Finally understanding the Caro Kann ...Bf5 line (part 1)

And the series continues with what I hoped would be the final part, but whilst doing my initial research, I quickly realized that, nope, this ain’t gonna fit. Have you ever had these presentations where the presenter runs out of time, and compresses half the presentation into the remaining minute? I don’t wanna be that guy. And lucky me, I can do whatever I want and set and break my own rules. So, the final part will be split up into black going 0-0 and 0-0-0. This will be the long castle part.

I cannot help it, but there is this voice in my head screaming that my working method isn’t efficient. Wouldn’t it be much better to spend a little time on multiple openings, instead of going all-in on this line you will probably play almost never...? Yes, but of course that’s true! But that ain’t no fun. Everything in this world is already about productivity and efficiency, and let’s give ourselves a break when playing board games, alright? Just my take... Others might disagree.

So, I push back the voices in their little dark box where they came from, as any sane person would do. And whilst chilling with Dire Straits from my speakers I continue my hunt for knowledge.

Some more moves in the main line

Before playing any moves, let me begin with explaining both players' general plan, and after that we will look at some concrete variations. So, we found in the previous part that going 0-0 for white is fine, but not very challenging. More exciting is using the kingside to attack. We use the bishop on g6 as a jump board to launch an early attack there, and we dump the clumsy king somewhere on the queenside. We don’t want the pesky light-squared B to snipe our king from far away, so it has to be traded. The pawns on f2 and g2 are currently blocked by knights, so we probably have to move our dear horsies first before launching the attack. The e4 and e5 squares are the most logical squares to go to.

As we found in the previous part, white goes h4 and threatens the bishop. Defending with ...h5 is almost never played, and I have some ideas why. Giving up the g5-square, making 0-0 more dangerous, ending up with a weak h5-pawn after the bishop trade: ...h6 seems more reasonable. White could go h5 immediately, but in practice Nf3 is played first, a slight nuance. The knight is looking at e5 and many of whites plans consists of hopping into that square, with some good timing. If black does not respond to this, for example ...Nf6, white is happy to hop in with tempo on the bishop, after which the knight can be supported with f4. Note that white was already planning to attack on the kingside, and gluing the knight into e5 helps quite a lot. After f4 trading on e5 becomes slightly unpleasant for black in many cases. White gets a strong center and a half-open f-file to work with. Therefore, the most popular response to 7. Nf3 is denying entrance by 7. ...Nd7.

Then, white continues the plan by going 8. h5. Bh7 9. Bd3. For some reason, I used to be afraid of trading pieces in the opening. Less pieces means more drawing chances. It’s what the popular people say, right? So yes, I like openings where all pieces remain on the board, and I naturally despise these type of moves. (Like, wouldn’t Be2 or Bc4 be fine developing moves as well? Why always the needless fighting for the same diagonal, when we can peacefully split up the available ones?) But I played some variations, and when white goes 0-0-0 the B on h7 gets really annoying. You want to use the c-pawn, but then the king loses its natural escape route to a1. You want to use e4 for a horse, but it’s defended too much. So in the end, you will play Bd3 eventually, so you might as well do it immediately. For experts this might be obvious, but I had to find out the hard way. But I’m convinced now.

...Bxd3 10. Qxd3. Blacks turn. Somewhere around this point, it gets time to get ready for castling. White hasn’t castled yet, and is in a position to launch an attack on either kingside or queenside, so a general rule seems to be to let white select first, and then follow the same direction. Opposite sides castle is very risky for the black player. In the next part, we will look at going 0-0 when white goes 0-0-0, which is possible, only if we know what we’re doing. I haven’t really found any good examples of the exact opposite: going 0-0-0 when white goes 0-0. And to be fair, whites queenside pawn storm does look pretty menacing. So whilst we politely urge white to hurry up castling, it’s sensible to play some flexible developing moves. ...Ngf6 and ...e6 make a lot of sense, whilst ...Qc7 is also alright, although it hints a little more at preparing for castling.

There is some logic behind ...Qc7 here. White wants to go 0-0-0, so a wild bishop might appear on f4 very soon. How is momma queen going to explain to her kids that daddy castles directly into the line of fire of this evil bishop? We need to be proactive and claim the diagonal now, with ...Qc7. However, once again in chess, this logic turns out to be rather flawed. Say, we allow this by playing ...e6 (...Ngf6 works similarly), then white goes Bf4, then black has the sequence Qa5+. White drives the Q away with Bd2 and we finish with Qc7. Bang! Mission complete, diagonal conquered. Would I myself ever find such a sequence over the board? Well, let’s be real, of course not. But that’s why we learn the entire openingstree. To skip the thinking part. By the way, if you pause for 5 minutes before playing ...Qa5+, you can always bluff you found it yourself. Now... there is this minor detail that Bd2 isn’t forced. The alternative c3 isn’t bad either. But you can keep the Q on a5 for a while (she prevents 0-0-0 by targeting a2 at the moment), and we’ll have a game.

After castles: following some normal games

My initial idea was to check for plans in certain ‘stable’ positions, but so far I spend most time explaining how these positions are reached. But now we are finally there: 13. 0-0-0 0-0-0. (In practice, the slightly more flexible ...Ngf6 is often played first, but I don’t think it really matters.) From here, I just want to see how games typically continue, and which general ideas I spot. So, the next section will just be a collection of ideas I deem important. With every variant I choose, I neglect five others, but that’s life. I just hope to give a fair overview of possible directions a game can be heading.

[3 days later]

Arghhhh. Alright, there’s a million different lines which are all more or less equal, I found. It’s no use trying to make any kind of summary. I’ll just give a few examples, and then I’ll give some examples how things can go terribly wrong. That’s more interesting.

To start, let’s go through some lines that just trade pieces on more-or-less equal conditions. I found one that trades a bunch of minor pieces, and one that trades queens. Black has quite a few attracting options available to trade, and it’s good to name them, I think. The g3- and f6-knights often get traded on e4. The f3-knight might get traded on e5. The tension of the bishops on the f4-square is often reason for trades. The pawn-break ...c5 (and sometimes ...e5) gets rid of some central pawns, after which the open file is susceptible for rook trades. From a GM point-of-view, black just wants to equalize, and these trades help tremendously. For more ambitious lower rated players, the story is slightly more difficult. Actively avoiding trading white central pieces and avoiding the ...c5 break might be primarily a losing strategy. There is still some life left in the endgames, I would say, so just try to push for a win this way.

The first diagram shows a line that involves trading without changing the structure. The good white knight is not enough to really claim a significant advantage. Blacks setup is just too healthy. The second diagram does change the structure, and I can imagine the open f-file helps white a little. However, it will take time for white to start making threats, and the black knights should be able to find squares just as nice as the white knights have.

Keeping pieces on the board is more complicated. An example line is shown in the next diagram, which reveals some typical ideas. First, the N on g3 does not have that many places to go, so it very often ends up being traded. It’s a little sad, to help black with its space disadvantage, but what can you do?... But then again, both N’s are pretty well positioned. The d7 one often helps with the ...c5-push, and the f6 one controls two important central squares. So getting rid of one for the helpless g3-knight is defensible. White also grabs more influence in the center with c4, which can be accompanied by Kb1, if the lord wishes some extra safety. We continue thematically by undermining this center with ...c5. White goes d5 creating a passed pawn, and I can spend five paragraphs checking whether this pawn gets lost or not (in the long run, of course), but I’m just gonna cheat and say that 120 people in the database did this analysis for me, so it’s fine. White has space and targets on the kingside, black has good piece coordination. It’s a game.

Another idea I like is the following: going Qe2 early to stick a knight into e5. The queen is not the happiest on d3 anyways, with the rook looking right at her and the ...c5 break coming. So spending an extra tempo on a repositioning is acceptable. Black is not really keen on taking on e5 (see next section), so the knight will stay alive on a nice central square, attacking f7 as an added bonus. I like it, because it’s not so completely obvious in my eyes how black can deal with this pest, and in the meantime we can shuffle some pieces, pretend we are planning something, threatening something, and then suddenly black might get nervous and make hasty decisions, such as eating the e5 knight, and getting a bad endgame.

How things can go terribly wrong

Now we get to the juicy part: how to not play, a.k.a. how we want the opponent to play. Chess is often not about identifying why strong moves are strong, but more why weak moves are mistakes. Obvious blunders aside, I found some ways both players can go wrong strategically.

One interesting concept that I severely underestimated, is the option for white to go c5. With tempo, if a B is on d6. In normal play, black does not give white this chance, as it’s often black that goes ...c5 first, but this shows the importance. I would have always assumed that the massive hole on d5 and the backwards pawn on d4 would make white cry in the long run. However, black does gets pushed back quite a bit, and white does get some dark squares to work with. The knight doesn’t get to d5 that easily, and whites pieces have some nice forwards moves. This concept depends very much on the details though, I must warn. Like, if we would swap Q and B from e7 and c7, the b8-h2 diagonal is under control, and I think white is just positionally lost.

Another way black can find himself in trouble, is by taking an annoying e5-knight. White recaptures with the d-pawn and can support with the f-pawn, making the center very stable and gaining massive amounts of space. Blacks ...c5 has nothing to bite on anymore, making this position pawn-break-free. It gives white some time to maneuver. The clumsy g3-knight suddenly becomes a proud fighter, as the e4 square is nice and the d6 square is potentially beautiful. If white manages to find a stable way to jump into d6, by trading bishops for example, black might just be lost. At the same time, the g-pawn doing g4-g5 creating kingside weaknesses is a potential danger that will be looming in the position for quite a while. I just realized that I completely contradict what I have written in the previous post, where Ne5 was not too great. I blame the king-placements and rooks-placements (especially the d1-rook) for this discrepancy.

I have spoken very positively about blacks ...c5, so one might think: why not do it directly? This leads me to the third way black can go wrong: doing ...c5 too early. The diagram gives two positions, both with black not having its g8-knight into play. In both cases, everything opens up, and the white pieces are just more in attacking mode at the moment. The rook lift Rh4 is always nice, asking black why the heck he opened up his king. The knights jump around. The bishop is happy on c3, I guess. Just jolly good times for the white player. Whether it leads to some permanent advantage, I don’t know, but black shouldn’t want this from the opening. What I found, was that ...c5 is almost always fine after white has gone c4 himself. With the pawn still on c2, black doesn’t have threats along this file and white has, whereas with the pawn on c4, the opposite starts being true.

I tried to find examples how white can mess up the opening in the database. There are a few things that deserve an honorable mention: losing the endgame because of poor technique, losing h5 with too little compensation and losing d4 in bad accident. These are kind-of game specific though, and are more like tactical errors. What does happen once in a while, is white losing central control. The black rook is already pinning the d4-pawn (there are some pieces in the way, but moving your own pieces is usually not so problematic), the ...c5 break is coming, the Q is opposite the king, and under the right condition, the black pieces can find ideal square within the blink of an eye. The example diagram shows this is action. Without making any really obvious mistakes, white suddenly has to fight for not losing any material directly. Perhaps you can say that going both Ne4 and c4 is mixing together two different plans, which is typically cause for alarm.

And now?

I think I’m happy with this overview. Next time the 0-0 part, and then I’m done hopefully. It’s a completely different way to look at chess and my brain needs to digest it all. It’s currently doing a skeptical but optimistic ‘hmmmm’......