The Dunst: Attacking the Caro-Kann with an Early Queen Sortie

Usually, surprising sidelines tend to be objectively dubious, but the following is totally sound according to several sources (I’m keeping some of them under my hat.) The idea I have in mind is similar to the Caro-Kann “two-knights,” which occurs after 1.e4 c6 2.Nc3 d5 3.Nf3. Here, white holds back the d-pawn at least for a while and simply develops. This line was a favorite of Fischer’s, but it doesn’t cause black any immediate trouble and his plans are straight forward and in keeping with the themes of the Caro-Kann. However, after the move 3.Qf3!? instead of 3.Nf3, we have the following eye popping position, which may occur through several move orders but properly belongs to the 1.Nc3 Dunst opening more than any other.

I hope to show that this position is more than playable for white- it is sound- and moreover black has no clear path to equality. This may be reached through a number of different move orders, including the Scandinavian 1.e4 d5 2.Nc3 c6 3.Qf3, as well as through the Dunst via 1.Nc3 d5 2.e4 c6 3.Qf3. Here, there a number of moves that look playable for black. There are: d4, e6, Nf6, e5 and dxe4. Black is immediately confronted in the center and must decide on a plan early. This position will almost certainly be a surprise to your opponent, who will no doubt think that the queen sortie is incorrect and therefore try to “punish” you. The upside to this is getting out of book and playing chess. The queen sortie is perfectly sound and supported by some theory, and this idea was first pointed out to me by Asa Hoffman who uses it regularly. He said he got it from a book on 1.Nc3 by Keilhack, titled “Der Linksspringer.” Since then, I have seen it in a few different places, and even had a brief conversation with Yaacov Norowitz (who plays the caro-kann exclusively) about it. Norowitz seemed to think it wasn’t that frightening for black, but admitted that he had wrestled with how to meet it and decided upon the straightforward move 3…e5, but seemed a little uncertain still.


Objectively best is 3…dxe4, but before we get there you’re probably curious about what happens after the tempting push 3…d4 hitting the knight. The answer is that white ignores it with 4.Bc4, when the pressure of the f7 square is worth the knight and then some. My engine gives the following as best play in that line

The mainline for the variation with 3…d4 goes 4.Bc4 Nf6 5.e5 dxc3 6.exf6 cxd2+ 7.Bxd2 exf6 8.0-0-0 when white clearly has more than enough compensation for the pawn. The following game is also an instructive example for the line:


Now, white can hit the center with 4.d4, when 4…exd4 is still the best move. If black plays 4.Nf6, as in the following game, white should emerge with an advantage, as the following game fragments demonstrate:


As I mentioned before, this was the move that was favored by Norowitz during a brief kibitzing session. In my opinion, this seems like a perfectly logical response, yet again it isn’t as solid as the simple 3…dxe4.

This game shows a straightforward plan for white against this move. Simply push e5, play Qg3- prepare f4, and play for an attack.

THE MAIN LINE 3...dxe4 4.Nxe4

Finally, we get to the mainline. Here, black will likely play Nd7, Nf6, and challenge the white center with a pawn break. White can either play conservative from this point forward, holding onto his strong point d4, or white may gambit the d pawn for an aggressive initiative.

The following represents black's most testing attempt to refute white's gambit of the d pawn. If this swashbuckling chess isn't to your taste, below, find a sample game as well where white simply plays c3 and looks for a more quiet attempt to play for an edge.

The following game shows a straightforward game in which white simply holds onto the strong-point d4 by playing c3 and looking for a small long term plus as opposed to a knock out attack.

The following game, played by Asa Hoffman the champion of this line, illustrates some ideas for white that involve the simple and solid move Ne2.