What we are going to look at here is a game where White tries to get away with not playing the theoretical move a3 and what Black should do in response to it. We will be looking at my Round 2 game played at the Kansas Open two weeks ago.
Kansas Open, Round 2
W: IM Michael Brooks (2375)
B: Patrick McCartney (2050)
1.e4 e6 2.d4 d5 3.Nd2 Nf6 4.g3 c5 5.Bg2 Nc6 6.Ngf3 Be7 7.O-O O-O 8.Re1 b5 9.e5 Nd7 10.h4 a5 11.Nf1 Ba6
Whether Black plays 11...Ba6 and White plays one of his normal moves first, or if Black immediately goes for the 11...b4 and 12...a4 idea is typically just a transposition, and it's no different here.
That said, there is a model game by Fischer (Fischer - Myagmarsuren, Sousse Interzonal 1967) in the other line where Black's inferior play lead to him putting his f1-Knight on e3 instead of h2, but the more relevant thing about it is White's Kingside Attack. After 11...b4 12.Bf4 a4 13.a3 (the move we are going to look at omitting) bxa3 14.bxa3 Na5? (Better is to play the Rook to b8, possibly preceded by developing the Bishop to a6, and so something like 14...Ba6 15.N1h2 [or 15.Ne3] Rb8 16.Ng4 Nd4 17.h5 Nxf3+ 18.Qxf3 Qb6 with dynamic equality) 15.Ne3 Ba6 16.Bh3 d4 17.Nf1 Nb6 18.Ng5 Nd5 19.Bd2 Bxg5 20.Bxg5 Qd7 21.Qh5 Rfc8 22.Nd2 Nc3 23.Bf6 Qe8 24.Ne4 g6 25.Qg5 Nxe4 26.Rxe4 c4 27.h3 cxd3 28.Rh4 Ra7 29.Bg2 dxc2 30.Qh6 Qf8 31.Qxh7+ and Black Resigned as 31...Kxh7 32.hxg6+ is mate on the following move with either 33.Rh8 or 33.Be4, depending on Black's response.
12.N1h2 b4 13.Bf4 a4
This is the main theoretical position in this line. Here, it is thought that White should play 14.a3 in order to avoid 14...a3 by Black, which opens up the c3-square for a Knight to attack a2 and White is tied down to hold the a2-pawn and keep Black from creating a very dangerous passed pawn. Here, White ignores the threat.
So White ignores the issue, but this also leads to a dilemma for Black. If you go strictly based on this supposed "theory", then 14...a3 screams to be played. At the same time, as also mentioned in the King's Indian Attack article from September, the early Ng4 also calls for 14...Nd4. Which should Black do?
I played the automatic move. This move does not by any means lose, and against most other "non-a3" moves by White, it is probably best. That said, given White's response in the game, while this move may be "OK", it is a little better to execute the other idea. After 14...Nd4!, White has a major decision to make. If White trades on d4 or allows Black to trade on f3, White has one less piece to attack the Black King with. One of the downsides for Black in the KIA vs French is that he has four pieces, the Rook on a8, Bishop on a6, Knight on d7, and Knight on c6, that make no contribution to the defense of the Black King. Therefore, this is a major achievement for Black in that he has removed a potential attacker of the Black King in return for the removal of one of the four Black pieces that was making no contribution to the defense of his master. If White avoids the trade, Black can also place his d4-Knight on f5, adding another contributor to the defense of his own King. An example of a line that could result from this stronger move could be 15.c4 (viewed as best by Shredder) 15...bxc3 16.bxc3 Nxf3+ followed by 17...Rb8, regardless of how White recaptures, and only after that should Black consider pushing the a-pawn to a3. The position is probably still equal, though Black might even be able to claim a very slight edge rather than White in this scenario.
15.bxa3 bxa3 16.c3
So this was White's idea when he ignored the a3-push. This also shows why 14...Nd4 is stronger than 14...a3. We have a very similar position to the 14...Nd4 line with one major difference. The addition of the Knights, and as mentioned, when you compare the f3-Knight to the c6-Knight, you have a useful attacker going up against a useless defender. Hence why a trade of these pieces favors Black.
16...Bb5 17.Rb1 Ba4 18.Qd2 Rb8?
This move fails for tactical reasons. I saw White's idea when I played this, but underestimated its strength. 18...Qa5 was better and any advantage White has is minimal.
The lesser evil was 19...Ndxb8, which gives White two free moves for his attack on the Kingside, and after 20.h5 Nd7 21.h6, White has the advantage, but it's not as bad for Black as the game move. Again, I saw White's next move back when I played 18...Rb8, but I highly underestimated its effect.
Black has no choice. 20...Nxf6?? 21.exf6 drops the Bishop as 21...Bd6?? 22.Bxd6 Qxd6 23.Qg5 leads to a position where mate can't be avoided, and 20...gxf6? 21.exf6 Qb2 (21...Bd6?? leads to the same mate as in the 20...Nxf6 line) 22.fxe7 Nxe7 23.Qxb2 axb2 24.Rb1 Ng6 25.Bd6 Rc8 26.h5 and Black's position is a train wreck.
21.exf6 Qb2 22.fxg7 Rc8 23.Bd6
Here, the idea of trading Queens and playing Rb1, like in the 20...exf6 line, is inferior. After 23.Qxb2?! axb2 24.Rb1 e5!, the tables turn and Black has the advantage. Best for White is 23.h5! Qxd2 24.Bxd2! Kxg7 25.h6+ and White has the advantage. The move played in the game looks strong as it keeps the Bishop active and appears to tie Black down, but looks can be deceiving. The next few moves lead to a fairly forcing sequence, and believe it or not, this turns out to be Black best line of defense.
23...Qxd2 24.Nxd2 Bc2 25.Bf1 d4 26.Rc1 dxc3 27.Nf3 Nb4
Here is where Black starts falling apart. Better was 27...Bb3! 28.axb3 a2 29.Ra1 c2 30.Bf4 Ra8 and White has to be really careful and is in a very dangerous position. For example, after natural moves like 31.Be3 Nb4 32.d4?, Black gets the advantage after 32...c1=Q! and both 33.Rxc1 Nc2 and 33.Bxc1 Nc2 34.Bb2 Nxa1 35.Bxa1 Rb8 are good for Black.
If White wants to avoid these problems, he could have played 27.Rxc2 instead of the 27.Nf3 that was played, but after 27...cxd2 28.Rxd2 Ra8, any advantage that White has is minimal, if any at all!
And now Black loses his shirt. 28...Bb3 is still better than the move played in the game, but it's not nearly as effective. With the Knight off of f3, White has another trick that gives him a clear advantage. After 29.Bxb3, the move 29...a2 can now be answered by 30.Bg2!, stopping Ra8, and the immediate 29...Ra8 is worse as it gives White time to play 30.d4, which means a lot. After 30...a2 31.Ra1 c2 32.Bf4 Ra3 33.Nd3! Rxb3 34.Nxb4 cxb4 35.Rxa2 and White's winning.
29.Rxc2 Nb4 30.Rxc3 Ra8 31.Rc1 a2 32.Ra1 Ra3 33.d4 Rb3 34.dxc5
This pawn will be a major problem for Black.
34...Rb1 35.Nc2 Kxg7 36.c6 Rxa1 37.Nxa1 Nxc6 38.Bb5 Ndb8
After 38...Nde5, the move 39.f4 is a major problem for Black.
Now, what we see is the Black Knights are dominated, and the extra pawn in return for the piece is useless for Black here. White has a simple approach that is extremely slow, but it would have been enough for me to resign immediately. I would, if I was White, trade on b8. After 39.Bxb8 Nxb8, the Knight is completely dominated, and at any point in time, White could play Bc4 and take the pawn, and then play the piece-up endgame. It would take for ever to win, but White will win it with correct play.
Instead, what White does here is actually quite amusing. The Black King is going to get mated by nothing but pawns and minor pieces!
39.Ba4 Kf6 40.f4 Kf5 41.Bc2+ Kg4 42.Kg2 Na6
And now, while Black takes the opportunity to get the Knights out, White drives the Black King back, and weaves a mating net, partially through the help of Black himself.
43.Bd1+ Kf5 44.g4+ Kg6 45.h5+ Kg7 46.Ba4 Nab4 47.Kf2 Kf6 48.Kg3 e5 49.fxe5+ Kg5
I had reached my hand toward the c6-Knight, and right before I touched it, I realized that 50...Nd4 doesn't work because 51.Nxd4!! a1=Q 52.Nf3+ Kh6 53.Bf8# is mate! The move played, trying to escape, doesn't work either as White has other routes for the Knight besides going through d4.
51.Bf8+ Kg5 52.Nc5 1-0
Black can't stop 53.Ne4# (or 52...f5 53.Ne6#).
This is the first time in the series that we have seen the King's Indian Attack, and as mentioned in the theory article from September, this line is full of tricks and traps, and this game is no different. Both sides end up playing inferior moves that aren't always obvious. Probably the most astounding is White's 23rd move where 23.Bd6 looked extremely natural, but probably leads to nothing better than equality with correct play by Black, in which Black had that opportunity that he threw away on the 27th move, and then completely went off the handle on move 28.
Another aspect that you should pick up from this game is that when the opposing side violates theory, and two ideas become available to you, weigh both options. Here, the ...Nd4 idea was of dominance over the ...a3 idea because it allows Black to remove a passive piece for an active one, and if White doesn't allow that trade, Black gets an extra piece into the defense of his own King. The move 14...a3 is by no means what cost Black the game, but his life would have been easier if he had played 14...Nd4. When you are familiar with an idea that is supposed to be good, and Black was in totally fine shape until his 18th move, at which point the game went back and forth between better for White and equal, always be on the lookout for an even better move before playing the automatic move.
That concludes this article. Good luck in all of your French games, Black or White!