HgMan wrote on 09/29/17 at 23:35:07:
This is well worth exploring. My lone problem with 4...Nd7 is that Black needs to prepare for a series of fifth moves, all of which tend to take the game in wildly different directions. This isn't the end of the world, of course—and maybe I'm answering my own original question—but Black does put the ball back in White's court to dictate the course of subsequent play in a manner that isn't consistent with other lines of the Caro-Kann. When I played this in correspondence chess, my opponents were pretty evenly split between 5.Ng5, 5.Bc4, and 5.Nf3. If I could be sure they'd play 5.Ng5, then I'd happily delve into this more carefully.
Sorry. Back to 5...Ndf6 (which I never played: stuck to 5...Ngf6 and Khalifman's Karpov repertoire).
I played the Smyslov Caro-Kann over the board for some years, almost exclusively, and I had the same experience in that you encounter all three white replies almost evenly. Black has at least two reasonable options against each of the three "big" replies, but you cannot really force white to do anything special if he doesn't want to. After a while the three lines felt pretty much the same to me. I could vary a bit as black and my biggest problem was that equalizing often meant reaching an ending which I did not manage to win as white had no weaknesses. This was fine as long as my main objective was grinding out draws against stronger opponents. I will elaborate a little.
- Against 5. Nf3, there is 5. ... Ndf6 (even though this is never played) and of course 5. ... Ngf6, and after 5. ... Ngf6 6. Ng3 there is 6. ... e6 as well as the immediate 6. ... c5. After. 6. ... e6 Black still has the choice between rather calm development and a quick ... c5 for several moves. White can often choose between c2-c3 and c2-c4, indeed dictating the course of events.
- Against 5. Bc4, there is again 5. ... Ndf6 or 5. ... Ngf6, or even 5. ... Nb6 aiming for ... Nd5. After 5. ... Ndf6 6. Ng5 e6 7. Qe2 Nb6 both of white's choices are rather dull for black.
- 5. Ng5 has already been discussed.
In all these cases the resulting positions resemble nothing else in the Caro-Kann (unlike the 4. ...Bf5 variation), but then again, they make up a type of position of their own. The endgames tend to be quite similar and there seems to be only a handful of critical positions.
What annoyed me most is that it takes quite long to equalize, often resulting in endgames where white still has no structural weaknesses. I had this especially against 5. Nf3 and 5. Bc4. White doesn't need to risk much if he doesn't want to. There's not much room for tactical oversights on white's side and after exchanging the last attacking piece there is still no hook in white's position. I guess for the real Smyslov/Karpov aficionado, this is where the game starts.