Classical Caro-Kann

The classical Caro-Kann has the following starting position.

Though the “propperly” classical variation occurs after 5.Ng3 Bg6, we will consider one early sidelines before delving into that. I mean the interesting 5..Nc5!?.

The move is somewhat unprincipled but has been tested out by high-caliber players such as Carlsen, Ivanchuk, Radjabov, and Svidler. Superficially, the idea is the same as the more typical 5.Ng3: to move the knight to safety with tempo (the b7 pawn is hanging). However, Black can equalize relatively quickly with an early e5 break, exploiting the fact that the d-pawn is overloaded.

After 5…e5!? 6.Nxb7 Qb6 7.Nc5 exd4 8.Nb3, the pressure on the d-pawn is released with 8…Bb4+ and Black faces essentially no problems, leading in development and central control.

If White tries 8.Na4 instead of 8.Nb3, attacking the queen instead of the pawn, the simple 8…Qa5+ 9.c3 dxc3 10.Nxc3 Bb4 leads to standard play.

Having said that, we now turn to consider the typical setup occurring after 5.Ng3 Bg6. Though White’s main response here is 6.h4, seeking kingside expansion, a few sidelines are worth considering: 6.f4, 6.Nf3 and 6.Bc4.

The position after 6.f4 is seen below.

Of course, the idea is to get a strong grip of e5, playing for a bind á la Maróczy. Black should aim for solid play, typically has no rush to castle, and sets his pieces as seen below.

The knights on e7 and d7 support the bishop on g6, prevent potential f5 breaks, and support the c5 break. Black’s position has no weaknesses. Black will eventually trade on d3 and get a pleasent game.

The 6.Nf3 variation is equally timid.

Black should respond with the typical Nd7, which prevents Ne5, and it is possible to transpose to the main 6.h4 line within a few moves. This line is very drawish and deserves little comment.

The last sideline we will consider is the move 6.Bc4!?, which after 7.N1e2 might be followed by $(a)$ castling, $(b)$ h4 or the mainline $(c)$ Nf4.

The idea of this setup is to play for the f5 break, which justifies Bc4 and N1e2. If White plays 8.O-O it is imperative to know the refutation, which consists of transfering the knight on b8 to b6 via d7, forcing a trade of LSBs (the bishop on b3 would be passive) and then playing g6 the crucial move g6.

If in this setup White feels adventurous and attempts f5 anyway, gxf5 should give Black good play. The sacrifice Nxf5 complicates the position a lot, but does not work because Black can achieve a good defensive setting after Be7-Nfd5, even giving up the f7 pawn.

I believe this Bc4 line is a good attacking weapon for White, since Black must play precisely in order to equalize. On the Lichess database, White won 62% (!) of games after 9.f4.

The move 8.h4 makes little sense, since after 8…h6 Black’s LSB has a retreat square already, making Nf4 a weaker idea. So more worthy of consideration is the immediate 8.Nf4. Of course, Black can never play h6 now, and the correct defense is to exert pressure on the knight quickly via Bd6 and Qc7. As seen in the line, this can lead to an early exchange of queens and a safe game for black.

Advanced Caro-Kann

Short variation

In the Short variation of the advanced Caro-Kann, White plays the seemingly quiet 5.Be2. Curiously, White’s argument is that he needs no special arguments! White claims Black’s self-induced slow development and tempo-lag is sufficient to claim superiority, and nothing special is needed on his end. The line has some sharp and interesting variations.

Short variation: Main line

Here, White has many interesting options. We will begin with the sharper ones.

The first variation we will examine is what happens if White plays 8.Bg5!?. This idea carries two positional threats. $(a)$ When pinning the e7 knight, White hinders the liberation of the f8 bishop, making d6 weak. A slow reply like Bg6? would be met by Nb5! and it is impossible to prevent White’s knight from penetrating. $(b)$ Secondly, the move prevents normal development (e.g. Nbc6) due to the threat of Nxf5. The pin on the e7 knight would force Black to reply exf5 and enter the middlegame with a weak structure.

Having understood those threats, it’s clear that Black must unpin with tempo, so as to continue with development normally.

After 8…Qa5+ 9.Nc3 (other replies are harmless) Nbc6, White has two critical ways to continue. The stubborn one is to insist on getting a knight on d6 via Ndb5, while the other is to play Bb5.

After 10.Nb5!?, 10…Ng6 is pretty much forced, which leads to tactical complications because now Black’s LSB lacks retreat squares and therefore g4 is lurking.

The immediate 11.g4 would not be possible because the e4 square is available for the bishop, and f3 would be met with d4, a sharp pawn sacrifice which gives Black a lot of activity.

Thus, White always plays 11.a3 threatening b4, but leading to nothing conclusive after 11…Be7 12.Bxe7 Kxe7! 13.b4 Qd8 14.Nd6 with a very interesting position.

Tactically, a property of the position is the overloaded e6-pawn, which prevents Black from immediately playing 14…Ngxe5? due to 15.Nxf5 exf5 16.Qxd5 and Black loses a pawn, has the inferior structure, and the unsafe king (+-). But it is essential to capture on e5 and weaken White’s strong knight, so Black must play 14…d4! to release the pressure on d5. After 15.Nb5 Ngxe5 the position is equal, but White’s knights will dominate and the b7 pawn will most likely fall. I think White has good winning chances.

White’s most common reply, however, is 10.Bb5, which poses a threat but is far less complicated.

The threat clearly is 11.Bxe7, leaving the Nc6 unsufficiently defended. But after 11…Qc7 White has nothing better than castling, and Black can continue with the usual Bg6 Nf5 Be7 and castling. Perhaps White can try 12.Qg4!? with the idea of h4-h5 and kingside aggression. The key is that Black must play a6 and, if White responds Ba4, play h5 himself!

Having covered the curious Bg5 line, we may proceed with my personal favorite (for White, not for Black!), 8.c4!?.

The move is clearly logical: it fights for the center, and since Black is lagging in development, the opening of lines should benefit White. Black must dready very carefully in order not to lose quickly. In the Lichess database, White wins 61% (!) of games after 8.c4.

Of course, taking the pawn via 8…dxc4 makes little sense. Black should worry about getting his pieces out. After 9.Nd2 Nbc6 10.Nxc4 (aiming for the outpost on d6), White’s position is already to be favored. The only move which prevents an explotation of the outpost is Nd5, where White can simply castle and enjoy a good game.

So, the correct response is 8…Nbc6, continuing development. If takes Black’s point of view (as we are doing), it is clear that we should strive for rapid simplification. Thus, slow moves like 9.Nc3 fall right into our hands and pose no problems: 9.c3 Nxd4 10.Bxd4 dxc4 11.Bxc4 Nc6 (or 11.Qa4+ Nc6 12.Rd1, a bit more agressive, but harmless after 12.Bd3! an idea which will apear again later.) It is in a similar fashion that Black responds to the equally slow 9.O-O.

Thus, after 8…Ncb6, White’s only way to keep the initiative is the sharp 9.Qa4, reaching the following very interesting position.

The anti-positional response 9…Qd7 seems to work fine. However, the modern solution is the simple 9…a6, after which White has three options which must be seperately considered.

$(a)$ If White castles after 9…a6, Black should resolve central tension via 10…cxd4. This exploits White’s Qa4 move, since now Bxc4 is impossible for obvious reasons. If Black is allowed to play b5 he will consolidate, so White must reply with energy, its only move being 11.Rd1.

11.Bf3 merits consideration. But after 11…b5 12.Nxc6 bxa4 13.Nxd8 Rxd8, Black should survive: the initiative is gone. See Portych vs Nouveau, 2023.

After 11.Rd1 Black again has the resource 11…Bd3!. Black must continue to look for simplification: 12.Bxd3 cxd3 13.Rxd3 Qa5! and White has nothing better than trading queens. The normal play should ensue.

We see then that 10.O-O poses no major issues.

$(b)$ So White might try 10.Nc3, the main move and the one leading to sharper play! Black will again explot the position of the queen with dxc4, only to see 11.O-O-O!? with a very interesting position. Here Black must play 11…Qa5 and simplify that way: everything else fails.

11…Bd3? now does not work: after trades occur and 13…Qa5, now White can decline the trade with Qb3, and the position is very tricky. For instance, it seems that White can play 14…Nxe5, but after 16.Rdd1 it is very hard to develop.

11…Qc8?! also fails due to 12.Nxf5 Nxf5 13.Bd6! and White will get the queen for a bishop and a rook. See the game Fabiano Caruana vs. Sebastien Cossin, which Caruana won nicely, for an example of this specific line.

After 11…Qa5, White really has nothing better than simplification via Nxf5 and Qxa5, in whichever order. Avoiding the trade with 12.Nxf5 Nxf5 13.Qc2 allows 13…Nxe3 and Black is much better. That said, though a very strong player will probably hold the position resulting from the trade of queens, it is clearly White who’s pushing.

Exchange Caro-Kann

One of the advantages of the Caro-Kann over the French defense, from a psychological perspective, is the fact that the exchange variation, though less entertaining than others, is not that dull. (The French defense player knows of the somniferous effect of the exchange French.)

We consider the following position, with White playing Bd3 and c3, as our starting one. Other attempts by White pose no problems.

Double pin variation

However, before going into that, I would like to take a look at a variation I face frequently in Blitz, and which is very popular in Rapid as well. I denominate it the double pin variation, though it lacks an official denomination. See the starting position below.

Double pin variation: Early castling

The hyper-aggressive h3-g4-Ne5 line

First, let’s consider a few popular lines. By far, the most common move is 6.O-O, so let us look at that variation first.

6.O-O is a harmless variation and Black should continue normally.The most common is to see 6…e6 7.h3 Bh5, mantaining the pin.

The weakening 8.g4 followed by 9.Ne5 is the second most common continuation, so we need to know it. The necessary reply is 9…Ne7, defending c6 and g6. If White over-extends with 10.f4 we simply reply a6, drive the bishop away, and we are fine.

If White attempts the artificial 10.Bg5, it loses a pawn by force after the natural-looking 10…f6.

If White strikes with 10.c4, we take the opportunity to harass the bishop and the following exchanges favor us. (If White does not trade, the line 11.Ba4 dxc4! wins immediately.)

Lastly, 10.Nc3 gives nothing: 10…a6 and Black’s fine.

So, it seems like the hyper-aggressive approach of 7.h3 followed by 8.g4 and 9.Ne5 is not good.

Qd3

The most common reply after 7.h3 Bh4 is 8.Qd3, unpinning the knight. But this poses no problems to Black after 8…Bd6. Again, White has options for its ninth move.

9.Bg5!? f6 accomplishes very little. Black will play Ne7-Bf7 and the position is fine. It’s not clear how White should exploit the e6 weakness., since it cannot be targeted by any of White’s minor pieces. Black has won the battle for the center.

9.Ne5 Ne7 is also fine for Black. Even after the pressing 10.c4! Black simply castles and it’s a game.

Anyhow, in almost all variations, Black simply plays Ne7, short castle, and everything is okay.

Other lines

Other lines in the double pin variation where White castles early require no particular attention.

Double pin variation: immediate h3

If White immediately replies 6.h3 to Bg4, we are happy to trade immedately. After White castles, it is best to immediately challeng the bishop with a6: whichever way White replies, this will hinder a c4. If White plays c3 we develop normally, since c4 would make less sense. There is little to say about this variation as a whole.

Main line of the exchange variation

Recall the main position of the exchange variaiton, where White plays the setup with an early Bd3 and c3, with Black responding with Nc6 and Qc7.

Main line: 6.f4

The least challenging setup for White might be the weaking 6.f4?!, which is rare.

The move 6.f4 allows for 6…Nh6 with a subsequent Bf5, solving all of Black’s problems. White will place a knight on e5 eventually, trying to justify f4, the placement of the knight on h6 makes f6 possible, if push comes to shove. Nothing special really.

Main line: 6.Bg5

This is a weak attempt at preventing Black to develop comfortably with Nf6. However, that’s precisely what Black should play, since 7.Bgf6 gx6 gives Black no problems. White will play e5 whatever White plays. Even if White pressures d5 and f6 via Qf3, e5! gives a lot of compensation for the sacrificed pawns. This line deserves little attention.

Main line: 6.h3

6.h3 is one of the most common moves, and the one which is coherent with White’s Bd3 move. The idea is to complicate the development of Black’s LSB. To me, the most pressing idea for Black is to go for a Nf6-g6-Bf5 setup, allowing for White to trade on f5 and double Black’s pawns.

The resulting lines of play will involve long castling, e6, Ne4, etc. Standard ideas. The open g-file favors Black. Whether Black’s DSB belongs to g7 or d6 merits considerationo, but it seems to me d6 is more active.

Main line: 6.Ne2

The move 6.Ne2 is an attempt at getting Bf4 in. To my eyes, it is a logical attempt to exploit Black’s early Qc7 move. However, it allows 6…Bg4, which solves Black’s LSB problem (even if after f3 it seems like it doesn’t).

The point is that after 6…Bg4, White cannot continue with 7.Bf4 due to 7…Qxf4!, entering an equal queenless middlegame where Black is doing perfectly fine. It seems clear that after 7. Bf4 Qxf4 8. Nxf4 Bxd1 9. Kxd1 e6, Black should be able to comfortably carry out a minority attack on the queenside. This is particularly easy now that White’s DSB does not control the h2-b8 diagonal, making Rb8 ideas possible. And without the queens, it will be difficult for White create counterplay on the kingside.

It is for these reasons that 6.f3 is mandatory, entering a variation which leads Black into an interesting isolani position. After

1. f3 Bd7 (the only retreat which gives anything) White will continue with his plan of 8.Bf4. However, here Black has the strong 8…e5! After 9. dxe5 Nxe5
2. O-O Bd6 11. Nd4 Ne7 12. Na3 O-O, Black’s position is excellent.

Fantasy Variation

The starting position of the main line Fantasy Variation, with 3.fe, goes as follows.

This is a sharp position with a centre in strong tension. Black correctly overprotects e5. White has a series of optoins now. We will first consider the agressive 9.Qb3

Fantasy Variation: 9.Qb3

The move 9.Qb3 attacks b7 and f7 simultaneously. The correct reply is to simply castle.

If White plays 10.Qxb7, there is a very interesting and forced line which exploits combinatorily three positional factors which make White’s position vulnerable: $(a)$ the bad positioning of the queen, $(b)$ Black’s lead in development, and $(c)$ the lack of defenders around the White king, itself a consequence of under development of the queenside pieces.

This line is so pretty that I want to comment it separately.

The first moves involve removing the Kf3, defender of White’s kingside, followed by exd4 with the idea of opening up our bishop’s diagonal. On sees in the resulting position that the elemnts for a kingside attack are there. First let us discuss the line where White responds 12.gxf3 to 11…Bxf3.

After 11.gxf3 exd4 12.cxd4, Black has a beautiful piece sacrifice: 12…Nxe5!. The idea is to give our queen some freedom of movement. After 13.fxe4 Qh4 the response 14.Rf2 is forced, since no other move prevents Qxh2 mate.

The resulting position leads to a new sacrifice with 14…Nc5!, attacking White’s queen. If the knight is accepted, mate is unstoppable after Bxc5. The only sensible move for White’s queen is 15.Qxc6, where after 15…Nxe4 again we have a forced line:

16.Be3 (only move) Rc8! (with a skewer) 17.Qd5 (or Qb7) Nxf2! 19.Bxf2 Qxh2+ 19.Kf1 Rxc4! 20. Qxc4 Qh3+ 21.Ke1 (only move) Re8+ 22.Kd2 Qg2 getting back the piece. Black is totally winning and the line is forced.

Now let us see what happens if White responds 11.Rxf3. Again, we open our bishop’s diagonal with 11.exd4 12.cxd4 Ng4!

12…Ng4! comes naturally to mind if one understands that after 12…Nxe4 we are not threatening anything. h2 is falling with check and White should ignore it and develop with Nd2. However, the most common replies are all attempts at saving the pawn.

13.e5 fails 13…Ndxe5! 14.dxe5 Bc5+ 15.Kh1 (only move) Qd1+ 16.Rf1 Nf2+ with forced mate. If 16.Bf1 is played 16…Qxc1 is sufficient to win.

13.h3 fails to the same 13…Nc5! 14.Qxc6 Bh2+ followed by Qxd4 wherever White goes.

13.Bf4 fails to 13…Nc5! as well. After 14.Qxc6, Black trades bishops and throws in the deadly Qxd4+.

As we can see, after 10.Qxb7 Bxf3! almost all lines are forcing and lead to a victory.