Finally understanding the Caro Kann ...Bf5 line (part 2)

I’m slowly going crazy about this very specific Caro line. So far, everything went smoothly, and all 3 previous parts came into existence without any troubles. This final one, oh boy... For some strange reason I just cannot be bothered to care about the opposite sides castle line. Whilst most people enjoy launching fearsome kingsides attacks, I just think playing for checkmate is a very impolite strategy. Only cavemen go straight after their main goal. We are civilized people and we should aim for more nuanced strategies. Just me saying.

Good to mention is that the previous work has helped me a little. Got a draw against a GM in a blitz game in the Nh3-line, that was pretty nice. Haven’t encountered the main line(s) very often, but I’m sure it will come one day, when I vaguely remember I once analyzed it...

So, here we go one last time, the ...0-0 line. I liked the ‘how not to play’ section last time, so I’m going to show more examples of things that can go wrong.

How things change by going 0-0.

The first thing that springs to mind when talking about opposite castling, is attacking possibilities. I learned as a kid that both players should attack on the flanks and the first one to the enemy king usually wins. However, I am tempted to say this is a straight-out lie. Very often, one player is just waaay slower in its attack, and you’ll end up wasting moves by moving random pawns forward whilst being crushed on the other side of the board. Whilst going through the database games, I got the impression that black is often quite slow with its attack, and when white attacks on the flanks, you should push in the center. Of course, it all depends on circumstances, because I have also seen games where blacks goes b5 and a5.

The games of thousands of masters in the database show that going ...0-0 is perfectly fine for black, but it does require knowledge of some precise defense. Allowing a well-timed g4! sacrifice by white can be very dangerous. Not accepting this pawn allows it to go another step to g5, and a file will open. Accepting the pawn allows a white rook to go opposite the black king, pinning the g7 pawn, after which h6 becomes undefended. The line I show below is a good example. Black goes ...Bd6 and ...0-0, after which g4 causes trouble.

However, like all things in life, it depends on the details too. In the example above, a vital element is e6 being targeted by the white queen. In other situations, like the line I show below, the g4 sacrifice is more easily accepted. I must admit though, it does still look a little scary with this opened g-file.

There are many possibilities, but an important defensive resource which makes opposite sides castles viable for black, is the use of d5 for the queen. After chopping and moving some knights, the Q can make use of central d5 square and offer all sorts of queen trades. (Targeting a2 is a nice bonus.) If the queens are traded on for example e4 or b5, the remaining position gets very dry. Black is really stable and the kingside attack is not longer a major concern with so much material already traded. Avoiding the queen trade is more ambitious for white, and the game continues, but black has at least stirred up enough trouble to give himself time to defend.

Because of whites idea of attacking with g4, the clumsy g3-knight is often moved forward to e4, effectively always trading it. There are some lines I noticed in some GM-games, which consists of not trading this knight. The idea of playing Qe2 and then jumping into the center with Ne5 is available also here again. I don’t have much more to say then: just don’t do anything stupid with black here. Like, trading on e5 with your Q still on d8, allowing Bxh6. Don’t do that. Otherwise, there appears to be several options.

An idea which I do really enjoy, is going Rhe1 and Qe2, creating a pin on the e-file. Then, the clumsy g3-knight finds itself a nice square on f5. If the knight stays alive, it can often be brutally sacrificed on h6. It looks crazy at first glance, but it’s a theme I found in at least one game. The sacrifice shows once more how careful black must be with his king. White really has some space to work with, and the pieces can start moving towards the monarch. The line below shows an (artificially crafted) example of the principle.

The high-level way to go 0-0

Clicking through the database there is this funny line that’s often played, see the board below. It involves white going Bf4 as a normal developing move, then black gives a check on a5 and after the Bd2 retreat, black puts its bishop on the diagonal as well, on b4. Black doesn’t seem to gain particularly much with this maneuver, as the white c-pawn pushes the Q and B back to headquarters. If we compare the position with the direct Bd2 line above, we notice the difference that white has its pawn on c4 whilst black the queen on c7. In main lines, white likes to have the pawn there anyways, so why would black do this?

There’s many answers to this question. The queen is simply reasonably developed on c7, to start with. But another interesting idea is to undermine the white center with ...b5! The c-pawn is the only obstacle stopping black from having the d5-square for himself forever, so damaging the white structure makes sense. The king is on the queenside too, so if b5 is lost in the process, black wouldn’t mind. If white pushes the pawn to c5, the structure gets fixed. It’s easy to think that we should go all-in with a pawn storm (e.g. a5-a4 or b4) towards the king. This is pretty slow however, and white is in no danger here. The games I checked followed a different pattern: white tries to attack the black king with a pawn storm or rook lift made possible by the space advantage, whilst black tries to defend his king and plays in the center, where he has some long term positional advantage (d5 being a strong outpost).

Delaying castling altogether

Explaining things to others can be very misleading. I almost even convinced myself. I could finish here now pretending I discussed everything worth discussing, and many wouldn’t really notice anything suspicious. In life, you often get away with this (as a teacher), but chess is much less forgiving. Whenever you think you know everything, your opponents will always manage to surprise you with something random you’ve never seen before. Splitting the topic into two parts, long castles and short castles, sounds very covering, but excludes the option of black not castling.

It’s not so strange if you think about it. White answers with a kingside attack as a response to ...0-0, and with central action against ...0-0-0. Leaving the black king in the center a little longer, leaves white with some uncertainty. Launching (and sacrificing) g4 is often premature when the king is not on g8 yet. Going for the central plans is fine, but black will then invest a well-timed tempo into ...0-0 or ...0-0-0, and the white attack will come with a lot less poison. Black does need room to insert this tempo, which isn’t always easy when the center blows open. The statistics seems to like delayed castling though, as the win rate for black is often higher than the win rate for white.

Have I solved chess now?

Having done absolutely everything within my abilities to contribute my part in our collective struggle to solve chess, I’m going to take a look what to do next. Perhaps I’ll find some other opening I’d like to study, but perhaps I’ll focus on some other topic.