Albin Counter-Gambit- Alternatives to the lines with a3 and/or Nbd2

Accepting the gambit and following up with a3 and/or Nbd2, with the aim of preventing ...Bb4+ and rounding up Black's d4-pawn, has become established as White's most critical response, but White has plenty of other responses, some of which are quite good, but all of which allow Black reasonable counterplay and at worst just a slight theoretical disadvantage.

White has various ways to decline the gambit at move 3 but none of them are convincing.  Black can generally choose between 3...exd4, which usually equalises, or the more ambitious approaches stemming from 3...e4, though these can leave Black's e-pawn weak.  If 3.cxd5 Qxd5 4.dxe5 Qxd1+ 5.Kxd1 Nc6 then Black gets very good compensation for the pawn in spite of the queen exchange. 

Lines with an early e2-e3

White should avoid 4.e3?! because of 4...Bb4+ 5.Bd2 dxe3!, with the idea of the underpromotion trick 6.Bxb4? exf2+ 7.Ke2 fxg1N+.  Instead White's best is 6.fxe3 and Black gets at least an equal game with 6...Qh4+ 7.g3 Qe4.  Slightly better is 4.Nf3 Nc6 5.e3, but following 5...Bb4+ 6.Bd2 dxe3 7.fxe3, but Black chooses between 7...Bb4+ and 7...Qe7 and gets a good game.

If White wishes to undermine Black's d4-pawn with e2-e3 and encourage a queen exchange on the d-file, it is better to play 4.a3 first, preventing ...Bb4+, and then 4...Nc6 5.e3 follows.  Here Black can get compensation for the pawn by playing 5...Be6 or 5...Bf5, whereupon 6.Nf3 dxe3 7.Qxd8+ Rxd8 8.Bxe3 follows.  However, White may be able to get an edge against 5...Bf5 in particular, because although the bishop on f5 gives Black queenside counterplay, particularly on the sensitive c2 and d3-squares, it takes the f5-square away from Black's knights.  Probably most reliable is 5...Nge7, intending 6.Nf3 Nf5 (or 6...Bg4) which gives Black good compensation for the pawn.

5.g3

After 4.Nf3 Black normally continues 4...Nc6, since 4...Bb4+ 5.Bd2 is now quite awkward for Black.  The old main line then runs 5.g3, which is probably good enough to give White a slight edge, but with some counterplay for Black.
Black can launch a crude kingside attack against this with ...Be6/f5/g4, ...Qd7 and ...0-0-0, with the idea of ...Bh3 exchanging off White's light-squared bishop and pushing the h-pawn.  

5...Bf5 is more reliable than 5...Bg4, because after 5...Bg4 6.Bg2 Qd7 7.0-0 0-0-0 8.Qb3! White intends Rd1, rounding up the pawn on d4.  In the same variation following 5...Bf5, the bishop on f5 takes some flight squares away from the white queen and so Black can get good play with 8...a6, and then if 9.Rd1?? then 9...Na5 wins the white queen.  5...Be6 can be met by 6.Bg2!? encouraging Black to take on c4, at the cost of time and the initiative, and envisaging meeting 6...Qd7 with 7.0-0 0-0-0 8.Qa4 although Black retains decent hacking chances on the kingside here.

A common position for this line arises after 5...Bf5 6.Bg2 Qd7 7.0-0 0-0-0 8.Nbd2 Bh3 (Note that if 8.Qb3 then Black should avoid 8...Bh3?! because then 9.e6 is strong, as White's queen on b3 and bishop on g2 both point at b7, but as noted above, 8...a6 is a good reply)

Objectively White may be slightly better but these positions are attractive for exponents of the black side who like to have the initiative and attack.  For example, 9.Qa4 Kb8 10.Rd1 Be7 prepares to play ...h7-h5-h4.

A rather different approach for Black is 5...Nge7, favoured by Alexander Morozevich, which generally intends ...Ng6 and winning the pawn back.  It can lead to complications after 6.Bg2 Ng6 7.Bg5, attempting to make it difficult for Black to regain the pawn on e5, for example 7...Qd7 8.e6!? fxe6 9.0-0 e5, or 8.0-0 h6 9.Bf4 Nxf4 10.gxf4 g5, with board-wide chaos and equal chances for both sides.  The main snag with this variation is that White can let the pawn go and strive for a riskless edge with 7.0-0, with the idea 7...Ngxe5 8.Nxe5 Nxe5 9.b3 (Ivanchuk-Morozevich, Astana 2012).

Games and analysis

The coverage is divided as follows:
1.  3.dxe5 d4 4.e3, and 3rd-move alternatives for White
2.  4.a3 Nc6 5.e3 Bf5 (or ...Be6)
3.  4.a3 Nc6 5.e3 Nge7
4.  4.Nf3 Nc6 5.g3 Bg4
5.  5.g3 Be6
6.  5.g3 Bf5
7.  5.g3 Nge7

[Event "Cappelle op 11th"] [Site "Cappelle la Grande"] [Date "1995.??.??"] [Round "9"] [White "Shumiakina, Tatiana"] [Black "Raetsky, Alexander"] [Result "0-1"] [ECO "D08"] [WhiteElo "2350"] [BlackElo "2510"] [PlyCount "88"] [EventDate "1995.03.??"] [EventRounds "9"] [EventCountry "FRA"] [Source "ChessBase"] [SourceDate "1996.11.15"] {Here I focus on coverage of White's attempt to trade off Black's d4-pawn with a quick e2-e3. Game 1 deals with the dubious 4.e3, as well as White's alternatives to 3.dxe5. In Games 2 and 3, I deal with the more critical 4.a3 Nc6 5.e3, which prevents ...Bb4+ and generally forces the d4-pawn and the queens off the board, but Black is able to generate enough tactical tricks to compensate for the pawn. Game 2 features 5...Bf5, which is interesting, but probably slightly better for White with best play (5...Be6 and 5...a5 are also discussed in the notes). Game 3 features 5...Nge7, which is probably the most reliable response.} 1. d4 d5 2. c4 e5 {Black sacrifices the e5-pawn. The idea is that White has weakened the b4 and d4-squares by playing c2-c4, and therefore Black can meet 3.dxe5 with 3...d4, causing White some discomfort and threatening ...Bb4+ in various lines. White should take the pawn, but occasionally prefers to try something else.} 3. dxe5 (3. cxd5 {Why not take the d5-pawn and bring the black queen out into the open? Unfortunately for White, the queen turns out to have considerable nuisance value. Instead, White can try to get back into regular Queen's Gambit lines by playing Nc3, Nf3 or e3, but all of these moves have the drawback that Black can instead take on d4, or play ...e5-e4, forcing the game into independent channels which are less likely to yield an advantage for White than if White takes on e5.} Qxd5 4. dxe5 {Tempting, as it exchanges queens and wins a pawn, but White ends up paying a heavy price by losing castling rights and struggling to find a safe place to tuck the king away.} (4. Nf3 {is probably best.} e4 $5 5. Nc3 Bb4 {then offers roughly equal chances.}) 4... Qxd1+ 5. Kxd1 Nc6 {Black has very good compensation for the pawn, for White is missing a pawn on c2 and therefore can't move the king to safety with c3 followed by Kc2. The e5-pawn is also in danger of dropping off.} 6. f4 {This is quite greedy, but it is unclear if White really has anything better than this.} (6. Nf3 {makes sense to me, protecting the pawn and developing a piece, but it scores poorly for White in practice, and doesn't really help White's king situation.} Bg4 7. Bf4 Nge7 8. Nbd2 Ng6 {is a logical continuation, where White will struggle to both keep the king defended and hold onto e5, and in general the e5-pawn tends to drop off.}) 6... Bf5 {Black has ideas of ... Nb4 or ...Nd4, creating threats against c2.} (6... Nh6 {is a good alternative, intending ...Nf5 and ...Be6.}) 7. Nc3 {White develops a piece and adds some protection to the e4-square. There are some tactical points which take the sting out of Black's potential threats to c2.} O-O-O+ {Black castles and continues to attack the white king.} 8. Bd2 f6 {, as played in Li Zunian (2378)-Wang Hao (2605) Suzhou 2006. Black challenges the e5-pawn and is not afraid to persist in sacrificing a pawn.}) ( 3. Nc3 exd4 4. Qxd4 (4. Nxd5 c6 5. Nf4 Nf6 {offers equal chances.}) 4... Nc6 5. Qxd5 Be6 {offers Black reasonable compensation for a pawn, although in practice Black often hasn't followed up accurately. For example,} 6. Qxd8+ Rxd8 7. Bg5 f6 8. Bf4 {I think Black's best way to get sufficient compensation for the pawn is probably the immediate} Bxc4 {, for example,} 9. Bxc7 Rd7 10. Bf4 Nb4 {with considerable threats against the white queenside, which force White to return the extra pawn by giving up the pawn on a2.}) (3. Nf3 {can also be met by} exd4 (3... e4 {is also worth considering, which forces White's f3-knight to a less-than-ideal square, though the downside is that Black's e4-pawn can become vulnerable to being undermined by the white forces.} 4. Nfd2 Nf6 5. e3 {is a logical continuation.}) 4. Nxd4 Nf6 5. cxd5 Qxd5 {with approximate equality.}) (3. e3 {can transpose to a French Defence, Exchange Variation after} exd4 (3... Nc6 {is also fine, transposing to a variation of the Chigorin Defence, 1.d4 d5 2.c4 Nc6 3.e3 e5, which is not too troublesome for Black.}) 4. exd4) 3... d4 {This is the main point of Black's gambit- White has weakened the b4 and d4-squares by playing c2-c4, and so Black takes the opportunity to get a strong-point on d4, which has a cramping effect on White's position.} 4. e3 $6 {White wishes to exchange off Black's d4-pawn and encourage a queen trade down the d-file, but this doesn't work because Black has a sneaky tactic.} (4. e4 {occupies the centre, but removes the threat of e2-e3 challenging the pawn on d4, so this line should not concern Black too much.} Nc6 5. f4 {White decides to try to hold onto the pawn, but Black gets very dangerous compensation, as White has played a couple of pawn moves too many in the opening.} (5. Nf3 Bg4 {is an irritating pin for White to deal with, especially as Black is threatening ...Nxe5 here, putting more pressure on the f3-knight. For example,} 6. Qb3 (6. Nbd2 Bb4) 6... Bb4+ 7. Nfd2 Qh4 8. a3 Bxd2+ 9. Nxd2 O-O-O {This complicated position offers Black good attacking prospects. } 10. g3 Qh5 11. f4 g5 $15 {Janowski,D-Marshall,F Suresnes 1908}) 5... f6 {A standard reaction, challenging White's e5-pawn and offering a dangerous gambit. } (5... Bb4+ 6. Bd2 Nh6 {intending ...Ng4 is a good alternative for Black, e.g. } 7. Bxb4 Nxb4 8. a3 Nc6 9. Nf3 Ng4 {Dubov, D (2638)-Potapov,P (2460) Khanty-Mansiysk 2013, although I think Black stands better in the final position.}) 6. exf6 $6 {Black's lead in development offers more than enough compensation for a pawn after this rather obliging capture.} (6. Nf3 fxe5 7. fxe5 Nge7 {intending ...Ng6 offers Black at least equality, but it is less difficult for White than what happens after White takes on f6.}) 6... Nxf6 7. Bd3 Bb4+ 8. Nd2 Ng4 9. a3 Ne3 10. Qe2 O-O $5 {Black chooses to sacrifice a piece in return for increased attacking chances against the white king.} (10... Bg4 {is a good alternative, where Black is probably better in spite of the pawn minus.}) 11. axb4 Nxb4 12. Ndf3 Nxg2+ 13. Qxg2 Nxd3+ 14. Kd1 Qe7 $44 { Markos,J (2596)-Hrabusa,M (2344) Czechia 2012}) 4... Bb4+ 5. Bd2 dxe3 $1 { White cannot safely take on b4, and must therefore accept a compromised pawn structure with f2xe3.} 6. fxe3 (6. Bxb4 exf2+ {is deadly because White cannot take on f2 because of the attack on the queen on d1, and the underpromotion} 7. Ke2 fxg1=N+ $1 {is a key resource for Black, which forces White to accept the loss of a piece due to Black's various threats against the white queen.} 8. Ke1 (8. Rxg1 $2 Bg4+ {picks up the white queen.}) 8... Qh4+ 9. Kd2 Nc6 {is easily winning for Black, Klocker,B (2253)-Konrad,E (2190) Oberwart 2012}) 6... Qh4+ { Black decides to create further disruption in the white kingside.} (6... Bxd2+ 7. Qxd2 {is less appealing for Black as Black cannot avoid trading queens without ceding a significant share of the initiative to White.}) 7. g3 Qe4 8. Qf3 (8. Nf3 Qxe3+ 9. Qe2 {tries to get the queens off, but Black can back out of the queen trade and achieve an equal possition with} Bxd2+ 10. Nbxd2 Qb6 $11 {Farkas,A (1854)-Marodi,M (2158) Miskolc 2004}) 8... Bxd2+ 9. Nxd2 Qxe5 {White has a slight lead in development, but the white position is full of weaknesses, so Black should be at least equal here.} 10. O-O-O (10. Qe4 {is more popular, forcing the queens off the board, but Black probably has a slight edge following} Nc6 {, for example,} 11. Qxe5+ Nxe5 12. Ngf3 (12. O-O-O $6 Ng4) 12... Ng4) 10... Nf6 {Preventing White from continuing with 11.Qe4, which would give White an improved version of the queenless middlegame that follows 10.Qe4.} 11. Qf4 Qe7 {Black would rather not trade queens in a way that improves White's pawn structure after the recapture on f4.} 12. Ngf3 O-O {This helps to unbalance the position, castling on the opposite side of the board to White.} 13. Bd3 Ng4 {A common plan in this variation, threatening a nasty fork on f2, although there was also a strong argument for developing with ...Nc6 or ...Re8.} 14. Ng5 $6 (14. Ne4 {is stronger.}) 14... h6 15. Nge4 (15. Nh7 $6 Rd8 {looks aggressive for White but White has no follow-up and has to spand a tempo dealing with the attack on the d3-bishop.}) 15... Nc6 {Compared with the position after 14.Ne4, White's other knight is on d2, rather than f3, which improves Black's chances.} 16. Rde1 Be6 17. Bb1 Rad8 {Black is now clearly better, and Black went on to win the game.} 18. h3 Nge5 19. g4 $6 f5 20. Ng3 g5 {It is unusual to see a pawn storm against the opposing king refuted by advancing pawns on the same wing, in front of your own king, in this type of position, but the point is that White doesn't have enough behind the attack, and Black is able to pick off the kingside pawns.} 21. Qf1 fxg4 22. Qg2 Nb4 23. Ref1 Ned3+ 24. Bxd3 Nxd3+ 25. Kb1 Qg7 26. Nde4 gxh3 27. Qc2 Nb4 28. Qb3 Nc6 29. Qc2 Bg4 30. a3 Kh8 31. Nf2 Bf3 32. Nxh3 Ne5 33. Rh2 Kg8 34. Nf2 Ng4 35. Nf5 Rxf5 36. Nxg4 Bxg4 37. Rxf5 Bxf5 38. Qxf5 Qh7 39. e4 Qxf5 40. exf5 Kg7 41. Re2 Rf8 42. Re5 h5 43. Kc2 Kf6 44. Ra5 h4 0-1 [Event "Petersburg Summer rapid"] [Site "St Petersburg"] [Date "2013.09.01"] [Round "8"] [White "Aleksandrov, Aleksej"] [Black "Moiseenko, Vadim"] [Result "1/2-1/2"] [ECO "D07"] [WhiteElo "2626"] [BlackElo "2465"] [PlyCount "50"] [EventDate "2013.08.31"] [EventType "rapid"] [EventRounds "11"] [EventCountry "RUS"] [Source "ChessBase"] [SourceDate "1996.11.15"] 1. d4 d5 2. c4 e5 (2... Nc6 3. e3 e5 4. dxe5 d4 5. a3 {was the move-order actually used in the game to get to the position after White's fifth move, but it is easier for purposes of this article to stick with the conventional move-order.}) 3. dxe5 d4 4. a3 {A refinement of the idea behind the dubious 4. e3 of the previous game. White prevents ...Bb4+ and will then follow up with e2-e3, with the aim of exchanging off Black's d4-pawn and encouraging a queen trade down the d-file, leaving White a pawn ahead in a simplified position. It sounds good, but in reality, it isn't one of the most critical tests of the Albin Counter-Gambit.} Nc6 5. e3 {The point behind White's fourth move, although 5.Nf3 is objectively superior, as it keeps White's options open, and is covered under the line 4.Nf3 Nc6 5.a3.} Bf5 $5 {This is quite a tricky line, as White has to deal with potential threats against d3 and c2, although with accurate play White should be able to get some advantage against it, despite Black having scored well in practice. The bishop is actively placed here, and can support threats on c2 and d3, but the downside is that it blocks the possibility of putting a knight on f5. Probably most reliable is 5...Nge7, with the idea of playing the knight to f5 to reinforce the d4-pawn, as features in the next game.} (5... Be6 {is probably a slight improvement over the 5...Bf5 line because the g8-knight can go to f5 here.} 6. Nf3 dxe3 7. Qxd8+ Rxd8 8. Bxe3 Nge7 (8... Nh6 {intending ...Ng4 is also worth considering.}) 9. Nc3 Nf5 {Black goes after the bishop on e3 with the aim of wrecking White's pawn structure, rather than going after the e5-pawn.} 10. Nb5 Rd7 11. Rd1 Nxe3 12. fxe3 g6 {with equal chances for both sides, as White's poor pawn structure heavily devalues the pawn plus. Agrest,E (2605)-Sadler,M (2626) Germany 2003}) (5... a5 {is quite a popular alternative, but it usually leads to a queen trade anyway, and White may be able to get an edge against it.} 6. Nf3 Bc5 7. exd4 Bxd4 8. Be2 ({Trading queens with} 8. Nxd4 Qxd4 9. Qxd4 Nxd4 {leaves White with difficulties on the queenside light squares, e.g.} 10. Ra2 Bf5 11. Nc3 O-O-O 12. Be2 Be6 $44 {Lund,D (2270)-Holst,P Hedehusene 1992}) 8... Nge7 9. O-O O-O 10. Nc3 Bxe5 11. Nxe5 {and the players agreed a draw in Stern,R (2490) -Rabiega,R (2508) Berlin 2010, where I think White stands better in the final position:} Nxe5 12. f4) 6. Nf3 dxe3 (6... Bg4 {amounts to a loss of tempo, and White is better after} 7. Be2 dxe3 8. Qxd8+ Rxd8 9. Bxe3) 7. Qxd8+ Rxd8 8. Bxe3 {White has exchanged queens and has kept an extra pawn, but Black has active pieces and White has to take care with the weak light squares on the queenside. White often has to be prepared to return the extra pawn in this line in order to get any advantage.} Nge7 {Black intends simply to play ...Ng6 and win the pawn back.} (8... f6 {is the computer's preference, but I doubt that Black can get enough compensation for the pawn with this approach.} 9. exf6 Nxf6 10. Be2 {looks good for White, for example,} Ne4 11. Nbd2 Nc5 12. b4 Nd3+ 13. Bxd3 Rxd3 14. O-O $14) (8... Nh6 {is an interesting alternative, intending ...Ng4 harassing the bishop on e3. With best play, White is slightly better but Black has counterplay on the queenside light squares.} 9. h3 {is White's only try for advantage, preventing the knight from hopping into g4.} Bc2 10. Nc3 Nf5 11. Bg5 Rd7 12. Rc1 Bb3 13. Be2 $14 {Foord, D (1950)-Lyell,M (2187) Great Yarmouth 2007}) 9. Nc3 {With the threat of Nc3-b5, attacking c7 and a7.} Ng6 (9... a6 { is a good alternative, preventing Nc3-b5, and should enable Black to regain the pawn in most lines, although White tends to end up with the more active pieces, e.g.} 10. Be2 Ng6 11. Nd5 (11. O-O Ngxe5 12. Nxe5 Nxe5 13. Rfd1 Be7 14. Rxd8+ Kxd8 15. Rd1+ Kc8 {leads to a roughly level position, Holland,D (1963) -Abbott,P (1893) Parramatta 2009}) 11... Rd7 12. O-O-O Ngxe5 13. Nxe5 Nxe5 14. Rhe1 {with a slight advantage for White.}) 10. Nb5 (10. Nd5 {intending} Rd7 11. O-O-O {is a good alternative for White.}) 10... Rd7 11. Nfd4 Nxd4 (11... Nge7 { might be a slight improvement, reinforcing the knight on c6 rather than exchanging on d4. For example,} 12. Rd1 Nxe5 (12... a6 {is objectively superior.}) 13. Nxf5 Nxf5 14. Rxd7 ({White might be able to get away with the rather greedy} 14. Bxa7 b6 15. Bb8 Rxd1+ 16. Kxd1 Ng4 17. Ke1 {but I can see why White would have been reluctant to go in for this.}) 14... Kxd7 15. Bf4 ( 15. Bxa7 b6 {leaves White's bishop shut out of play.}) 15... Ng6 16. Bd3 Nfh4 { with equal chances, Horcajada Reales,G (2207)-Santos,J (2265) Madrid 2012}) 12. Nxd4 Bg4 {White has some advantage in this position, but White's next move is probably imprecise.} 13. h3 (13. f3 {is stronger, and gives White a significant plus, which is a strong reason for Black to look at earlier alternatives.}) 13... Bc5 {Due to Black's pressure against d4 and e5, White's best option is to give up the e5-pawn by pushing it to e6.} 14. e6 fxe6 15. Nc2 $6 (15. Nxe6 Bxe3 16. hxg4 Bd2+ {gives Black reasonable counterplay.}) (15. Nb3 {would have been better though. White would have kept some advantage after} Bxe3 16. fxe3 Bf5 17. Nc5) 15... Bxe3 16. Nxe3 Bf5 17. Rd1 Ne5 18. Rxd7 Nxd7 19. f3 {This position is equal, and the players soon agreed to a draw.} Ke7 20. Kd2 Rd8 21. Kc3 Nc5 22. Be2 Na4+ 23. Kb3 Nc5+ 24. Kc3 Na4+ 25. Kb3 Nc5+ 1/2-1/2 [Event "FRA-chT2 West 0809"] [Site "France"] [Date "2008.11.23"] [Round "3.2"] [White "Vareille, Francois"] [Black "Goldsztejn, Gildas"] [Result "0-1"] [ECO "D07"] [WhiteElo "2406"] [BlackElo "2415"] [PlyCount "80"] [EventDate "2008.10.19"] [EventRounds "11"] [EventCountry "FRA"] [Source "ChessBase"] [SourceDate "1996.11.15"] [WhiteTeam "Le Chesnay"] [BlackTeam "Stade Francais Jeen"] [WhiteTeamCountry "FRA"] [BlackTeamCountry "FRA"] 1. d4 d5 2. c4 e5 3. dxe5 d4 4. a3 Nc6 5. e3 Nge7 {I think this is Black's best bet in this position, intending ...Nf5 to reinforce the pawn on d4.} 6. Nf3 Nf5 (6... Bg4 {is a good alternative for Black, which gives Black the useful extra tempo ...Ng8-e7 over the line 5...Bf5 6.Nf3 Bg4.} 7. Be2 dxe3 8. Qxd8+ Rxd8 9. Bxe3 Nf5 $44 {is very acceptable for Black.}) 7. exd4 {White takes the opportunity to remove Black's annoying d4-pawn.} (7. Be2 dxe3 8. Qxd8+ Kxd8 ({Or} 8... Nxd8 9. fxe3 Nc6 10. Nc3 Be6 11. Nd5 O-O-O 12. b4 $14 { is not very appealing for Black, Grabuzova,T (2379)-Muzychuk,M (2387) Dagomys 2008}) 9. Bxe3 Nxe3 10. fxe3 {leaves Black with reasonable compensation for a pawn because of White's poor pawn structure.}) (7. e4 {looks strong at first glance but Black then has the strong} Nh4 $1) 7... Nfxd4 8. Be3 {White puts pressure on Black's knight, encouraging the knight to move and allow a queen swap.} Bg4 {Black pins the f3-knight, helping to reduce the pressure on the d4-knight.} 9. Nbd2 {White reinforces the f3-knight, but now the d4-knight is in less danger.} Bc5 10. Be2 Nxf3+ {This initiates complications and a series of exchanges.} 11. Bxf3 (11. Nxf3 Bxe3 12. Qxd8+ Rxd8 13. fxe3 Ke7 {(I think this may be better than giving up bishop for knight and taking on e5) is a little better for Black in spite of the pawn minus. Black will bring a rook to e8 and will tuck the king away on f8. White will find it very hard to hold onto the weak isolated e-pawns.}) 11... Bxe3 12. Bxg4 ({I think Black has an edge following} 12. fxe3 Qh4+ 13. g3 Bxf3 14. Qxf3 Qe7) 12... Nxe5 13. Be2 Bxf2+ $5 {Black decides to sacrifice a piece for attacking chances, a good practical decision although probably not objectively the best.} (13... Bd4 {is slightly better for Black, who has the more active pieces.}) 14. Kxf2 Qh4+ 15. Kf1 (15. g3 $2 {weakens the white kingside and leaves Black with more than enough compensation for the piece after} Qd4+ 16. Kf1 O-O-O) 15... O-O-O 16. Qc2 {White gets out of the pin on the d2-knight.} Qf4+ 17. Nf3 ({Not} 17. Ke1 $6 Rhe8) 17... Ng4 18. g3 $6 {White cracks, and plays this weakening move after all, which allows Black to get the piece back and simplify into a pawn-up ending.} (18. Qc3 {was White's best way to cover the threat of ...Qe3 and ...Qf2 mate. However, Black still has good compensation for the piece following} Rhe8) 18... Ne3+ 19. Kf2 Qxf3+ 20. Bxf3 Nxc2 {Now Black is a pawn up in the endgame, and goes on to win. Black's technique for the remainder of the game appears quite good, causing some problems with the knight, and then swapping off the rooks, which makes it easier to exploit the extra pawn.} 21. Rad1 Nd4 22. Bd5 Nb3 23. Bg2 Rhf8 24. g4 c6 25. Be4 g6 26. h4 Nc5 27. Bc2 Rxd1 28. Rxd1 Rd8 29. Rxd8+ Kxd8 {With the rooks off the board, this should be a win for Black, and White was unable to prevent Black from making use of the extra pawn.} 30. Kf3 Ke7 31. Kf4 Ne6+ 32. Ke5 f6+ 33. Ke4 Kd6 34. Bb1 Ng7 35. Ke3 f5 36. h5 Ke5 37. h6 Ne6 38. b4 Ng5 39. gxf5 gxf5 40. c5 Nf7 0-1 [Event "Khanty-Mansiysk ol (Men) 39th"] [Site "Khanty-Mansiysk"] [Date "2010.09.27"] [Round "6.2"] [White "Amir, Karim"] [Black "Abu Sufian, Sakil"] [Result "1-0"] [ECO "D09"] [WhiteElo "2234"] [BlackElo "2361"] [PlyCount "95"] [EventDate "2010.09.21"] [EventRounds "11"] [EventCountry "RUS"] [Source "ChessBase"] [SourceDate "1996.11.15"] [WhiteTeam "Pakistan"] [BlackTeam "Bangladesh"] [WhiteTeamCountry "PAK"] [BlackTeamCountry "BAN"] {Although most sources tend to recommend 4.Nf3 Nc6 5.a3 or 5.Nbd2 against the Albin Counter-Gambit these days, the old main line 5.g3 is still popular and probably sufficient to give White an edge with best play. Black can either try to round up the e5-pawn with ...Nge7-g6, or play for a crude kingside attack with ...Be6/f5/g4, ...Qd7 and ...0-0-0. Games 1, 2 and 3 deal with 5...Be6, 5.. .Bg4 and 5...Bf5 respectively. Despite 5...Bf5 being less common than the other two bishop moves, I think it is probably more promising, for reasons that I explore in the notes to Game 3. Finally, Game 4 features Alexander Morozevich using his preference of 5...Nge7, which is probably Black's most reliable option.} 1. d4 d5 2. c4 e5 3. dxe5 d4 4. Nf3 {This is probably White's best fourth move, developing a piece and attacking d4.} Nc6 {Black should simply develop and reinforce d4. Checking on b4 doesn't particularly help Black here because White blocks with the bishop on d2, and then sets about rounding up the d4-pawn.} 5. g3 {White prepares to fianchetto the king's bishop on g2 and put Black under pressure on the h1-a8 diagonal. This can also help White to generate a powerful queenside attack if Black opts to castle queenside and aim for kingside play, though the downside is that the kingside fianchetto formation leaves White open to ...Be6/f5/g4, .. .Qd7, ... Bh3 and the ...h7-h5-h4 lever. White also has a few alternatives here.} (5. Bf4 { develops and reinforces e5, but Black can get good compensation for the pawn starting with} Nge7 {intending ...Ng6 hitting the bishop on f4.} 6. Nbd2 (6. e3 Ng6 {is a worse version for White, since} 7. Bg3 {runs into} Bb4+ 8. Nbd2 dxe3) 6... Ng6 7. Bg3 Bf5 $44 {intending ...Qd7 and ...0-0-0 is one likely continuation.}) (5. e3 {has a similar objection to the immediate 4.e3 but is not quite as bad for White.} Bb4+ 6. Bd2 {still allows the tactical trick} dxe3 {which White must answer with} 7. fxe3 (7. Bxb4 {leads to a better position for Black following} exf2+ 8. Ke2 Qxd1+ 9. Kxd1 Nxb4) 7... Bxd2+ (7... Bg4 8. Bxb4 Qxd1+ 9. Kxd1 Nxb4 {is the main alternative, where it will be hard for White to hold onto the extra pawn.}) (7... Qe7 {is an interesting alternative, intending} 8. Nc3 Be6 {preparing queenside castling and avoiding any queen trades.}) 8. Qxd2 (8. Nbxd2 Nge7 {followed by ...Ng6 offers equal chances.} ( 8... Qe7 {is a good alternative, if Black would rather be assured of avoiding an early queen exchange.})) 8... Bg4 {is the other main alternative for Black, with the idea that if White trades queens then Black recaptures with the rook. White's isolated e-pawns are a problem. Peter Leisebein won a nice game from this position:} (8... Qe7 9. Nc3 {followed by Nd5 is a problem for Black.}) 9. Be2 (9. Qxd8+ Rxd8 {is OK for Black}) 9... Bxf3 10. Bxf3 Qxd2+ 11. Nxd2 Nxe5 { and White is saddled with a weak isolated e-pawn, Nittschalk,K-Leisebein,P TSC corr 2002}) 5... Be6 {This is the most popular move for Black in the database, just ahead of 5.. .Bg4, although 5...Nge7 has overtaken it in recent years. The move prepares .. .Qd7 and ... 0-0-0 with the idea of ...Bh3 and a h-pawn hack, while the bishop also attacks the pawn on c4.} (5... f6 $6 {is reminiscent of the Blackmar-Diemer Gambit, and is one of the computer's top preferences, but Black will struggle to get enough for the pawn after} 6. exf6 $1 Nxf6 7. Bg2 {Black's half-open f-file is not of much use in this variation (Black tends to rely more upon opening the h-file) and in some lines the king's knight would be better off going to f5 or g6 via e7.} Bf5 8. O-O Qd7 9. Qb3 {This move has a large plus score for White in the database.} Bc5 (9... Na5 10. Qb5 Qxb5 11. cxb5 O-O-O 12. Rd1) (9... O-O-O 10. Ne5 {is very good for White because of the strong pressure against b7 and c6, and the knight on e5 cannot be taken.}) 10. Bf4 {with advantage for White.}) 6. Bg2 $5 {This is rarely played, but returning the pawn on c4 makes sense because if Black takes, then White gains one or two tempi and opens up the c-file. Instead, White normally plays 6.Nbd2 to defend the pawn on c4.} (6. Nbd2 {is typically met by the standard attacking plan.} Qd7 7. Bg2 {Continuing with the quick fianchetto and kingside castling rather complies with Black's attacking plans, but the white king is generally safer on the kingside than in the centre or on the queenside, where there is comparatively little pawn cover and Black's pressure down the d-file can often be an issue.} ({However,} 7. a3 {is an interesting alternative, intending to press ahead immediately on the queenside with b4, and has only been played in one low-level game in my database. For example,} Nge7 (7... O-O-O 8. b4) 8. b4 Ng6 9. Bg2 Ngxe5 10. b5 Nxf3+ 11. Bxf3) 7... O-O-O ({The immediate} 7... Bh3 {is also playable, for White has nothing better than castling short, whereupon} 8. O-O O-O-O {transposes to the line 7.. .0-0-0 8.0-0 Bh3.}) 8. O-O Bh3 {Diagram [#] This is quite a common position in the 5.g3 line of the Albin Counter-Gambit and is strongly relevant to all three of Black's bishop moves at move 5. Many games from here tend to be combative with both sides attacking the enemy king.} (8... h5 $5 9. Qa4 (9. h4 Nh6 {intends to invade on g4.}) 9... Kb8 10. Rd1 Be7 $14 {intends to push ... h5-h4. Pink, F-Leisebein,P (2122) DESC email 2001}) 9. Qa4 (9. a3 Bxg2 10. Kxg2 Nge7 {with . ..Ng6 and ...h5 to follow gives Black good hacking chances,}) (9. e6 $5 Bxe6 10. Ng5 Nf6 11. Nxe6 Qxe6 {gives approximately equal chances, since White has gained the bishop-pair but has lost the extra pawn, and Black can still consider the h-pawn hack.}) 9... h5 (9... Bxg2 10. Kxg2 h5 {is most likely an improvement over the game.}) 10. b4 $5 {A pawn sacrifice to open up lines against Black's king.} Bxg2 11. Kxg2 h4 $6 ({Although it looks dangerous, Black should probably have taken the pawn here:} 11... Bxb4 12. Rb1 Bxd2 13. Bxd2 h4) 12. b5 Nb8 13. Ba3 $6 (13. Qxa7 hxg3 14. fxg3 $16) 13... Bxa3 14. Qxa3 hxg3 15. fxg3 $14 {Dao Thien Hai (2506)-Liew Chee Meng,J (2287) Manila 2013}) 6... Bxc4 {Black takes the pawn.} ({It might be better for Black to decline the offer with} 6... Qd7 {, for example,} 7. O-O O-O-O 8. Qa4 (8. Qb3 Na5) (8. Nbd2 {transposes to 6.Nbd2}) 8... Bh3 9. Rd1 Bxg2 10. Kxg2 h5 $14 {Diagram [#] with attacking chances, for example,} 11. Bg5 $6 (11. Nc3 {intending Nb5 is more threatening for Black, getting on with the queenside offensive.}) 11... Be7 12. Bxe7 Ngxe7 13. h4 {Preventing ...h5-h4, but now the g4-square is weak.} Kb8 14. Na3 Qg4 $15 {Schilke,M (1954)-Wasem,R (2106) Berlin 2010}) 7. O-O (7. Nbd2 {is a good alternative, gaining time on the c4-bishop immediately, though it will often transpose to the game continuation.}) 7... Qd7 8. Nbd2 Be6 9. Ne4 {Diagram [#] White has a slight advantage in this position as White's pieces are the more active. Black's d4-pawn is vulnerable, although so is White's e5-pawn.} (9. Qa4 {is another good option, preparing to attack on the queenside and/or play Rd1 and Nb3, ganging up on the d4-pawn.}) 9... Bd5 10. Ne1 {This defends the knight on e4 and has the idea of Nd3 followed by Nc5, but I would rather continue developing with 10.Qc2.} Qf5 (10... O-O-O 11. Nd3 { gives White a slight advantage as White has more attacking prospects on the queenside than Black does on the kingside.}) 11. Qc2 Qxe5 $6 {This takes the e5-pawn prematurely because White now gets to play Bf4 and take on c7.} (11... Nge7 {followed by queenside castling gives Black a reasonable game.}) 12. Bf4 Qe6 (12... Qe7 $6 {blocks the f8-bishop's coverage of the d6-square and allows the tactic} 13. Nd6+ cxd6 14. Bxd5) 13. Bxc7 Rc8 14. Bf4 Nb4 15. Qb1 {The queen is passively placed here but can soon emerge after White starts to expand on the queenside.} Be7 16. a3 Nc6 17. Qd3 Nf6 $6 {Black had a few better moves here.} (17... Bc4 18. Qc2 Nf6 {would have been better because now} 19. Ng5 $2 {is met by} Qxe2) 18. Ng5 Qd7 19. Bh3 Be6 20. Nxe6 fxe6 21. Nf3 O-O $6 {This castles into quite a strong attack.} (21... Nd5 22. Qe4 $16) 22. Ng5 Nd8 23. Be5 Bc5 24. Rac1 h6 25. Bxf6 hxg5 26. Bxg5 {White now has an extra pawn and a positional advantage, so is clearly winning at this stage. During the remainder of the game, White went relatively passive for a while but gradually simplified and maintained the extra pawn, and eventually the kingside attack returned and forced a win.} Nf7 27. Bf4 Bb6 28. e3 Rcd8 29. Rfd1 Qe7 30. exd4 Bxd4 31. Qe2 e5 32. Be3 Ng5 33. Bg2 Ne6 34. Re1 Qf7 35. Rcd1 Nc5 36. b4 Ne6 37. Be4 Bxe3 38. Qxe3 Nd4 39. Rf1 b6 40. h4 Qh5 41. Kg2 Kh8 42. Qg5 Qxg5 43. hxg5 Rd6 44. Rc1 Ne6 45. g6 Rfd8 46. Rh1+ Kg8 47. Rh5 Nf8 48. Rch1 {Black has to concede further material to avoid checkmate.} 1-0 [Event "Oberliga NRW 9900 II"] [Site "Germany"] [Date "2000.01.19"] [Round "4.1"] [White "Roeder, Mathias"] [Black "Hubert, Ralf"] [Result "1-0"] [ECO "D09"] [WhiteElo "2445"] [BlackElo "2355"] [PlyCount "83"] [EventDate "1999.10.26"] [EventRounds "9"] [EventCountry "GER"] [Source "ChessBase"] [SourceDate "1996.11.15"] [WhiteTeam "Aachen"] [BlackTeam "Solingen II"] [WhiteTeamCountry "GER"] [BlackTeamCountry "GER"] 1. d4 d5 2. c4 e5 3. dxe5 d4 4. Nf3 Nc6 5. g3 Bg4 {This move is a close second in popularity after 5...Be6. From g4, the bishop puts pressure on the f3-knight, providing the option of exchanging bishop for knight and reducing the pressure on the d4-pawn (although in many lines Black would rather advance the bishop to h3 and exchange off White's king's bishop). The bishop can potentially be kicked with h2-h3 in some variations.} 6. Bg2 (6. Nbd2 Qd7 7. Bg2 O-O-O 8. h3 $5 {is an interesting option, trying to take advantage of the position of Black's bishop on g4, although White has to take some care as this leaves the h3-pawn weak.} Bf5 9. a3 {This move intends to leave the white king in the middle, while launching a queenside pawn storm.} Nge7 (9... f6 {is a reasonable alternative.}) 10. b4 Ng6 11. Qa4 Kb8 12. Bb2 Qe8 $6 (12... d3 {is probably Black's best option, after which White has no more than a slight edge in a complicated position.} 13. b5 Ncxe5 14. Nxe5 Nxe5 15. Bxe5 Qe7 $13) 13. O-O $1 {is better for White, Servaty,R (2153)-Schebler,G (2482) Recklinghausen 2003}) 6... Qd7 7. O-O (7. a3 {is an interesting idea, trying to press forward immediately on the queenside, but I don't think that it works so well against the 5...Bg4 line.} O-O-O 8. b4 {is met by} Bxf3 {followed by ...Nxe5 with a reasonable position for Black, so at move 8 it is most likely better just to castle, but then White's best line against 5...Bg4 is ruled out.}) 7... O-O-O ( 7... Bh3 $6 {is premature here because of the response} 8. Bxh3 Qxh3 9. e3 $16 {for Black does not have enough control over the d4-square to be able to justify moving the queen away from the d-file yet. Black should wait until having castled queenside.}) 8. Qb3 $1 {White senses that Black is preparing to trade the light-squared bishops with 8...Bh3, and so puts pressure on b7, setting up the idea of meeting ...Bh3 with e5-e6 followed by Nf3-e5. The move also prepares Rd1, adding pressure to the pawn on d4.} (8. Nbd2 {gives Black a pleasant choice between 8...h5, 8...Nge7, and 8...Bh3 transposing to a line examined in the first game (5...Be6 6.Nbd2 Qd7 7.Bg2 0-0-0 8.0-0 Bh3).}) 8... Nge7 {Black prepares to bring the knight to either g6, attacking e5, or f5, defending d4.} (8... Bh3 $6 {is the most popular move, but as observed in the note to 8.Qb3, White gets the upper hand with} 9. e6 $1 Bxe6 10. Ne5 {which is irritating for Black due to the threats to c6 and b7.}) (8... h5 9. Rd1 h4 { doesn't work out very well for Black:} 10. Nxd4 $16) 9. Rd1 {Diagram [#] This move, putting pressure on d4, gives White some advantage.} Qf5 $6 {This leaves Black open to a powerful piece sacrifice.} (9... Bxf3 10. Qxf3 Ng6 11. Qh5 $14 {gives White the better chances, even though the e5-pawn is likely to drop off in the long run. White is threatening the disruptive Bg2-h3, pinning Black's queen against the king, Kortschnoj,V (2665)-Veinger,I (2200) Beersheba 1978}) 10. Na3 {Safe and sound, and good enough for a slight advantage, but I think that White has a way to break through on the queenside here.} (10. Nxd4 $1 Rxd4 11. Rxd4 Nxd4 12. Qxb7+ {is a very strong piece sacrifice which may well be close to winning for White. For example,} Kd8 13. Nc3 Nxe2+ 14. Nxe2 Bxe2 15. Be3 Qxe5 16. Bf4 Qc5 17. b4 Qb6 18. Qa8+ Kd7 19. Re1 {1-0 Offenborn, H-Benicke, B GER email 2008}) 10... Ng6 11. Nc2 {White reacts to the attack against e5 by putting more pressure on d4.} Bxf3 {This move is pretty near forced- otherwise White grabs the pawn on d4 and ends up a pawn ahead with the better position.} 12. exf3 Ngxe5 13. f4 {White has some advantage in this position as Black's d4-pawn has turned into more of a weakness than a strength.} Na5 {Black correctly counterattacks against c4.} 14. Qa4 Nexc4 15. Nxd4 b5 ({Trading queens was not forced in this position and Black could also have considered} 15... Qf6 {and this may well have been a better bet for Black as it keeps Black's queenside pawn structure in reasonable shape. However, White is still better in this position and has good attacking chances on the queenside.}) 16. Nxf5 bxa4 {The main problem with this line for Black is that the a4-pawn becomes very weak and soon drops off.} 17. Rxd8+ Kxd8 18. Ne3 Bc5 19. Nxc4 Nxc4 20. Bd5 Nd6 21. Bc6 Ke7 22. Bxa4 {Thus White goes a pawn ahead, though Black has some drawing chances as White's doubled pawns on the kingside devalue the pawn plus.} Rb8 23. Bc2 f5 24. b3 Bd4 25. Rb1 Ke6 26. Be3 c5 27. Re1 Kd5 $2 { This loses another pawn.} (27... Kf6 {was probably best but White is still a strong favourite.}) 28. Bxd4 cxd4 29. Re5+ Kc6 30. Bxf5 Nxf5 31. Rxf5 Kd6 32. Kf1 Rc8 33. Ke2 Rc2+ 34. Kd3 Rxa2 35. Rf7 Rxf2 36. Rxg7 h5 37. Rxa7 h4 38. Kxd4 Rxh2 39. Ra6+ Kd7 40. g4 Rb2 41. Rh6 Rh2 42. Ke5 {In view of the notes to this game, I don't think 5...Bg4 is as reliable as 5...Be6.} 1-0 [Event "Samara Polugaevsky Memorial 3rd"] [Site "Samara"] [Date "2013.07.07"] [Round "6"] [White "Tunik, Gennady"] [Black "Frolov, Ivan"] [Result "1/2-1/2"] [ECO "D09"] [WhiteElo "2435"] [BlackElo "2327"] [PlyCount "103"] [EventDate "2013.07.02"] [EventRounds "9"] [EventCountry "RUS"] [Source "ChessBase"] [SourceDate "1996.11.15"] 1. d4 d5 2. c4 e5 3. dxe5 d4 4. Nf3 Nc6 5. g3 Bf5 $5 {This is the least popular of Black's three bishop moves, but it may well be the most promising alternative to the main line with 5...Nge7. From f5, the bishop covers the key squares d3 and c2, which can leave the white queen short of squares in some variations.} 6. Bg2 Qd7 7. O-O {This is almost invariably played.} ({Again the attempt at an immediate queenside attack with} 7. a3 O-O-O 8. b4 {is not too convincing here.} d3 {may be the best way for Black to generate some trouble and take advantage of White's reluctance to castle early.}) 7... O-O-O {As with the 5...Be6 and 5...Bg4 variations, Black intends to swap off the light-squared bishops with ...Bf5-h3 and push the h-pawn, with the aim of opening up the h-file.} 8. a3 (8. Qb3 {is the most critical move when Black's bishop is on g4 rather than on f5, but with the bishop placed on f5, taking the c2 and d3-squares away from the white queen, Black has one way to get a good game here.} a6 $1 {looks to me to be the right move, despite having only been played twice in my database and with a score of 0/2. Black prevents 9. Rd1 due to 9...Na5 which would trap the white queen, and also prevents any ideas based on Qb5.} (8... Bh3 $6 {is the most popular move, but as observed in the variation with 5...Bg4, White gets the upper hand with} 9. e6 $1 Bxe6 10. Ne5 {with various threats against c6 and b7.}) 9. Bd2 Bc5 {#[Position after 9...Bc5] This gives a complicated position in which Black has, at worst, a small disadvantage, with long-term attacking prospects on the kingside. For example,} 10. Rd1 Nge7 11. Na3 Be4 12. Nc2 Qf5 13. Nce1 Ng6 {[%eval -32767,0] with a complex position offering equal chances, Nenciulescu,S (1909)-Fetisov,I (1504) Lechenicher SchachServer 2010}) ({If} 8. Nbd2 {then} Bh3 {is fully playable, and transposess to the line 5... Be6 6.Nbd2 Qd7 7.Bg2 0-0-0 8.0-0 Bh3.} (8... h5 {also gives Black just a small theoretical disadvantage with some counterplay, e.g.} 9. Qa4 Kb8 10. Rd1 Be7 {intending ...h5-h4.})) 8... Nge7 {With the idea of ...Ng6, putting pressure on e5 and also supporting the . ..h7-h5-h4 pawn push.} (8... Bh3 {is not quite as strong as it is against 8. Nbd2 because White can ignore Black's kingside threats and press forward with a queenside offensive.} 9. b4 Bxg2 10. Kxg2 {#[Position after 10.Kxg2]} h5 $6 ( {Black should probably cause some disruption in the centre with} 10... Qe6 { before going ahead with pushing the h-pawn.}) 11. b5 Nce7 12. h4 (12. Qa4 Kb8 13. Rd1 {is very good for White.}) 12... Ng6 13. Qd3 Nh6 {A common theme in Albin Counter-Gambit lines where White meets ...h7-h5 with h2-h4. White prevents the h-pawn from pushing any further but leaves the g4-square weak and is vulnerable to piece sacrifices on h4 to open up White's king.} 14. e3 $6 ( 14. Bg5 {should give White an edge.}) 14... Qg4 15. exd4 $6 Nf5 {with a dangerous attack on the kingside, Bellon Lopez,J (2470)-Cirabisi, F Genova 1989 }) 9. Bg5 $6 {I don't think this is particularly strong because this bishop move is an invitation for Black to get the kingside pawns rolling.} (9. b4 {is probably best here, following up a2-a3 by pressing forward with the queenside attack, while unlike in the line with 7.a3 above, White's king has already castled to comparative safety on the kingside.} Ng6 (9... Bh3 $6 10. b5) (9... h5 $6 10. b5 Na5 11. Qa4 b6 12. Bd2) 10. Bb2 {#[Position after 10.Bb2] and Black can regain the pawn on e5 here, but White has a slight advantage because the queenside attack is currently more potent than Black's kingside attack.}) ( 9. Nbd2 Ng6 {followed by pushing the h-pawn is quite OK for Black.} 10. b4 h5 11. Bb2 (11. b5 $6 Ncxe5) 11... h4) 9... h6 10. Bxe7 {Giving up the bishop-pair like this is not ideal, but otherwise White would have had to retreat the bishop to a less-than-ideal square.} Bxe7 11. Nbd2 h5 12. b4 Bh3 $6 (12... h4 {is probably stronger here, striving to open the h-file immediately. White cannot safely take on h4 because Black's e7-bishop controls that square.} ) 13. Qa4 {A standard idea in this variation. White ignores Black's kingside build-up and goes for a queenside onslaught.} h4 14. b5 Nb8 15. Qxa7 {This is not as disastrous for Black as it looks, since Black's b8-knight turns out to be quite a good defensive piece, but White certainly has the upper hand here.} hxg3 16. fxg3 Bxg2 17. Kxg2 Qh3+ (17... g5 {intending g4 may well be a stronger way to keep attacking chances going.}) 18. Kg1 g5 {Black prepares to dislodge the knight on f3 with ...g5-g4, so White must be careful.} 19. c5 (19. Rf2 {is the most secure way of defending h2, with a large advantage for White.} ) 19... g4 20. c6 bxc6 21. bxc6 Nxc6 22. Qa6+ Kd7 23. e6+ $6 {After this move, Black's attack forces White to make concessions.} (23. Rac1 $16 {is a better way to try and force Black to attend to the problems on the queenside, rather than going ahead with the kingside offensive. White threatens to take on c6 and I think Black's defence from here is difficult.}) 23... fxe6 24. Rac1 Bd6 25. Qxc6+ Ke7 26. Nh4 {This is played to prevent Black from either winning a piece or mating on h2, but Black has a strong sacrifice in response.} (26. Qb7 {is probably White's only decent alternative, which leads to positions with roughly equal chances, e.g.} gxf3 27. Rxc7+ Bxc7 28. Qxc7+ Ke8 29. Nxf3 {and White has two pawns for the exchange and has repelled Black's onslaught down the h-file for the time being, though must still take some care.}) 26... Rxh4 $1 27. Qg2 Qxg2+ 28. Kxg2 Rh5 {Black has the advantage in this endgame due to bishop vs. knight and the powerful central pawns.} 29. Nc4 Rdh8 30. Rh1 e5 31. e4 Rb8 32. Rb1 Rhh8 33. a4 Ke6 34. Rhc1 Rb4 35. Rxb4 Bxb4 36. Rb1 c5 37. a5 Ra8 38. Ra1 d3 39. a6 d2 40. Ne3 Bc3 41. Ra4 $2 {This leaves White with trouble holding onto the a6-pawn when the black king gets to b5.} (41. Ra2 {should give White fair drawing chances.}) 41... Bd4 42. Nd1 Kd6 43. Kf1 Kc6 44. Ke2 Kb5 45. Ra3 Rxa6 46. Nc3+ Bxc3 47. Rxc3 c4 48. Kxd2 Ra2+ 49. Ke3 Rxh2 50. Rc1 Rg2 $2 {This allows White to force a draw by perpetual check or stalemate.} ( 50... Rb2 {prevents White from checking the black king, and Black has very good winning chances.}) 51. Rb1+ Kc5 52. Rb5+ {A strong "stalemate defence" which obliges Black to take a draw since there is no way to escape the barrage of rook checks.} 1/2-1/2 [Event "Amber-blindfold 13th"] [Site "Monte Carlo"] [Date "2004.03.25"] [Round "5"] [White "Gelfand, Boris"] [Black "Morozevich, Alexander"] [Result "0-1"] [ECO "D09"] [WhiteElo "2709"] [BlackElo "2732"] [PlyCount "56"] [EventDate "2004.03.20"] [EventType "rapid"] [EventRounds "11"] [EventCountry "MNC"] [EventCategory "19"] [Source "ChessBase"] [SourceDate "1996.11.15"] 1. d4 d5 2. c4 e5 3. dxe5 d4 4. Nf3 Nc6 5. g3 Nge7 {Black prepares to round up the e5-pawn with ...Ng6, and Black has been doing rather well with this approach, although Vassily Ivanchuk may have found a way for White to get a slight edge against it.} 6. Bg2 Ng6 (6... Be6 {is also playable, but should be slightly better for White. Black changes plans, with the idea of playing ... Qd7 and ...0-0-0 after all, but Black is committed to an early ... Nge7 which narrows down Black's options.} 7. Nbd2 Qd7 8. O-O O-O-O (8... Ng6 9. Ng5 { returns the pawn but aims to chop off the e6-bishop for the white knight, which should be to White's advantage.} Ngxe5 10. Nxe6 Qxe6 11. Nf3 O-O-O 12. Qb3 Na5 (12... Qxc4 $6 13. Nxe5) 13. Qb5 Nxf3+ 14. Bxf3 $14 {Hort,V (2605) -Gasic,B (2360) Sarajevo 1972}) 9. b4 $1 Ng6 {Black needs to play this after all, because of the threat of b4-b5, forcing Black's c6-knight to a less-than-ideal square. After ...Ng6, the knight can now capture the pawn on e5. Taking the pawn on b4 is risky as it opens the b-file up and White can launch a faster attack against Black's king.} 10. Qa4 Kb8 11. b5 Ncxe5 12. Bb2 $14 {Lauber,A (2380)-Manhardt,T (2135) Medellin 1996}) 7. Bg5 {White tries to disrupt Black's plan of rounding up the pawn on e5. This line leads to considerable complications.} (7. O-O {is a quiet line which allows Black to regain the pawn on e5, but White can hope to get a slight positional edge.} Ngxe5 8. Nxe5 Nxe5 {Diagram [#]} 9. b3 {is the most common continuation, where White reinforces the pawn on c4 and prepares Bc1-b2.} (9. Na3 {may, however, be a slight improvement for White, as played by Vassily Ivanchuk:} Be7 10. Nb5 O-O 11. Nxd4 Nxc4 12. b3 $14 {Ivanchuk,V (2769)-Morozevich,A (2770) Astana 2012 }) 9... Bc5 10. Bb2 O-O {has, however, scored well for Black, and in some lines Black can afford to let the d4-pawn drop in return for piece play. Here is one practical example which saw Black work up a kingside attack:} 11. Nd2 a5 12. a3 (12. Ne4 Ba7 13. Qd2 {is probably White's best continuation, with approximate equality- instead, White drifted into a worse position.}) 12... Bg4 13. h3 Bh5 14. Nf3 Nc6 {with a good game for Black, Pedersen,S (2457)-Sadorra, J (2431) Beijing 2008}) 7... Qd7 {This blocks in the c8-bishop but it appears to be the least of the possible evils.} (7... Be7 $6 {either loses a pawn for inadequate compensation, or loses castling rights:} 8. Bxe7 Qxe7 (8... Kxe7 9. Qd2 $16) 9. Nxd4) 8. O-O (8. e6 $5 fxe6 9. O-O e5 {is an interesting alternative, with a complex position in which chances are approximately equal. Black has a lot of control over the centre but the central pawns can potentially be a weakness as well as a strength.} 10. Nbd2 h6 11. Bh4 {should be followed up by Michael Goeller's suggestion} Bb4 12. Qc2 ({Perhaps} 12. Ne4 {may be superior.}) 12... Qf7 13. Ne4 O-O 14. a3 Nxh4 15. Nxh4 Be7 $15 {Gohla, U (2085)-Schmidt,T (2115) DESC email 2006}) 8... h6 $1 {Diagram [#] Black should generate trouble on the kingside rather than being too focused on regaining the pawn on e5.} (8... Ngxe5 9. Nbd2 {is better for White, for instance Black's d4-pawn is in danger of dropping off.}) 9. Bf4 (9. Bc1 {is a more popular alternative, which gets the bishop out of the way of any attacks, but returning to the starting square means that White falls a little behind in development.} Ngxe5 10. Nbd2 Be7 {has scored rather well for Black, for example,} 11. Qa4 Nxf3+ 12. Nxf3 O-O 13. Rd1 Rd8 14. a3 Bf6 {with at least equality for Black. Capuano,E (2173)-Salvador,R (2416) Bergamo 2005}) 9... Nxf4 10. gxf4 g5 {A double-edged continuation, very much in keeping with Morozevich's preferred style, disrupting White's defence of the e5-pawn and gaining kingside space.} (10... Be7 11. e3 {should be a little better for White.}) 11. Nbd2 gxf4 12. Ne4 {#[Position after 12.Ne4]} Be7 {Black leaves the g-file open for a potential kingside attack.} (12... Bg7 {attacking e5 is also playable, and may be even stronger, but there have been no high-level games with it.}) 13. Qd2 (13. Kh1 {is probably a slight improvement over the game, getting the king off the g-file immediately. One practical example saw Black go in for a sharp, probably unsound, attack which paid off:} Qg4 (13... Rg8 14. Rg1 Rg6 {may be best, with approximately equal chances.}) 14. Nxd4 Nxe5 15. Nb5 {with adavantage for White, but Black went on to win, Kozak,M (2257) -Bukal,V (2405) Znojmo 2005}) 13... Qg4 14. Kh1 Bf5 15. Nxd4 $6 (15. h3 Qh5 16. Qxf4 {is a slight improvement but Black is still on top after} Bxh3) 15... Rd8 (15... O-O-O {may be even stronger.}) 16. Nxf5 $6 {This does not give White sufficient material to compensate for losing the queen.} ({The best continuation for White was probably} 16. Nf6+ Bxf6 17. Bxc6+ bxc6 18. exf6 { although I think Black still has a slight advantage following} c5) 16... Rxd2 17. Nxe7 Kxe7 18. Nxd2 Qxe2 19. Nf3 Rg8 20. b3 Nb4 21. Nd4 Qg4 22. Be4 (22. Bf3 {is better, though White is still losing, probably best is the active} Qh3) 22... Rg5 (22... Nd3 $1 {is the quickest winning move- Goeller, citing Raetzky and Chetverik. White has to make heavy concessions in order to avoid being quickly checkmated.}) 23. Rg1 Qd7 24. Nf3 Rxg1+ 25. Rxg1 Nd3 26. Rg2 c6 27. Bh7 a5 28. Nh4 Nxe5 0-1