I play the Caro–Kann Defence against 1.e4 as black, and for whatever reason it feels like I get the following line way too frequently:

[FEN ""]

1.e4 c6 2.d4 d5 3. exd5 cxd5 4. Nc3?!

This line is a "non-critical" line, so much so that I just got an opening book on the Caro-Kann and there are a whopping 0 lines that discuss how to play against this setup. It's very rarely played in the masters databases, but on lichess.org it's the 4th most common move in this position, and you can also reach the same position by playing 1.e4 c6 2.Nc3 d5 3.exd5?! cxd5 4.d4 which isn't too uncommon.

Similar to the c3 knight in the Richter–Veresov Attack or the "Jobava" London, mostly I just see people dismissing this knight, claiming off-hand that it's bad, that it "doesn't belong there". But for whatever reason, when I look at my results in the different lines in the Exchange Variation, this is my worst line! I'm probably running into problems with a small sample size, but I'm wondering if anyone has any general tips for how to make white feel this "bad knight".

  • Why does 2.Nc3 deserve a ?! in the alternative move orden when it can easily transpose into the mainline? – David Nov 3 '20 at 12:24
  • 2
    @David Yes 2.Nc3 isn't dubious since White can play 3.d4. The dubious move in this line would be 3.exd5. – Inertial Ignorance Nov 4 '20 at 0:54
  • I think the better question is: why isn't this discussed so much in books. For which I suspect the answer will be that it's a very "boring" opening which means just developing and then looking for an opening. – paul23 Nov 4 '20 at 13:01

I'm wondering if anyone has any general tips for how to make white feel this "bad knight"

The fact that you manage to put "bad" and "bad knight" in scare quotes shows that there is some hope for you. White does not have a bad position. I suspect that this line isn't played much at master level because it isn't the best way to squeeze every last scintilla of advantage out of white's starting advantage of having the first move.

But there is no universe in which you can look at white's position and seriously think that white has a bad position which black must punish. If you think like that as black then you are on the way to losing.

Four moves into the game and white has already -

  • Staked a claim to the center with a pawn on d4 (black matches this with a pawn on d5)
  • Developed a piece to a good central square, the knight on c3 (black is behind but it is black's turn so can remedy this on the next move)
  • Has free development for both bishops (black has free development for only one bishop)

White is not worse. Black is not better. Has black completely equalized? It doesn't look like it.

So, your best approach is to accept that white's first move advantage may have decreased slightly but your job remains the same during the opening - try and equalize via good development. How you develop your pieces will depend to a certain extent on how white develops. Pay attention to all the things you normally do during the opening and aim to reach an even middlegame where you can try and outplay your opponent.

At all costs avoid trying to punish your opponent. This will only lead to bad moves and an early loss.

  • 10
    With the c pawn blocked, the normal ideas in the exchange caro (bishop on d3 protected by c3 from any Nb4 ideas, stopping black's bishop from getting outside the pawn chain with Qb3 ideas, etc) are mostly all invalidated. I think your answer would benefit from digging deeper into these specifics. – NoseKnowsAll Nov 2 '20 at 21:24

The knight is badly misplaced on c3 here. If you protect your d5-pawn with ...e6, what is the knight supposed to do? It doesn't put any pressure on the d5-pawn since there are no White pawns helping to attack it. If White's c-pawn were on c4 then the knight would be completely fine on c3. But now, even if all of White's pieces were attacking the d5-pawn, it wouldn't really matter since there's no White pawn involved in the attack. Capturing on d5 would entail sacrificing a piece for two pawns.

On c3 the knight also can't move to e4, which is firmly controlled by the d5-pawn. The best plan for it is to maneuver somewhere else, such as to the g3-square via e2. But now White will have spent 3 tempi in total just to put the knight on a mediocre square. Also, as NoseKnowsAll mentioned, with the c-pawn being unable to move to c3, Bd3 is in danger of running into ...Nb4; and, due to the c2-pawn, Qb3 ideas aren't possible (making it safer for you to develop your light-squared bishop).

There's no quick way to punish White's play, so just develop for now. E.g., ...Nf6, ...Bf5/...Bg4, ...e6, ...Be7/...Bd6, ...0-0, ...Nc6. Then, you could consider starting a queenside attack with ...a6 + ...b5. Maybe later on White will move the c3-knight away and push the c-pawn to c3, so then you'd be in a position to launch a minority attack with ...b4.

Another plan is to instead develop your king's knight to e7 (after having played ...Bd6), and aim to play ...f6 + ...e5 later in the game. Because White's c-pawn isn't helping to pressure the d5-pawn, this plan has a higher chance of succeeding than it normally would.

  • I think white has a little more bite in this position. I think the main idea of Nc3 is a combination of Nc3 and Bf4 and threat of attacking c7 square. For example, 1.e4 c6 2.d4 d5 3. exd5 cxd5 3. Nc3 Nf6 4. Bf4, now black should play 4...a6, and this creates two potential squares for the knight b6 and c5. For example, 5.Nf3 Nc6 6.Ne5 e6 7. Nxc6 bxc6 8. Na4 – Akavall Nov 4 '20 at 3:59

Just what is so bad about the White knight and its position? 4. Nf6 for Black is the best option, with a slight theoretical edge in scoring probability in the endgames, although getting there would take accurate play. Additionally, there are many moves that look plausible that would fail to do well. Overall, the position is very drawish.


My answer is to just play chess- playing on the queenside as normal is a good idea in this structure even without the pawn on c3, since here you get to knock away the knight. Pressure versus d4 is also easier without a c3 pawn, which is nice. Black is better than normal, but it's not really anything you can get a instant win from- this just doesn't even come close to proving an advantage.

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.