After 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4, the most popular move (main line) is 4...Bf5. But former world champion Anatoly Karpov seems to like 4...Nd7. Since Karpov's style is "solidly positional, taking no risks but reacting mercilessly to any tiny errors made by his opponents," I am wondering if 4...Nd7 is generally considered more solid and less risky than 4...Bf5, at least by Karpov.
I am hoping that Anatoly Karpov can answer or comment on this question, but it does not seem realistic.
Regarding your question, as far as I remember, both lines hold.
Bf5 gave White a lot of options to fight for an advantage. It got revamped with the help of the computers, Black seeking to generate more active counter-play by castling short instead of long, but I had no courage to try those positions out as they were too wild for my taste.
Nd7 was regarded as more solid, giving White less attacking options, but also requires more enterprising play from Black.
Sadly, user q.undertow is right. There is no safe line in the Caro-Kann.
Allow me to explain why:
Since it's creation, it was considered inferior to French defense, masters arguing that 1...c6 is too slow, Black not fighting energetically enough for the center. This fact gives White enough time to put pressure on Black. This will always be the case with Caro, Black argues that he will be able to withstand the pressure and cash in White's over-extension in the endgame.
In some lines, White plays outrageous moves, almost "all-in"-ing, forcing Black to invest reasonable risk, in order not to lose. These positions are ugly for the eye of the beholder, but that does not matter. What matters is that objectively, the positions are sound and not inferior for Black.
Before the endgame, comes middlegame. Relinquishing initiative to White, Black will have to defend some serious threats and nobody in the old days knew just how far can White go with his aggression. They tried to play positionally, relying on general principles, but this approach gave only a small plus for White, so he turned to "all-in"-ing, backed by computer analysis. These new moves broke every principle they could, but still remained viable, forcing Black to reply in equal manner in order not to get crushed, generating mind-boggling positions.
Sadly, it is White who steers the game, so if he wants to, he can drag you into sharp positions. Positional lines hold for Black (last time I checked) so you should have no problems there as their plans are not difficult to grasp.
If you do get dragged in the complications, book yourself up, you won't be able to save yourself with calculation and general principles. From my experience, it was harder to play as Black, more stress, easier to make a mistake, less options to generate counter-play...
IMO it is an overstatement to say that Karpov took no risks. Just to give one example which is curiously on topic, in one of his Caro games he played an early .. Ke8-e7 !? in order to connect his Rh8 to his Queen, and got an advantage that way. I would call his style "positional dynamic" or something like that.
Perhaps the larger point is that in chess you simply can't win without taking at least some minimal risk. And I'm hoping that winning (sometimes) is your goal, rather than always drawing.
And there's a risk to systems like .. Bf5 too, though of a different kind: that of being stuck in a minimally worse position with no counterplay. Playing that sort of position against a master is no fun at all.
From Fundamental Chess Openings (FCO) by Paul van der Sterren:
"Theory of this opening system (which sadly lacks any widely accepted name) has been enriched with a huge number of sharp and original variations"
Here he is referring to the Karpov variation (which now has a widely accepted name!).
So it's definitely not slow and solid necessarily- white can drag you into an attack still.
If you want a slow and solid variation, may I recommend the following variation: 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nf6 5. Nxf6 exf6
FCO says of this: "It is solid and fairly simple strategically, but it must be admitted that Black is making a positional sacrifice of some sort. In exchange for obtaining open files, free development for all his pieces and a safe place for his king (supposing that Black castles kingside) Black accepts a doubling of pawns which may eventually boil down to a virtual pawn sacrifice. For if White manages to obtain a passed pawn on the queenside, Black will have nothing to show for it on the kingside. This line is clearly a matter of taste"
I play this from the black side regularly and agree with the assessment above.
