Challenging the Sicilian with 2.a3!?

The Bulgarian chess publisher “Chess Stars” has put out some fantastic literature. Their opening series according to Anand/Kramnik are well known and sought after, and their book entitled “getting an edge against the grunfeld” is one of the most in depth and impressive high level books I own. Unfortunately, there are a few of their books that are so rare and hard to find that they practically don’t exist. “Challenging the Sicilian with 2.a3!?,” by Alexei Bezgodov is one of them.

This repertoire is really based around the wing gambit, however, before getting to the meat of the theory, it’s necessary to cover a lot of ground, since playing 2.a3 leaves options open for black to play many different moves.
Playing 2.a3 against 1…c5 leaves the direction the game will take squarely in black’s hands. Thus, it is a kind of waiting move, though the author is careful to point out that it is a “high class” waiting move. Unlike other systems that allow a player to escape playing the open sicilian, such as the c3, or closed variations, here the nature of the game can go in many directions, and white must be familiar with many different kinds of positions to succeed.

However, the main idea behind playing 2.a3 is to prepare the pawn thrust b4, and thus attain a kind of “improved” wing gambit against the sicilian. This works better than you might think, and often involves lifting the queen’s rook to a4. The resulting positions are often odd looking but fun to play.

Possible second moves for black include: e5, Nc6, d5, Nf6, g6, e6, and d6, all of which lead to very different kinds of positions.


Obviously, the least “sicilian” looking of these moves is e5, and so I will mention a few things about this move first. Where black plays e5, the game takes a turn towards the open games. White should respond with a vienna/king’s gambit set up, though the position may also take on a giocco piano flavor. The addition of the moves a3 for white, and c5 for black, will make the positions unique, such that the arising positions will not be pure and simple vienna/king’s gambit games. White will argue that the black pawn doesn’t belong on c5, while a3 is a good positional move. To illustrate this, compare the following two positions:

Anand-Van Wely, Monaco 2003

Kroshk-Shasa, internet, 2004

first of all, I should mention that in first diagram it is black’s turn to move, while in the second diagram it is white’s. This in and of itself is enough to distinguish the two diagrams, but now let’s look at the positions. Black’s idea in the first diagram is to play a6, preparing b5, winning a tempo off the bishop and expanding on the queenside. In the second diagram we see that the pawn on a3 gives the bishop a nice escape square, thus taking the teeth out of this plan.

Here is a nice instructive game using this vienna/king’s gambit set up in this variation:


It is also worth mentioning that instead of playing 3.Nc3 and heading for a vienna setup, white may prefer to play 3.f4 immediately, giving us this uncommon looking king’s gambit position:

There are two things worth saying about this position before we move on. (1) white’s pawn on a3 both gives the bishop a nice spot on a2 after Bc4, and also controls the b4 square so that no black knight or bishop can harass white on the queen side. (2) One of the most annoying lines to face for a king’s gambit player is the King’s gambit declined with 2.Bc5, in fact it is recommended in several recent opening books on how to play against the open games. Here, that variation is impossible because the black pawn on c5 prevents the bishop from coming to this useful square! Thus, this is already not your normal king’s gambit, and particularly in a blitz game where your opponent is reliant on a tightly knit repertoire, the move 2.a3!? may prove to be as useful as it is surprising.


Another way in which Black may meet 2.a3!? is with the principled 2…d5. This move, which is popular against the c3 sicilian, is equally playable here, however the flavor the game takes is more scandinavian than sicilian. After the moves 3.exd5 Qxd5 4.Nc3 the queen must move, but where to is the question. the check on e5 isn’t best, as white is able to play Be2 and rapidly develop while black flounders positionally with his queen and king stuck in the center. Of course, 4…Qa5, which would be the mainline scando move is unavailable because of the pawn on c5, while Qd8 is simply too passive. So, the obvious answer for the queen is to put her where Tiviakov would: d6, giving us the following position:

White plays 5.b4! immediately, seizing the moment. Black has three logical ways to respond with (a) Nf6 (b) Nd7 and the critical (c) cxb4.

After 5…Nf6 6.Nf3 Bg4 7.bxc5 Qxc5 8. Rb1!
(black can’t adequately defend the b pawn, for instance, if the logical b6, then 9. Bb5+ Nbd7 10.h3 Bf5 11.0-0 a6 12. Bxa6 Bxh3 13. Bb5 Bf5 14. Re1 e6 15. Ne5 Bd6 16. Qf3 Rc8 17. Nxd7 Nxd7 18. Bxd7+ Kxd7 19. Na4 Qxc2 20. Nxb6 +-)
8…Qc8 9.h3 Bh5 10.g4 Bg6 11.Ne5 gives us the final position:

White has an impressive lead in development and arguably a space advantage as well. It shouldn’t be too difficult to turn this into a material advantage.

Line (b) 5…Nbd7 6.Nf3 Ngf6 7.Bb2 cxb4 8.Nb5 (though axb4 is also interesting)

8…Qb8 9.axb4 a6 10. Na3 e6 11.b5 axb5 12.Bxb5 Here the material is equal, but white’s position is clearly preferable.

Line (c) 5…cxb4 6.axb4 Qxb4 7,Nb5 Na6 8.Qf3! giving us the the diagram:

White has some serious threats, and black’s only move is not easy to find: 8…Qc5! White is facing the check on e5…so:
9.Rxa6 bxa6 10.Qxa8 axb5 11. Bxb5+ Kd8 (obviously if 11…Qxb5 then Qc8#) 12. c4 e5 13. Nf3 Bd6 14. Ng5! (this move wins-and now black’s position crumbles in all variations).


Where black plays 2…Nf6, the game takes a turn towards the Alekhine’s defense, and white should accordingly play 3.e5. Note that 3…Ne4?? loses a piece to 4.d3.

After 3…Nd5 the options for white on move four are open to 4. Nf3, 4.c4, and 4. d4. Theoretically, 4.d4 is the best option for white to fight for an advantage, and the theory of the line is complicated, and I will not include it here. However, I prefer the simple 4.Nf3 as in the following game, played by Lawrence Day, who is a hero of this variation and perhaps one of the first to ever use 2.a3!? in tournament play. This game illustrates nicely how white should play against the 2…Nf6 lines.


Responding to white’s odd second move with g6 is logical, in fact it is probably the move that I would play if surprised by 2.a3!? It goes without saying that the arising positions share little to no similarity to the dragon proper, however the key feature of these positions is black’s fianchettoed dark square bishop. After 2…g6, white has two logical options: (a) c3 and (b) b4.

Since the move b4 is more in keeping with the ideas behind this opening, i will only give the following sample game as a taste of the complications that arise after 3.c3. The tactics are complicated, as you can see, and white invests some material for a long term python like grip on black’s king.

Moving on to line (b), after 3. b4. It should be clear by now that this pawn thrust is the basic idea in nearly all variations of the 2.a3!? system. After 3…Bg7 (it’s worth noting that 3…cxb4 4.axb4+/= eval, since freeing the rook so early gives white an enormous development advantage) the best move for white is 4.Nc3, giving us the following diagram:

Here, black has several options: (a) 4…cxb4, (b) 4…Nf6, (c) 4…b6, (d)…d6.

line (a) 4….cxb4 is inferior because it activates white’s queenside play, for instance 5. axb4 Nc6 (if Qb6, black will win a pawn but achieve an unplayable position) 6. b5! Nd4 7. Nf3 d6 8. Nxd4 Bxd24 9.Ra4! Bg7 (this rook lift is key to the entire 2.a3 system.) 10. Bb2 Nf6 11. Qa1! (this queen move is thematic as well, pressuring the dark central squares and creating a battery on the a-file. It’s difficult to dispute white’s lasting advantage in this position.

line (b) after 4…Nf6 5. e5! (a kind of alekhine’s defense, but the knight must go to the odd h5) Nh5 6.bxc5! Bxe5 7.d4 Bg7 8.Nf3 0-0 9.Rb1 Here, white’s queenside play as well as his advanced central pawn duo give him the better chances.

line (c) 4…b6 is perhaps the most logical reply. It supports the c pawn and mobilizes black’s forces on the queenside. After 5.Nf3 Nc6 6.Rb1, black has a number of options, but the position seems much more solid than in the other lines.

line (d) 4…d6 is the best move. Black defends his c-pawn, and controls the central dark squares. Here, white has many moves that work, from the thematic 5.bxc5 to the position 5.g3, but my choice is 5.Rb1. This move resembles line (c), and seems to be the most solid choice. After 5…Nf6, 6. g3 0-0 7.Bg2 Nc6 8.Nge2 Rb8, the following position has been reached:

This position seems to be roughly equal, however in practice it has favored white.


Where black plays 2.d6 without Nc6, he is likely heading for a scheveningen structure. Here, the basic idea is to play b4, park the bishop on b2, and play for a kingside attack with 6.g4 just as you would in a keres attack. The key to playing this position isn’t theoretical knowledge but rather a stomach for considerable risk and a thirst for blood. A typical continuation might me 1.e4 c5 2.a3 d6 3.b4 Nf6 4. Nc3 e6 5. Bb2 Be7 6. g4!

Naturally, other moves are playable here other than 6.g4, such as f4, however, the text move is most in the spirit of this sideline, and if your going to go for it, you might as well really go for it. For the attacking player, this position really plays itself. Obviously, knowledge of the Keres attack is helpful here, as it certainly resembles it, at least on the kingside of the board.


The line with 2…e6 is solid, and black adopts a safe sicilian pawn structure reminiscent of a Taimanov. After 3.b4, black can play either (a) 3…b6, (b)3…cxb4, or (c) 3…d5(which resembles the scando lines above).

After 3…b6 4.bxc5 bxc5 5. Nf3 Bb7 6. Nc3 Nc6 7.Rb1 Rb8 8.Bc4 Nge7 9.d4 cxd4 10.Nxd4, a lively position has arisen, with some of the familiar themes from earlier lines.

The following is a game that illustrates some of the ideas in this variation.

The other possibility for black, 3…cxb4, is stronger.
The following game covers the ideas in this line nicely, and is a good place to start.


After 2….Nc6 3. b4! (what else?) Black can choose to decline the pawn, though accepting is likely best. 3…cxb 4. axb4 Nxb4 5. Ra4! gives us the following position:

This position represents a main line tabiya for this entire system. From here, the game may take many different directions, from the solid to the surreal. Black has six different options to choose from here, so many in fact that I won’t bother to list them, as you can probably guess them all from the above material. A key feature of the position is obviously white’s rook on the queenside, which has the ability to come into play rather early and often does swing to the center of the board. Another key concept in this variation is to play white’s queen to the a1 square, thus hitting the dark central squares and creating a battery on the a-file.

The following game represents a nice win for white in the “main line” of this system, if there is such a thing. Note that in the game, white plays d4 instead of the rook lift. The rook lift is really the new key idea which Bezgodov presents in his book, arguing for it in nearly all lines where it is a legal move. I like this thematic approach, and am convinced that the move Ra4 is not as insane as it looks.