I'm trying out the Tartakower variation of the Caro-Kann. It begins

[fen ""]
[startply "11"]
[startflipped "0"]

1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nf6 5. Nxf6+ exf6 6. Bc4 Bf5 (6... Bd6)

I have not had much success with it thus far. I lost a recent LiChess game seemingly without much effort from my opponent. It continued from the above 6. Bc4 Bf5. It seems for black 6...Bf5 is inferior to 6...Bd6, being played in far fewer master games (and judging from Stockfish's evaluation: +0.6 vs. +0.3), although I'm not sure if this is correct, and I'm not sure why.

Question: Is 6...Bf5 inferior to 6...Bd6 in the Caro-Kann Tartakower after 6. Bc4? And if so, why?

5 Answers 5


6...Bf5 certainly doesn't feel right to me. I would think Black wants to get ready to castle king-side quickly and 6...Bd6 allows that.

Additionally Bd6 prevents White from playing Bf4 which is quite active compared to the alternatives.

By-the-way, I think this line is risky for Black as you already have the potential for a lost endgame due to the pawn structure.

  • I think you are slightly overestimating the pawn structure when you call this line risky. True, black would be in danger in a king and pawn ending. But black doesn't have to allow this to enter a king and pawn endgame. There are obvious advantages to this structure, which are that both of black's bishops are active and have free diagonals, the knight has squares to go and there are no weaknesses.
    – ktm5124
    Mar 20, 2019 at 19:14
  • @ktm5124 White enjoys the same but with no crippling of his structure. In other openings like the Exchange Ruy Lopez where such pawn structures occur Black gets the bishop pair advantage in compensation; here White is "free-rolling". Of course the endgame is far off as you pointed out.
    – Ywapom
    Mar 20, 2019 at 19:25
  • I wouldn't use the word "crippling" though; that's too dramatic. Black's structure is not at all crippled. White's queenside pawns could also come under attack, if black ever gets a knight on c4. Eventually black might play c5 and white's best response is to allow the exchange or initiate the exchange, to get a 3-on-2 in the queenside. But it's possible the c-pawn will eventually be vulnerable. This is all longer-term, of course, after black castles. The nice thing about the f6 pawn is that together with c6 black controls d5 and e5. His knight may be stronger than white's.
    – ktm5124
    Mar 20, 2019 at 20:13
  • I also wouldn't be too quick to compare this to an Exchange Ruy Lopez, because I think that's solely fixating on a single aspect of the position (i.e. black has doubled pawns and he will lose a king and pawn ending). In the Exchange Ruy, black's pieces are often a lot less active than they are here, especially if black has pawns on e5 and f6 restricting the dark-squared bishop.
    – ktm5124
    Mar 20, 2019 at 20:16

It is a bit early in the game to be saying that one move is decisively better than another, but there are general principles that can be applied.

One of these is to ask whether you are already sure where certain pieces belong, and if you are sure, put them there at once unless there is some tactical reason not to. In this variation, where Black recaptures on f6 with the e-pawn, he has already conceded a slightly disadvantageous endgame because he cannot create a passed pawn on the K_side, but White can do so on the Q-side. As compensation, Black has achieved mobility for all of his pieces, and can try to use this to bring off an attack with ....Bd6, 0-0, Qc7 and Bg4 that makes any endgame irrelevant.

So Black is fairly sure that his KB belongs on d6. His QB could belong on g4, f5, perhaps e6, or it might stay where it is while Black plays Re8 and Nd7-f8. He might even consider ..f6-f5. So ..Bd6 has the advantage of preserving options for the QB. There is less value in maintaining options for the KB because we already know where we want it.

The possibility of 6...Be6 is dubious. After 7. Bxe6 (or first 7.Qe2) the Pawn on e6 will be a permanent weakness even it gets to e5.


6...Bf5 looks like a good developing move. It fights for control of the e4 square and puts pressure on the c2 pawn. However neither side is trying to push a pawn to e4 because there are no e pawns left and the c2 pressure is pretty meaningless.

Meanwhile white's 6. Bc4 fought for control of the d5 square and now 7. d5 is a possibility although probably not yet desirable. 6...Bf5 does nothing to address this and basically lets White have control.

6...Bd6, on the other hand, directly contests the d5 square. Furthermore it leaves white with an immediate choice, either acquiesce to the exchange of bishops thereby unravelling black's pawn structure with near equality, or move the bishop again.

6...Bd6 is clearly superior.

  • 6
    6 ...Bd6 does NOT contest the d5 square.
    – CWallach
    Mar 17, 2019 at 7:20
  • 1
    What exchange of bishops? Are you talking about Be6? The question was about Bd6.
    – bof
    Mar 18, 2019 at 6:49
  • @bof You're right but the underlying principle remains. The light squared bishop does not belong on f5. Either e6 or, g4 if the white knight goes to f3 rather than e2 , would be better. Playing Bd6 first to allow speedy castling is probably best.
    – Brian Towers
    Mar 18, 2019 at 10:22
  • I disagree with this evaluation. The idea of playing Be6 isn't as good as you make it out to be as it creates a weakness on e6 after Bxe6 fxe6. White can attack a pawn on e6 with the open e-file. There isn't a desperate need to "unravel black's pawn structure" as the four kingside pawns aren't weak and provide good protection for the king. The only thing to really worry about in this pawn structure is a king and pawn ending where white can create a passer on the Qside, but black doesn't have to allow that situation; black can keep pieces on the board.
    – ktm5124
    Mar 20, 2019 at 18:50

I would play Bd6 first, because it's more flexible.

First, it allows you to castle more quickly. Second, you know that you want your bishop on d6, but you don't know exactly where you want your light-squared bishop. Perhaps it belongs on g4, pinning the knight. Third, if you do want to put your bishop on f5, it might be worthwhile to first prepare this with h6, so that you can meet Nh4 with Bh7. The pawn formation on f7, g7, f6 and g6 didn't work out too well for you in the game, as your king didn't really have a good square; it was vulnerable on g8 to the light-squared bishop, it would have been vulnerable on h8 to checks along the h-file and it would have been vulnerable on f8 to a check on h8.

I think f5 is a fine square for the bishop, and if the knight moves to f3 then g4 would also be a good square. I personally would prefer g4 in that case since it seems more active to me; it pins the knight whereas a bishop on f5 has no target. But I don't think it's a mistake to eventually place the bishop on f5. It's just that the timing is important. You might have to prepare it with h6 to meet the threat of Nh4. The one square I don't like putting the bishop on is e6, since after Bxe6 fxe6, the e6 pawn becomes a weakness. It can easily be attacked along the e-file, and by a queen on b3 or g4. And if you ever play e5, then either you end up with an isolated pawn on e5 (after dxe5 fxe5) or you end up with pawns on dark squares, which is the wrong color to place your pawns when you have a dark-squared bishop.

I think the real question is not whether you play Bd6 or Bf5 first, but how you meet the threat of Nh4 when you have a bishop on f5. That seems to me the reason you lost the game. You had no good way to meet this threat and that gave white a lot of initiative.


Not directly answering the question but 6...Qe7+ followed by Be6 is considered to be an equaliser against Bc4 (idea is if Be3 then Qb4+ wins and otherwise the bishop on c4 has to meekly retreat or Qe2 is played with lots of trades on e6 and equality)

If just between ...Bd6 and ...Bf5: 6...Bf5 is also little bit off because you are going to castle kingside and are still two moves away from doing so. Two moves is actually a lot more worrying than one move away from castling (if your opponent does something scary when you are one move away, you just castle).

Also as @Philip Roe says it is a bit committal - it's not clear if the bishop belongs on e6, f5, g4 or what yet. In general you want to play the moves you know you have to do first - i.e. castling.

Hopefully this explains why Bd6 is better than Bf5

@Ywapom ...exf6 C-K is only slightly risky IMO - this line with ...h5 against the Qc2+Bd3 +Ne2 setup (1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nf6 5. Nxf6+ exf6 6. c3 Bd6 7. Bd3 O-O 8. Qc2 Re8+ 9. Ne2 h5) has totally revitalised it (big plus score for black after 9...h5 although I would take that with a pinch of salt)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.