Castling for White constitutes an indirect defense of the pawn at e4. According to theory, Black is offered a choice between two standard plans, generally called the Max Lange Attack (5....Bc5 6.e5 d5) and the Anti-Max Lange (5....Nxe4 6.Re1 d5), both of which have been extensively analyzed over many years and now appear to give chances for both sides with careful play. In my own view, the more direct methods of defending the pawn at e4 with either 5.Ng5 (The Perreux Variation) or 5.e5 (The Modern Variation) give White more power to dictate the course of play and are therefore superior. But for players willing to learn a large amount of theory, the Max Lange Attack and its lesser sibling can lead to very interesting tactical exchanges. And with White's King safely castled, White can devote all of his attention to the attack.
While taking the pawn appears dangerous due to the fact it opens up the e-file for an attack on the Black King, it is actually the safer choice, mainly because it generally leads to piece exchanges and simplification of the position. In the main line of the Max Lange Attack, there are fewer piece exchanges, leading to quite complex positions that remain open to discussion despite over 150 years of close analysis. Because of its complexity, in fact, the Max Lange Attack is rarely seen in over the board play and is most popular among correspondence players, who have the time to analyze the many potential lines to discover new ideas. The main line generally runs 5.....Bc5 6.e5 d5 (6....Ng4!? is also playable) 7.exf6 dxc4 8.Re1+ Be6 9.Ng5 Qd5 10.Nc3 Qf5 11.Nce4 O-O-O 12.g4 Qe5 13.Nxe6 fxe6 14.fxg7 (14.Bg5!? is also seen) 14....Rhg8 15.Bh6 d3! 16.c3 d2!? 17.Re2 Rd3 with many continuing complications. You have to wonder about a main line that runs for over a dozen moves!
The temporary Bishop sacrifice is the standard line here. In 1924, the Canal Variation with 7.Nc3!? was equally popular, and White continues to have some success with it. But after 7....dxc4! 8.Rxe4+ Be6 9.Nxd4 Nxd4 10.Rxd4 Qf6! (stronger than the standard 10....Qc8) Black gets at least equality.
This double pin on Knight and pawn is an attractive motif and may be the main reason players sometimes fall in love with this line as White.
The Queen has several choices, but only 8....Qh5!? and 8....Qa5 are considered best by theory. Even in 1924, 8....Qa5 was considered standard. After 8....Qh5!? play can run 9.Nxe4 Be6 10.Bg5 (not 10.Neg5 O-O-O! 11.Nxe6 fxe6 12.Rxe6 Bd6! and Black has the initiative, but 10.Nxd4 Qxd1 11.Rxd1 O-O-O 12.Be3 Nxd4 13.Bxd4 Bf5 is about equal) 10....Bd6 with approximate equality.
The more direct 9.Rxe4+?! is good for Black after 9....Be6 10.Nxd4 O-O-O 11.Be3 Nxd4 12.Rxd4 Bb4.
An interesting move that can lead to some interesting tactical complications. More common is the direct approach to regaining the lost material with 10.Neg5 O-O-O 11.Nxe6 fxe6 12.Rxe6 Bd6 with about equal chances.
More common today is 10....Bb4 (10....Qh5 is also good) 11.Nxd4 Nxd4 12.c3 Be7 13.cxd4 Qd5 14.Bb4 Bxb4 15.Qa4+ Qc6 16.Qxb4 O-O-O with about equal chances. There is nothing inherently wrong with Black's chosen move. His follow-up, though, seems a mistake.
Black should not so readily surrender the ability to castle in this position. Best is 11....Bd6 with the idea of castling kingside after 12.Bf6 O-O! 13.Nxd4 Nxd4 14.Qxd4 Qxd4 15.Bxd4 Rfd8 =.
White puts pressure along the e-file and tries to recover his pawn by an exchange at e6. A better plan was demonstrated in the game Zhuravlev-Dzhanoev, USSR 1976 13. Qd3 Rad8 14. Qa3+! Kd7 15. Nc5+ Kc8 (White aids Black's monarch in making his way to the Queenside so that he can initiate an attack on that wing) 16. b4! a6 17. b5 Na7 18. bxa6 b6 19. Nxe6 fxe6 20. Re5 Qc4 21. Qd3 Qxd3 22. cxd3 Nc6 23. Rxe6 Nb4 24. Re7 Nxa6 25. Rxg7 Nb4 26. Ne5 Rhe8 27. f4 Rxe5 28. fxe5 Nxd3 29. e6 c5 30. Rf1 Nb4 31. Rff7 Nd5 32. Ra7 Kb8 33. Rgb7+ Kc8 34. Rxh7 Kb8 35. Rad7 Rxd7 36. Rxd7 1-0
Another standard attack plan, though not one for the faint of heart, is 13. c4! dxc3 14. Qc2. A 1985 game between Jansen and Mostertman continued 14....Rhd8 15. Neg5 Qd3 16. Qb3 Rd6 17. Nxf7 Kf6 18. Rxe6+ Rxe6 19. N3g5 Re7 20. Rd1 Rae8 21. h4 Qf5 22. Qxc3+ Ne5 23. Rd5 Kg6 24. h5+ Kxh5 25. Qd4 Qb1+ 26. Kh2 Ng4+ 27. Kg3 Nf6 28. Ne6+ 1-0
That White has several strong continuations
here suggests that Black's strategy is not the best.
Black prepares to castle by hand, bringing his King to the queenside. The Rook also helps to support the pawn at d4. If instead 13....Rae8 14.Nxe6 fxe6 15.Qg4 Kd8 and now White can risk 16.Qxg7 due to the threat of Nf6, forking Queen and Rook.
Black must lose his e-pawn eventually. If 15....Kf7 16.Ng5+ etc. Now Black hopes for 16.Qxg7+?! Kc8, when Black's King is suddenly much safer than White's with the open g-file along which to create an attack. But White has a clever way of picking up the e-pawn instead.
Obviously not 16....Qxc5?? 17.Qxe6 mate.
Black creates a useful pin on the Knight, which White does best to break immediately with 18.Qxg7 Qxg7 19.Nxg7, though Black has excellent play for the pawn after 19....Nb4! 20.Re2 Rd5!
White continues with his misguided idea of leaving his Knight at e6 where it is in some danger. Now was the time to run for the draw with 19.Nc5 Qxg4 20.Rxg4 Re2 21.Rc1 Rhe8 22.Kf1. After 19.Rae1, though, the Knight is securely pinned since moving it allows Qxg4 and if White recaptures there follows Rxe1#.
A clear case of chess hallucination. Necessary was a careful retreat with 21.Nc5 or 21.Ng5, with approximate equality.
White clearly overlooked this simple defense. Now he is lost.
Forcing the exchange of Queens due to the threat of mate on the back rank with 25....Re1+. The ending, however, is hopeless, and White should have resigned immediately. Likely he was still kicking himself for his blunder.
Clearing the way for the triumphant march of the d-pawn, which is often Black's strongpoint in these lines of the Two Knights Defense. White must have been continuing the game out of inertia.
An interesting struggle, marred, however, by White's glaring mistake at move 21.