Chess Opening Theory/1. d4/1...Nf6/2. c4/2...e5/3. dxe5/3...Ng4/4. Bf4/4...Nc6/5. Nf3/5...Bb4/6. Nbd2

With 6.Nbd2 White gives back the gambited pawn in order to keep a healthy pawn structure and to get the bishop pair. After 6...Qe7 White has two important possibilities:
 * 7.a3 to force the immediate exchange of the Bb4 for the Nd2 and get the bishop pair, a space advantage and a minority attack on the queenside.
 * 7.e3 to win a tempo over the 7.a3 variation but it may end up with the Bb4/Nd2 exchange made in less favorable circumstances, or not made at all.

The maverick gambit 6...f6 also exists.

The line 6.Nbd2 Qe7 7.a3
1.d4 Nf6 2.c4 e5 3.dxe5 Ng4 4.Bf4 Nc6 5.Nf3 Bb4+ 6.Nbd2 Qe7 7.a3

The Bb4 is attacked but Black does not have to move it for the moment, and instead plays 7...Ngxe5 to get the gambitted pawn back. Now Black threatens both to take the c4 pawn and to take the Nf3, when White will either have to accept doubled pawns or move his king. For example this is seen after 8.e3?! when 8...Nxf3+ forces either 9.gxf3 or 9.Qxf3 Bxd2+ 10.Kxd2, when White cannot castle anymore.

White cannot play 8.axb4?? because of the Kieninger trap 8...Nd3 mate. White does not want to play 8.Bxe5?! either because it would lose the bishop pair, which is the main source of White's hopes for an advantage in the Bernstein line. So White is more or less forced to exchange a pair of knights with 8.Nxe5 Nxe5.

White still cannot win a piece with 9.axb4?? because the mate threat by Nd3# is still in force. White also cannot win a piece by 9.Bxe5?! because Black would play the zwischenzug 9...Bxd2+ 10.Qxd2 Qxe5 with an equal game. White accordingly plays 9.e3 in order to protect the c4-pawn that was attacked by the Ne5.

Now there is no more mating threat on d3 so the Bb4 is really attacked and Black has to move it. 9...Bd6 (or 9...Bc5 10.b4 Bd6, intending to meet 11.c5?! with 11...Nd3+ 12.Bxd3 Bxf4) would misplace the bishop, and 9...Ba5?? would lose the bishop to 10.b4 Bb6 11.c5. That leaves 9...Bxd2+, when after 10.Qxd2 we get the real starting position of this variation.

It is important to note that for Black, the sequence 7...Ngxe5 8...Nxe5 9...Bxd2+ is not only cunning, but also the best move-order as another sequence would give White an early opportunity to realise the advantageous c4–c5 push. For example after 7...Bxd2+?! 8.Qxd2 Ngxe5 9.Nxe5 Nxe5 White should not play the usual 10.e3?! but should strive for more with the immediate 10.c5! as Black cannot take in c5 without losing the c7-pawn (because of the possibilities Ra1–c1 and Qd2–c3).

After 10.Qxd2, Tseitlin explains that "opening manuals assess this position as favourable to White on the basis of the bishop pair. However, considering the closed nature of the position, White faces substantial difficulties in the realisation of this nominal advantage." Black has not a lot of things to be proud of as there are no targets in White's camp, but can put up a lot of resistance thanks to some small assets:
 * Black's Ne5 is strongly centralized, attacks the c4-pawn, and restricts the Bf1 from moving to the natural squares d3 and f3. Moreover, exchanging the knight with Bxe5 is not appealing for White, since that would mean losing the advantage of the bishop pair.
 * the Bc8 can sometimes become better than its counterpart the Bf1, if it makes it to the good squares b7 or c6 while the Bf1 remains restricted by the Ne5.

This explains the most natural plans for both sides. White will try a minority attack on the queenside, in order to increase its space advantage and to create some weaknesses in the black pawns (e.g. an isolated pawn or a backward pawn). So White will try to use the advances b2–b4 or c4–c5 in good conditions, supported by the queen and the rooks on the c-file and the d-file. On the other hand, Black will try to keep the position closed, most importantly keep the c4-pawn where it is in order to keep the Bf1 at bay. This can be achieved by moves like b7–b6 and d7–d6, and sometimes the manoeuvre Ne5–d7–f8–e6.

The first move by Black has to be 10...d6! because otherwise White plays 11.c5! and gets a clear advantage immediately. For example 10...b6? loses a pawn to 11.Qd5 Nc6 (forced) 12.Bxc7, and 10...O-O?! is bad because of 11.c5! when Black should not take with 11...Qxc5? because of 12.Rc1 Qe7 13.Rxc7 and White is winning already.

The line 6.Nbd2 Qe7 7.e3
1.d4 Nf6 2.c4 e5 3.dxe5 Ng4 4.Bf4 Nc6 5.Nf3 Bb4+ 6.Nbd2 Qe7 7.e3

In this variation White tries to avoid the move a2–a3 in order to gain a tempo over the 7.a3 variation. After the standard moves 7...Ngxe5 8.Nxe5 Nxe5 9.Be2 Black has the main line 9...0-0 or the less-explored 9...b6. While 9...0-0 will give him the option on the tenth move to keep his Bb4 bishop or part with it, 9...b6 is more forcing as Black will be more or less compelled to exchange his bishop for the Nd2.

Lalic thinks the strategies in which Black gives up the bishop pair (by exchanging its Bb4 for the Nd2) for nothing are a mistake. He does not like the strategy to retreat the Bb4 in d6 either, because they are too drawish. He recommends the strategy to retreat the bishop in c5, and keep it there thanks to the push a7–a5.

The sideline 6.Nbd2 f6
1.d4 Nf6 2.c4 e5 3.dxe5 Ng4 4.Bf4 Nc6 5.Nf3 Bb4+ 6.Nbd2 f6

With 6...f6 7.exf6 Qxf6 Black tries to take advantage from the fact White has moved his dark-squared bishop away from the queenside, meaning the b2-pawn is left without protection. Lalic reckons that "although unsound at the highest level, this line has every chance of success as a surprise weapon at club level because White must play actively and accurately to secure the advantage."

One game saw the continuation 8.e3 Qxb2 9.Be2 d6 10.O-O O-O 11.Ne4 Nf6 12.Bd3 Bf5 13.Nxf6+ Qxf6 14.Rb1 Bc5 15.Ng5 Nb4! with Black trying to invade the d3-square. According to Lalic "fortunes could have turned either way." The correct plan for White was shown by Gleizerov who played 8.e3 Qxb2 9.Be2 d6 10.O-O O-O 11.Nb3 Qf6 12.c5! to open the a2–g8 diagonal that was weakened precisely by the gambit move 6...f6. The move 11.Nb3 is not only useful to support the c4–c5 push, but also to exchange the knight against Black's dark-squared bishop after a possible a2–a3 forcing the retreat Bb4–c5. As Lalic puts it, "I doubt if Black has a satisfactory answer to White's play in this game".

The sideline 5.Nbd2 d6
1.d4 Nf6 2.c4 e5 3.dxe5 Ng4 4.Bf4 Bb4+ 5.Nbd2 d6

To reach this line Black has to avoid 4...Nc6 5.Nf3 Bb4+ and play immediately 4...Bb4+. Although this gambit has common themes with the other gambit 6...f6, it is basically more sound as Black does not weaken his kingside (not opening the a2–g8 diagonal) and opens the way for his Bc8. After 6.exd6 Qf6 White can react to the attack on his Bf4 in several ways:
 * After 7.e3 the best for Black (according to Lalic) is 7...g5!, with the possible continuation 8.Bg3 h5 9.h4 Qxb2 (threatening 10...Bxd2+ winning a piece) 10.Ngf3 Bf5 threatening to win the exchange with 11...Bc2 12.Qc1 Ba3.
 * Also possible is 7.Bg3, with the hope of playing the advance e2–e4 in one go in a distant future. In one game play continued 7...Qxb2 8.Ngf3 Bf5 9.a3 Bxd6 10.Bxd6 cxd6 and Black was fine.
 * The best move is considered to be 7.Nh3 as it develops a piece and protects both the Bb4 and the f2-pawn. It also helps that the Bb4 is still guarding the Nd2, so that after 7...Qxb2? there is not the threat of winning the exchange (8...Bxd2+ would be answered by 9.Bxd2) and White can repel Black's attack with 8.Rb1 Qa3 9.Rb3 Qa5 10.dxc7 Nc6 11.a3! Be7 12.e3. Instead, Black must play energetically with 7...Nxf2 8.Kxf2 Bxh3 9.g3 Bxf1 10.dxc7!? Nc6 11.Rxf1 and here Lalic recommends 11...O-O 12.Kg2 Rfe8.

Theory table
1.d4 Nf6 2.c4 e5 3.dxe5 Ng4 4.Bf4 Nc6 5.Nf3 Bb4+ 6.Nbd2 Qe7