this.}) ({I don't think Black gets enough for the pawn following the Marshall Gambit-style} 8... Be6 9. Re1 Re8 10. Nxe5 Nxe5 11. Rxe5 c6) 9. a3 Nd4 {Black threatens to grab the bishop-pair by exchanging off White's bishop on b3, and encourages a knight exchange on d4, bringing the e5-pawn out to d4 where it may be less vulnerable to attack.} (9... a4 {is the most popular continuation, which prepares to play ...Kh8 and ...f6. Play then typically continues with} 10. Ba2 Kh8 11. Re1 f6 12. d4 exd4 13. Nxd4 Ndb4 14. axb4 Qxd4 15. c3 Qxd1 16. Rxd1 {with a slightly better ending for White due to Black's weaknesses, though a draw is the most likely result, Kramnik,V (2715)-Kasparov,G (2805) New York 1995}) 10. Nxd4 { White exchanges off the troublesome d4-knight.} (10. Ba2 {prevents the exchange of the light-squared bishop and avoids bringing the e5-pawn out to d4, but at the cost of a little time.} Ra6 (10... Nf4 {is another interesting idea, e.g.} 11. Nc3 Nxh3+ 12. gxh3 Bxh3 {is quite a strong sacrifice.}) 11. Nxe5 (11. Re1 {is a sterner test of Black's rook-swing idea, although I don't think White has much of an advantage, e.g.} Re6 12. Bd2 Nxf3+ 13. Qxf3 c6 14. Nc3 Rf6 ) 11... Re6 12. Nf3 Ne2+ 13. Kh1 {and Black has reasonable compensation for the pawn, Michiels,B (2489)-Kashlinskaya,A (2360) Pardubice 2011}) 10... exd4 11. Re1 Ra6 12. Qf3 ({If} 12. Bxd5 Qxd5 13. Rxe7 Rg6 {and Black's kingside attack appears to be sufficient.}) 12... Re6 13. Rxe6 Bxe6 14. Nd2 c6 {Diagram [#] If anything I would rather be Black here because White's forces are botled up.} 15. Ne4 a4 16. Bc4 (16. Bxa4 Qa5 {[%cal Ra5e1] gives Black good counterplay due to the threat of ...Qe1+ and the attack on the bishop on a4.}) 16... b5 (16... Qa5 {is again worth considering, with the idea of} 17. Bd2 Qb6) 17. Ba2 Nc7 {Black appears to be happy to trade pieces and head for the draw in this game.} ({I would be strongly tempted by} 17... f5) 18. Bxe6 Nxe6 19. Ng3 Qd5 {Probably best here, since otherwise White has some hope of generating a kingside initiative with Ng3-f5.} 20. Qxd5 cxd5 {Black's impending pressure down the c-file compensates for the doubled and isolated d-pawns, and the players soon agreed to a draw.} 21. Bd2 Rc8 22. Rc1 g6 23. Ne2 f5 24. g3 Kf7 25. h4 Bf8 26. Kg2 h6 27. f4 Bg7 28. Ng1 Bf6 29. Nf3 Ke7 30. Kf2 {I think that at club level there would still be scope for further play in this position, but at grandmaster level it is hard for either side to make progress.} 1/2-1/2