Black's two main gambit lines in this variation rely upon White playing 1.e4 d5 2.exd5 Nf6 3.c4 or 3.d4, but White also has several other alternatives which are quite popular. 3.Nf3 can also be met by 3...Bg4, which, like 3.d4 Bg4, is probably not fully sound, but is playable. Critical is 4.Bb5+ Nbd7 5.h3! (rather than 5.c4 e6, which probably amounts to an improved version for Black over the line 3.d4 Bg4 4.f3 Bf5 5.Bb5+ Nbd7 6.c4 e6, and certainly gives Black good piece play). Michael Goeller analysed some critical lines in his article, The Anti-Portuguese, but some recent games in the line by David Smerdon have suggested that Black gets a fully playable game. Sounder for Black is 3...Qxd5, with the idea of rapid queenside castling and kingside counterplay. White gets slightly the better of it with best play, but Black's position is playable and should appeal to fans of these gambits.
3.Be2 is a more definite way to wipe out Black's gambits, as now Black has no control over the g4-square. Black should probably continue again with 3...Qxd5, followed by ...Nc6, ...0-0-0 and ...e6.
An important alternative way for White to try and hold onto the extra pawn is 3.Bb5+. Then after 3...Bd7 4.Bc4, Black has to find 4...Bg4 in order to get sufficient compensation for the sacrificed pawn. White also has 4.Be2, which has the same aim as in the line 3.d4 Bg4 4.Bb5+ Nbd7 5.Be2, aiming to head for a risk-free edge in positions reminiscent of the Caro-Kann Defence. Black's most ambitious option is instead 3...Nbd7, committing fully to sacrificing a pawn, with double-edged play resulting after 4.c4 a6 5.Bxd7+ Bxd7 6.Nf3 e6.
Games and analysis
Australian grandmaster David Smerdon has been the leading exponent of these gambit approaches stemming from 1.e4 d5 2.exd5 Nf6 in recent years, so three out of the four illustrative games feature him playing the black side. The coverage is as follows: 1. 3.Be2 2. 3.Nc3 3. 3.Nf3 where Black goes for the double-edged 3...Bg4 4. 3.Bb5+
Games
[Event "Monarch Assurance 16th"]
[Site "Port Erin"]
[Date "2007.09.30"]
[Round "9"]
[White "Ashton, Adam"]
[Black "Smerdon, David"]
[Result "1/2-1/2"]
[ECO "B01"]
[WhiteElo "2295"]
[BlackElo "2461"]
[PlyCount "38"]
[EventDate "2007.09.22"]
[EventType "swiss"]
[EventRounds "9"]
[EventCountry "IOM"]
[Source "ChessBase"]
[SourceDate "2007.11.01"]
1. e4 d5 2. exd5 Nf6 3. Be2 {Diagram [#] White's idea is simply to wipe out
all of Black's possible gambits, since Black now cannot play ...Bg4, and White
has not created any weaknesses by pushing with c2-c4 to defend d5.} Qxd5 {
Black can also consider recapturing with the knight, but the queen recapture
is more aggressive, aiming for rapid queenside castling and kingside
counterplay.} 4. Nf3 {This is usually played, to cover Black's threat against
g2, but this allows Black to play ...Bc8-g4.} Bg4 {This puts some pressure on
White's f3-knight and thus discourages White from moving the light-squared
bishop to a more active square.} (4... Nc6 {is a reasonable alternative,
though it will often transpose, for example,} 5. Nc3 (5. d4 {leaves Black with
nothing better than the straight transposition,} Bg4) 5... Qd6 6. O-O a6 {(to
discourage Nc5-b5)} 7. d4 Bg4 8. Be3 e6 9. h3 Bxf3 10. Bxf3 $14 {Nguyen,A
(2439)-Bitoon,R (2455) Bandar Seri Begawan 2012}) 5. d4 {This leads to a
position that often arises from the Portuguese/Jadoul Gambit: 1.e4 d5 2.exd5
Nf6 3.d4 Bg4 4.Nf3 Qxd5 5.Be2. In that line I mainly examined the critical 5..
.Nc6 6.Nc3, but here White tries a different option.} (5. Nc3 {is the main
alternative.} Qd6 (5... Qf5 {is also playable, although it normally involves
conceding the bishop-pair, e.g.} 6. d4 Nc6 7. h3 Bxf3 8. Bxf3 O-O-O 9. Bxc6
Qe6+ 10. Be3 Qxc6 11. O-O e5 {and Black has succeeded in keeping the chances
level, Daulyte,D (2334)-Jarvinen,R (2119) Tartu 2010}) 6. d4 c6 7. O-O Nbd7 8.
h3 Bh5 9. Bg5 $14 {Movsesian,S (2688)-Tiviakov,S (2655) Wijk aan Zee 2013})
5... Nc6 6. h3 {This encourages an exchange on f3, giving White the
bishop-pair, but this allows Black to strike out in the centre.} Bxf3 (6... Bh5
7. c4 {expands in the centre and is slightly better for White, for example,}
Qd7 8. g4 Bg6 9. d5 Nb4 10. Ne5 Qd6 11. Nxg6 hxg6 {with a significant
advantage for White, Ferguson,M (2404)-Taulbut,S (2400) Hinckley Island 2009})
7. Bxf3 Qd7 {Black threatens to play ...Nc6xd4, so White defends the d4-pawn.}
8. c3 (8. O-O O-O-O (8... Nxd4 9. Bxb7) 9. c3 e5 {is the main alternative,
which gives approximately equal chances, Shabalov,A (2570)-Gershon,A (2540)
Paget Parish 2001}) 8... e5 {Diagram [#]} (8... O-O-O {is worth considering,
getting the king out of the centre and bringing a rook to the d-file to help
support the ...e7-e5 thrust.} 9. Qb3 {is then White's main try for an
advantage, attacking f7, and then} e6 10. O-O Bd6 {leads to a double-edged
middlegame, with a slight advantage for White.}) 9. d5 {White would rather not
exchange the central pawns, since after exchanging queens, recapturing on e5
and playing ...0-0-0, Black can take control of the central files.} Ne7 10. c4
e4 11. Be2 Nf5 (11... O-O-O 12. Nc3 {is slightly better for White, but Black's
chances are better than in the game.}) 12. g4 Bb4+ 13. Nc3 Nh4 14. Bg5 (14. g5
{leaves Black's f6-knight without a good retreat square, and White has a large
advantage following} Ng8 15. Bd2) 14... Ng6 15. Qb3 Qe7 16. O-O-O O-O {Diagram
[#]In a reversal of the usual roles in this variation, it is White who has
castled queenside, and Black has castled kingside. This position favours
White, however, as White's kingside attack is the more potent.} 17. h4 (17.
Bxf6 gxf6 (17... Qxf6 $2 18. Qxb4) 18. Rd4 {looks good for White.}) 17... Bd6
18. Nb5 Bf4+ 19. Kb1 Qe5 {A draw was agreed here, but I think White has good
attacking prospects on the kingside. Perhaps White was happy to take a draw
due to being the lower-rated player.} 1/2-1/2
[Event "WMSG Rapid 1st (Women)"]
[Site "Beijing"]
[Date "2008.10.06"]
[Round "6"]
[White "Kovanova, Baira"]
[Black "Hou, Yifan"]
[Result "0-1"]
[ECO "B02"]
[WhiteElo "2379"]
[BlackElo "2578"]
[PlyCount "70"]
[EventDate "2008.10.05"]
[EventType "swiss (rapid)"]
[EventRounds "9"]
[EventCountry "CHN"]
[Source "ChessBase"]
[SourceDate "2008.11.05"]
1. e4 d5 (1... Nf6 2. Nc3 d5 3. exd5 {is the more common route to the position
at move 3, as was actually used in this game.}) 2. exd5 Nf6 3. Nc3 {Diagram [#]
White adds to the defence of the d5-pawn, preventing Black from recapturing
with the queen.} Nxd5 ({Black can try to persist in seeking compensation for a
pawn with} 3... Bg4 4. Be2 Bxe2 (4... h5 5. h3 Bf5 6. Bb5+) 5. Qxe2 {but I
don't think this line is particularly sound.} Nxd5 6. Qb5+ {doesn't work out
very well for Black.}) 4. Bc4 {This is normally played, developing a piece and
undermining the position of the knight on d5.} (4. Nxd5 $6 Qxd5 {Diagram [#]
is a popular, but poor, alternative, which brings Black's queen out into the
open but there is no good way of attacking it, and if anything I think Black
can get the edge.} 5. Nf3 (5. Qf3 {aims to reduce the potential pressure by
exchanging off Black's active queen, but} Qc5 {Black refuses the possible
queen trade.} (5... c6 6. Qxd5 cxd5 7. d4 Nc6 {is also quite good, giving
Black two central pawns vs. White's one. One practical example continued} 8.
Nf3 Bf5 9. c3 a6 10. Bf4 h6 {Samolins,V (2412)-Luther,T (2515) Pardubice 2011})
6. c3 e5 7. Ne2 Nc6 8. b4 Qd6 9. Ng3 Be6 10. Nf5 Qd7 {with a slight advantage
for Black, Stromboli,C (2326)-Sanchez,J (2533) Milan 2012}) 5... Nc6 6. d4 Bg4
7. Be2 (7. c4 Qf5 8. Be2 O-O-O $15) 7... O-O-O 8. c3 e5 9. O-O exd4 10. Nxd4
Bxe2 11. Qxe2 Nxd4 12. cxd4 Bd6 $15 {Zallio,L (2121)-Matsuura,E (2480) Natal
2012}) 4... Nb6 {This is probably best, hitting out at the c4-bishop.} (4... e6
{is a solid alternative, although there is one way for White to try to win a
pawn against it, which was pointed out by Graham Burgess in his Mammoth Book
of Chess, suggesting that it was quite good for White, but I think that White
should avoid the pawn-grab line.} 5. Nf3 (5. Bxd5 exd5 6. Qe2+ {is the way to
win a pawn, but} Be6 (6... Be7 7. Qe5 O-O 8. Qxd5 Bd6 {also gives Black good
compensation for a pawn}) 7. Qb5+ Nc6 8. Qxb7 Nb4 9. Qb5+ c6 10. Qa4 Bd7 {and
Black stands better in spite of the pawn minus, 0-1 (16) Cabrera Moreno,D
(2191)-Chueca Forcen,A (2350) Aragon 2012}) 5... Be7 6. d4 O-O 7. Ne4 b6 8. Qe2
{Diagram [#] may give White a slight edge due to the superior development,
Fedorov,A (2595)-Minasian,A (2570) Krasnodar 1998}) 5. Bb3 (5. Bb5+ {may tempt
some players, but the bishop is less well-placed here.} c6 {Black blocks out
the bishop by pushing the c-pawn. This takes the c6-square away from Black's
queen's knight, but Black can develop the c8-bishop to f5 and then the
b8-knight to d7.} 6. Be2 Bf5 7. Nf3 e6 8. d4 Bd6 9. Bg5 Qc7 10. Bh4 N8d7 {and
Black has a comfortable equality and can consider castling to either side of
the board, Valent,P (2070)-Woeber,F (2153) Velden 2009}) 5... Nc6 6. Nf3 {
Diagram [#] Black has quite a wide choice of moves here.} Bf5 (6... e5 {is
aggressive, but may be a little committal, as White can castle and then snipe
at the black e5-pawn.} 7. O-O Bd6 8. d4 {is quite good for White.}) (6... g6 {
is a possibility, aiming for a kingside fianchetto.} 7. a4 a5 8. d4 Bg7 9. Be3
O-O $11 {Antal,G (2479)-Perelshteyn,E (2534) Lubbock 2009}) 7. a4 {[%csl Rb6]
[%cal Ga4a5] White prepares to dislodge the knight on b6 by playing a4-a5.} a5
(7... Na5 {is the main alternative, which puts the knight "on the rim", but
does not weaken the black queenside.} 8. Ba2 e6 9. O-O Be7 (9... Bd6 {is also
good, but queenside castling is difficult to arrange as moving the queen away
from d8 leaves it vulnerable to attack at this stage of the game.}) 10. Re1 O-O
11. Ne4 c5 $11 {Epishin,V (2460)-Horvath,C (2470) Leningrad 1989}) 8. O-O e6 9.
h3 Bd6 10. d4 O-O {Diagram [#] Chances are equal in this position.} 11. Nb5 h6
12. c4 Be7 13. Bf4 Rc8 14. d5 Nb4 15. Nfd4 (15. Nbd4 {would have been better.}
Bd3 16. dxe6 Bxf1 17. Qxf1 $44 {is then a stronger exchange sacrifice than in
the game. White's knights are better placed to add to an attack against the
black king.}) 15... Bd3 16. dxe6 (16. Re1 exd5 17. cxd5 Bf6 $15) 16... Bxf1 17.
Kxf1 c5 18. Nf5 fxe6 19. Nxe7+ Qxe7 {Diagram [#] White does not really have
enough compensation for the exchange.} 20. Bd6 Qf6 21. Bxf8 Rxf8 22. Qe2 Nd3
23. Kg1 Nf4 24. Qe4 Qxb2 25. Rb1 Qe2 26. Nc3 Qxe4 27. Nxe4 Nd7 28. Rd1 Ne2+ 29.
Kf1 Nd4 30. Ba2 b6 31. Bb1 Kf7 32. Bd3 Ke7 {Diagram [#]} 33. Nc3 $2 Ne5 34. Nb5
{This allows Black to create a passed pawn by exchanging knights on b5.} Nxb5
35. axb5 ({Or} 35. cxb5 Rd8 {and Black has a passed pawn on c5 which is hard
to stop.}) 35... Rd8 {White's problem is that there is no good defence against
the plan of pushing the black a-pawn down the board.} 0-1
[Event "BCF-chT 1213 (4NCL)"]
[Site "England"]
[Date "2013.05.06"]
[Round "11.152"]
[White "Emms, John M"]
[Black "Smerdon, David"]
[Result "0-1"]
[ECO "B01"]
[WhiteElo "2483"]
[BlackElo "2498"]
[PlyCount "76"]
[EventDate "2012.11.03"]
[EventType "team-tourn"]
[EventRounds "11"]
[EventCountry "ENG"]
[Source "ChessBase"]
[SourceDate "2013.11.20"]
[WhiteTeam "Wood Green Hilsmark II"]
[BlackTeam "Guildford II"]
[WhiteTeamCountry "ENG"]
[BlackTeamCountry "ENG"]
1. e4 d5 2. exd5 Nf6 3. Nf3 {Diagram [#] This is another "spoiling" idea to
try and cut out Black's gambits, but Black can still continue with ...Bg4 here.
} Bg4 {David Smerdon, a long-standing exponent of the 3.d4 Bg4 line, has
tested the closely-related 3.Nf3 Bg4 recently. It probably isn't fully sound,
but then again, the same could equally be said for 3.d4 Bg4, and it similarly
gives Black practical chances.} (3... Qxd5 {is a more definitely sound
alternative, which will normally transpose into lines examined under 3.d4 Bg4
4.Nf3 Qxd5, and 3.Be2 Qxd5.}) 4. Bb5+ (4. d4 Qxd5 {transposes to 3.d4 Bg4 4.
Nf3 Qxd5, a line which is slightly better for White but with counterplay for
Black.}) (4. h3 Bh5 5. c4 {is another idea for White, but Black can try the
gambit continuation} e6 6. dxe6 (6. g4 Bg6 7. dxe6 Nc6 8. exf7+ Bxf7 (8... Kxf7
{is probably the better recapture, e.g.} 9. Be2 Bc5 10. Nc3 Re8 $44) 9. d4 Bb4+
10. Nc3 O-O 11. Be3 $14 {Zeller,F (2417)-Dietrich,F (2241) Unterkochen 2003})
6... Nc6 7. Be2 (7. exf7+ $6 Kxf7 {is a bit too greedy.}) 7... fxe6 8. O-O $14
{and Black has partial compensation for the pawn.}) 4... Nbd7 (4... c6 {does
not provide Black with enough for the pawn, though it may be playable enough
for rapid games, e.g.} 5. dxc6 Nxc6 6. h3 Bh5 7. Nc3 Qc7 8. g4 Bg6 9. Qe2 $14 {
Papa,S (2349)-Filipovic,B (2432) Switzerland 2003}) 5. h3 {Jeremy Silman
established this as White's most critical continuation in one of his chess
articles.} (5. c4 e6 6. dxe6 $6 (6. h3 Bh5 {transposes to the game}) 6... fxe6
7. d4 c6 8. Ba4 e5 {gives Black reasonable compensation for the pawn.}) 5...
Bh5 {Black would rather not trade bishop for knight in this position, as a key
part of Black's compensation for the pawn is the strength of Black's
light-squared bishop.} (5... a6 {is an alternative, but it should be better
for White.} 6. Be2 {Recommended by John Emms, according to Michael Goeller,
and this should give White a significant advantage.} (6. hxg4 axb5 7. Qe2 Nxd5
8. Qxb5 e6 {is somewhat less clear.}) 6... Bxf3 7. Bxf3 Ne5 8. Nc3 {and Black
doesn't have much to show for the pawn.}) 6. c4 {White plays to hold onto the
extra pawn.} (6. Nc3 {prepares to return the pawn, in order to emerge with a
normal edge from the opening.} a6 7. Be2 Nb6 8. d4 Nfxd5 9. Nxd5 {This has
always been played in the practical examples that I can find, but White can
also consider playing Nc3-e4 or castling here.} Qxd5 10. O-O e6 (10... O-O-O {
is combative, and appears playable enough here.}) 11. a4 a5 12. b3 Bb4 13. c4
Qd8 14. Bg5 $14 {Adair,J (2399)-Smerdon,D (2523) England 2014}) 6... e6 (6...
a6 {doesn't give Black much for the pawn after} 7. Ba4 $1 (7. Bxd7+ $6 Qxd7 8.
O-O O-O-O {leaves Black only slightly short of full compensation for the pawn,
and Black can be happy with the attacking positions that this sort of
double-edged position offers.}) 7... b5 8. cxb5 Nxd5 9. Nc3 axb5 10. Bxb5 $16)
7. dxe6 (7. Nc3 {is also playable, returning the pawn in order to get castled,
which should give White an edge:} exd5 8. O-O (8. cxd5 Qe7+ 9. Qe2 {is not as
convincing for White, Landa,K (2613)-Vafiadis,K (2213) Kallithea 2008}) 8...
Be7 9. Re1 O-O 10. cxd5 Nb6 11. d4 $14) 7... fxe6 {Diagram [#] The outcome can
be considered fairly successful for Black, who has a fair amount of
compensation for the pawn due to the half-open d and f-files, and the fact
that White is still stuck in an awkward pin on the f3-knight.} 8. d4 c6 9. Ba4
Nb6 (9... Qc7 {followed by queenside castling is a good alternative.}) (9... e5
$6 {is ineffective here, however:} 10. g4 Bg6 11. dxe5) 10. Bb3 Bb4+ 11. Nc3
O-O 12. Be3 a5 (12... Ne4 13. Qc2 Bg6 (13... Bxf3 14. gxf3 {leaves White with
the half-open g-file pointing at Black's king.}) 14. Qc1 $14 {keeps everything
solid for White, though this may have been better for Black than the game
continuation.}) 13. a3 Bxc3+ 14. bxc3 a4 15. Ba2 Qe7 16. g4 Bg6 17. Ne5 Be4 18.
f3 {Diagram [#] If Black was forced to retreat the bishop here, then White
would stand clearly better, but Smerdon finds a strong piece sacrifice.} Nfd7
$1 {Black prepares to open up the white king by playing ...Qh4+.} (18... Bg6
19. Qd2 $16 {threatens to continue pushing the kingside pawns.}) 19. fxe4 (19.
O-O {may have been best, declining the piece sacrifice, and White may keep an
edge, e.g.} Nxe5 20. dxe5 Qh4 21. Qe2) 19... Nxe5 20. dxe5 Qh4+ 21. Ke2 Qg3 22.
Rf1 Qg2+ 23. Rf2 (23. Kd3 Rad8+ 24. Bd4 Nd7 $44 {threatens ...Nc5+, keeping
the attack going.} (24... Qxh3+ $2 25. Kc2 Qg2+ 26. Qd2 Qxe4+ 27. Kb2 {is no
good, though, as Black's checks run out.})) 23... Rxf2+ 24. Bxf2 Rf8 25. Qd4
Qf3+ 26. Ke1 $2 {This move loses.} (26. Kd2 {is likely to lead to a draw, e.g.}
Qg2 27. Ke2 Rf4 28. Qxb6 Qf3+ 29. Kd2 Qg2 30. Ke2 Qf3+) 26... Qh1+ 27. Bg1 c5
28. Qe3 Rf3 29. Qxc5 Nd7 30. Qd4 Nxe5 31. c5 Rf8 32. Bxe6+ Kh8 33. Qe3 Qg2 34.
Bf5 Nf3+ 35. Kd1 Rd8+ 36. Kc1 Rd2 37. Qxf3 Rc2+ 38. Kb1 Rb2+ 0-1
[Event "Sydney op"]
[Site "Parramatta"]
[Date "2010.04.09"]
[Round "5"]
[White "Li, Chao B"]
[Black "Smerdon, David"]
[Result "1-0"]
[ECO "B01"]
[WhiteElo "2613"]
[BlackElo "2530"]
[PlyCount "49"]
[EventDate "2010.04.07"]
[EventType "swiss"]
[EventRounds "9"]
[EventCountry "AUS"]
[Source "ChessBase"]
[SourceDate "2010.05.04"]
1. e4 d5 2. exd5 Nf6 3. Bb5+ {Diagram [#] This is another important way of
side-stepping the lines following 3.c4 e6 and 3.d4 Bg4, but it does not stop
Black from playing a gambit.} Nbd7 {This is the most ambitious response,
although perhaps objectively inferior to 3...Bd7. Black prepares to hit out
at White's b5-bishop with ...a7-a6 and sacrifice a pawn with ...e7-e6, in the
spirit of the 3.d4 Bg4 lines.} (3... Bd7 {is more than five times as popular,
and gives Black reasonable chances, but White does have a "killjoy" response
to it, dropping the bishop back to e2, which aims for a slight and largely
risk-free edge.} 4. Be2 $5 {White wants to avoid being pinned with ...Bd7-g4,
and having blocked the black queen's path to d5, encourages Black to head into
the more restrained type of positions that follow the knight recapture on d5.}
(4. Bc4 Bg4 5. f3 Bf5 {is the main line following 4.Bb5+, in which it is hard
for White to hold onto the extra pawn on d5.} 6. Nc3 {Diagram [#]} Nbd7 (6...
c6 $5 7. dxc6 Nxc6 {has been played by various players rated over 2000 Elo,
and gives Black better prospects than many of the other ...c7-c6 lines, though
White should still have an edge with best play.} 8. Nge2 e5 9. O-O Bc5+ 10. Kh1
$14 {Wood,C (2169)-Fischer,M (2129) Bayern 2011}) 7. Nge2 Nb6 8. Bb3 Nfxd5 9.
O-O e6 10. Nxd5 Nxd5 11. Ng3 Bg6 {with approximately equal chances although
White can hope to get a small edge by pushing the d and c-pawns. The game
Heredia Serrano,C (2206)-Feliciano Ebert,V (2267) Montevideo 2012, continued}
12. d4 Be7 13. c4 Nb6 14. Be3 O-O 15. Ne4 {and White already had a slight
advantage.}) 4... Nxd5 ({Maybe Black can get away with flicking in} 4... Bf5 {
, e.g.} 5. d4 (5. c4 e6 6. dxe6 fxe6 7. Nf3 Nc6 $14 {leaves Black with a fair
amount of compensation for the pawn.}) 5... Qxd5 6. Nf3 Nc6 7. O-O O-O-O 8. Be3
{with equal chances.}) 5. d4 Bf5 (5... c5 {is an interesting alternative, but
should still give White a small edge with minimal risk e.g.} 6. c4 Nf6 7. d5 e6
8. dxe6 fxe6 9. Nf3 $14 {Bauer,C (2627)-Filipovic,B (2374) Zuerich 2012}) 6.
Nf3 {Diagram [#]} e6 ({It may also be possible, though risky, for Black to
consider queenside castling, e.g.} 6... Nc6 7. O-O Qd6 8. c4 Nf4 9. Bxf4 Qxf4
10. d5 Ne5 11. Qa4+ Bd7 12. Qb3 O-O-O 13. c5 $14 {Hagarova,Z (2380)-Jurcisin,I
(2129) Tatranske Zruby 2003}) 7. Bd3 Bxd3 8. Qxd3 Be7 9. O-O O-O 10. c4 Nf6 11.
Rd1 {with a slight advantage for White, Delgado Ramirez,N (2583)-Matsuura,E
(2491) Sao Paulo 2013}) 4. c4 {White holds onto the extra pawn.} (4. Nc3 {is
also playable, but has scored well for Black. With this response White is not
bothered about holding onto an extra pawn.} a6 5. Be2 Nb6 6. Nf3 Nbxd5 7. Nxd5
Nxd5 8. d4 {Diagram [#] and White has the usual slight edge due to the pawn
centre, since White will soon play c2-c4.} g6 ({One practical example of a
more aggressive approach from Black continued} 8... Bg4 9. O-O e6 10. c4 Nf6
11. Qa4+ Nd7 12. h3 $14 {Xia,J (2204)-Barletta,M (2207) Arvier 2010}) 9. O-O
Bg7 10. c4 Nf6 11. Bf4 O-O 12. h3 $14 {Welling,G (2415)-Hebden,M (2570) Isle
of Man 1995}) 4... a6 ({Another option worth investigating is} 4... e6 {which
leads to similar positions to other lines that have been examined previously,
but in this situation I think it is probably inferior to 4...a6.} 5. dxe6 fxe6
6. Nf3 c6 7. Ba4 Nc5 (7... e5 {is untested, but may be a better way to try and
get good play, e.g.} 8. Qe2 Bd6 9. d4 Qa5+ 10. Nc3 O-O $14) 8. Bc2 Nd3+ 9. Bxd3
Qxd3 10. Qe2 $14 {Sax,G (2563)-Levacic,P (2382) Bibinje 2001}) 5. Bxd7+ {White
concedes the bishop-pair, but avoids the risk of Black attacking the white
bishop and gaining queenside space.} (5. Ba4 b5 {leads to complications.} 6.
cxb5 Nxd5 7. Nc3 N5b6 (7... Bb7 {is worth considering, simply developing and
putting the bishop on the long diagonal and reinforcing the d5-knight.} 8. Nf3
e6 9. bxa6 Rxa6 10. O-O $14 {Ivanov,A (2380)-Stefansson,H (2470) Vienna 1991})
8. Bb3 axb5 9. d4 (9. Nxb5 Ba6 10. Qf3 e6 11. Qc6 {, as played in Baramidze,D
(2265)-Szonyi,K (2118) Dortmund 2001, gives Black reasonable compensation for
a pawn after} Bd6 {and if} 12. Nxd6+ cxd6 13. Qxd6 Qg5) 9... Bb7 10. Nf3 $14 {
Erneker,J (2291)-Vojta,T (2365) Pardubice 2009}) 5... Qxd7 6. d4 (6. d3 {is
more popular, adding protection to the pawn on c4.} e6 7. dxe6 Qxe6+ (7... Qc6
8. exf7+ Kxf7 9. Nf3 Bb4+ 10. Nc3 Re8+ {and white Black has a lot of activity,
I am not convinced that it is worth two pawns.}) 8. Qe2 {forces off the queens,
though Black probably has enough play for the pawn after} Qxe2+ 9. Kxe2 Bf5 10.
Nf3 O-O-O {Sermek,D (2526)-West,G (2356) Queenstown 2006}) 6... c6 (6... e6 {
also looks fully playable.} 7. dxe6 (7. Nf3 exd5 8. O-O {is a good alternative
for White, simply giving back the pawn, but I doubt if White can claim much,
if any, advantage here.}) 7... Qc6 8. exf7+ Kxf7 9. Nf3 Bb4+ 10. Nc3 Re8+ {
gives Black dangerous compensation, and Black can regain one of the sacrificed
pawns by taking on c4, Jarvela,J (2242)-Seppala,M (2137) Finland 2008}) 7. Nf3
({A decent option for White is to take the pawn with the aim of subsequently
returning the pawn by letting Black take on c4:} 7. dxc6 Qxc6 8. Nf3 Qxc4 9.
Ne5) 7... cxd5 8. Ne5 Qf5 9. O-O dxc4 10. Nc3 {Diagram [#] Black can be quite
satisfied with the outcome of the opening, with the queen on an active square
and White saddled with an isolated pawn on d4.} b5 11. b3 h5 {This sort of
attacking idea appeals to me, but I think Black had to resolve the situation
on the queenside first.} (11... b4 12. Na4 c3 13. a3 h5 {would have improved
significantly over the game continuation.}) 12. bxc4 e6 (12... h4 13. h3) 13.
cxb5 Ng4 14. b6 Nxe5 15. dxe5 Qxe5 16. b7 Bxb7 17. Qa4+ {The point is that
White's attack comes through first.} Kd8 18. Rd1+ Bd6 19. Bf4 Qxc3 20. Rac1
Qxc1 21. Rxc1 Bxf4 22. Qxf4 Rc8 23. Qd6+ Ke8 24. Rxc8+ Bxc8 25. Qc7 1-0