Beating the Caro-Kann

[Kevin Pacey]

The Caro-Kann Defence can give White considerable trouble even at the elite player level, and a lot of class players find it annoying, if only since it can often result in rather dull positions that are also solid for Black, in a number of the theoretically approved main lines (which also can require a lot of understanding and/or memory work). How to cut down the workload for the average player who doesn't want to give up playing 1.e4, especially over an opening he faces far less typically than the Sicilian or 1...e5? I'll offer some suggestions of my own, using the PGN viewer capability of this forum as an aid:


[Event "It"]
[Site "Leningrad, USSR"]
[Date "1934"]
[Round ""]
[White "Botvinnik, M"]
[Black "Kmoch, H"]
[Result "1-0"]

1.e4 c6 2.d4 {White can cut down on the theory a bit already here, e.g. with 2.Nc3 and if 2...d5 3.Nf3 [3.Qf3 is an offbeat move I've also played, which may not be totally harmless in practice, and it's certainly not as weak as some books make it out to be]; however 2.d3 d5 3.Nd2 e5 4.Ngf3 Bd6 gives Black equality with a full share of the centre, which is one reason why diehard King's Indian Attack players might prefer 1.Nf3 to 1.e4. Otherwise, I've had much more trouble [than vs. the French] in coming up with quick and dirty ideas for White that have any sort of potency, as the Caro-Kann is a solid sort of animal indeed.} d5 3.exd5 {Alternatively 3.e5 is one of the main systems, and it may lead to closed positions of the sort that class players may not wish to go into when playing 1.e4; Besides that the Classical, beginning with 3.Nc3 [or first 3.Nd2] 3...dxe4 4.Nxe4, is arguably the main line of the Caro-Kann still, with all its heavy theory and/or solidity, and Black will probably not mind seeing it.} cxd5 4.c4 {The Panov [actually reached in this game via 1.c4 c6 2.e4 d5 etc.], a major system of the Caro-Kann that can lead to an open and sharp game, though there's still some theory to absorb. At least White usually can afford a misstep better than Black can. Alternatively, a way to really cut down on the theory is 4.Bd3, and after 4...Nc6 5.c3 Nf6 [if 5...g6 6.Nf3 Nh6 7.O-O Bf5 then 8.Ne5 or 8.Re1 could be tried] play 6.Bg5 Bg4 7.Ne2 e6 8.Qc2 as in Topalov-Ivanchuk, Linares 1994 [instead of playing the now clearly de-fanged 6.Bf4 main lines of the Exchange Variation that began at move 4].} Nf6 5.Nc3 Nc6 {White should also prepare something against the less spirited 5...e6 and 5...g6 lines. A bit of additional work, but at least White can be rewarded with his share of fun by playing the Panov.} 6.Bg5 {6.Nf3 Bg4 7.cxd5 Nxd5 8.Qb3 Bxf3 9.gxf3 can lead to an endgame, if White is into such; there are only a couple of ways for either side to sidestep that, and they're not looking so convincing.} e6 {6...dxc4 is a big alternative here that White should study too.} 7.c5 {7.Nf3 is a major alternative. However, it might transpose.} Be7 {7...Bd7!? guards c6, and so aims to get in ...b6. Then 8.a3 may be best.} 8.Bb5 O-O 9.Nf3 Ne4 10.Bxe7 Nxe7 {Also possible was taking with the queen.} 11.Rc1 Ng6 12.O-O Bd7 13.Bd3 {Trading bishops was also fine.} f5 14.b4 Be8 {Black's position is slowly becoming critical, as is often the case when he doesn't get in ...b6 or ...e5 after c4-c5 is played in the Panov. Instead it seems he should attempt to slow down White on the queenside with 14...a6.} 15.g3 Rc8 16.Re1 Qf6 17.a3 Ne7 18.Ne5 Qh6 19.f3 Nf2 20.Qe2 {Not 20.Kxf2 Qxh2+ 21.Ke3 f4+ 22.gxf4 Qxf4+ 23.Ke2 Qh2+ with a draw by perpetual check.} Nh3+ 21.Kg2 g5 22.Nb5 Bxb5 23.Bxb5 Rf6 24.Bd7 Rd8 25.b5 Qh5 26.c6 Rh6 {Now 27.c7 Nf4+ 28.Kh1 Rxd7 29.gxf4 is one way to win, but Botvinnik chooses a more cautious path.} 27.Kh1 1-0

[Kevin Pacey]

Here's the wiki on the Caro-Kann:

https://en.wikipedia.org/wiki/Caro%E2%80%93Kann_Defence

[Javier Gil]

Hmm. I'd like to say that most 1.e4 players do not particularly enjoy playing QG type positions, which is what a Panov attack is.
5...e6! (rather than 5...Nc6, which is a little dubious) is actually the main line, regularly chosen by the strongest GMs in the world. The game will often transpose into a Nimzo-Indian. I'd call the Panov a "high risk variation" for White...

[Kevin Pacey]

Unclear to me what is the mainline of the Panov, but GM Schandorff recommends 5...Nc6 in his GM Repertoire (Quality Chess) book on the Caro. He quotes a large number of top level games that have occurred in the well-known endgame line that has arisen from it over the years, such as Karpov-Kramnik, Linares 1993. ECO, 4th edition (circa 2004) has 5...e6 is their mainline, it would seem, but it's no longer such a distinguished authority as in the past.

Top Canadian GM Sambuev loves to use the Panov to play attacking chess, because of the isolated d-pawn positions that can often arise in other cases. Most 1.e4 players I play locally dislike even facing the Caro-Kann, as it is generally very solid (and perhaps dull, more to the point) in most of the main variations other than the Panov, if Black wishes.

[Javier Gil]

Unclear?? any book on the Caro-Kann or the Panov will tell you that. Also, if you check your database, 5...e6 is played at least twice as often as 5...Nc6.
5...e6 has been played by Karpov (dozens of times!), Kamsky, Anand, Ivanchuk, Morozevic, Svidler, Nakamura, Bareev, Jakovenko, Ponomariov, Dreev, Nissipeanu, Akopian, to name a few. By the way, Kramnik is an expert on both sides of that postiion and has played the lines which result from the transposition to the Nimzo, with ...e6, dozens of times.
I agree that 5...Nc6 can lead to very interesting games, but 5...e6! is defiitely the most solid and sound move in that position.
To call the move 5...e6 "unspirited" is bizarre.

[Kevin Pacey]

Wells' book GM Secrets on the Caro called the move 5...Nc6 'quite a spirited defence', as should be apparent (barring the Endgame Variation, that might be considered a mainline), which I was admittedly extrapolating by when it comes to 5...e6 (which Wells calls 'solid'). All my books have respect for either move. Wells says 5...Nc6 may be sounder, while IM Jouska says it may be riskier, even though she recommends it. Schandorff is a serious writer, and 5...Nc6 is his recommendation for Black too. To me, either move is worthy of mainline status, and who says there has to be just one mainline (well, Karpov's old co-authored book is the only one suggesting that 5...e6 is the one and only 5th move mainline, of the many books that I have)? Fwiw, as Black (and merely being a national master) I play 5...g6, and that's definitely not a mainline (though it is spirited), IMHO. In any case I could throw out names of many elite players who have played 5...Nc6 too, some of them the same ones who use 5...e6 at least now & then, e.g. (not counting the earlier mentioned Kramnik, or Ponomarionov, Euwe, and G. Kasparov): Bareev, Dreev, Leko, Adams...

Individual (even large) databases can have different number of games for any given variation. GM Gallagher in his old Starting Out book for the Caro at least quotes stats that show almost twice as many 5...Nc6 games being played as for 5...e6, 'with the main line of' ...Be7 played the next move, for example. What also counts is the latest trend at the top for e.g. the past decade or two.

I hope you're not trolling, as that's not taken too kindly on this discussion board by the moderator. Otherwise, perhaps a calmer approach will help any further discussion.

[Javier Gil]

Wells' book GM Secrets on the Caro called the move 5...Nc6 'quite a spirited defence', as should be apparent (barring the Endgame Variation, that might be considered a mainline), which I was admittedly extrapolating by when it comes to 5...e6 (which Wells calls 'solid'). All my books have respect for either move. Wells says 5...Nc6 may be sounder, while IM Jouska says it may be riskier, even though she recommends it. Schandorff is a serious writer, and 5...Nc6 is his recommendation for Black too. To me, either move is worthy of mainline status, and who says there has to be just one mainline (well, Karpov's old co-authored book is the only one suggesting that 5...e6 is the one and only 5th move mainline, of the many books that I have)? Fwiw, as Black (and merely being a national master) I play 5...g6, and that's definitely not a mainline (though it is spirited), IMHO. In any case I could throw out names of many elite players who have played 5...Nc6 too, some of them the same ones who use 5...e6 at least now & then, e.g. (not counting the earlier mentioned Kramnik, or Ponomarionov, Euwe, and G. Kasparov): Bareev, Dreev, Leko, Adams...

Yes, Karpov calls it the main line. I think his opinion is extremely qualified... not only was he World Champion when he played those games, no other living GM has ever defended the black side of a Panov in OTB games more often than he has. Dozens and dozens of games against the best players in the world.
If the fact that 5...e6 is played more than twice as often as 5...Nc6 doesn't make it a main line, it baffles me what criteria we are to use when calling a variation the main line.
As for books, here are a few more:

Gallagher, "Starting out The Caro-Kann"
1) He can defend his centre with the solid 5 ... e6. This is Black's most popular way of meeting the Panov-Botvinnik and is the line that often leads to the classic IQP position.
2) He can rapidly develop his queenside with ... Nc6 and ... Bg4 (or ... Be6). This would be quite risky if it wasn't for the fact that it is backed up by a lot of theory.

IM Edmar Mednis, "Understanding the Caro-Kann defence":

5...e6! By far Black's soundest procedure in the Panov. The critical d5-point is now sufficiently protected to allow Black to complete his Kingside development easily with . . . Be7 (...Bb4!) and . . . 0-0.
If this is too tame for you, there are two other important variations you may consider. You should know, however, that these other variations require much greater technical and tactical mastery than the main line and that Black also runs the risk of landing quite suddenly in an unfavorable position.

Jacob Aagard: "Easy guide to the Panov"
Those who do not wish to defend the endgame (and trust me, it's a painful task!) in the variation 1.e4 c6 2.d4 d5 3.exd cxd 4.c4 Nf6 5.Nc3 Nc6 6.cxd Nxd5 7.Nf3 Bg4 8.Qb3 Bxf3 9.gxf e6 10.Qxb7 Nxd4 11.Bb5 Nxb5 12.Qc6 Ke7 13.Qb5, often avoid it as follows:
1.e4 c6 2.d4 d5 3.exd cxd 4.c4 Nf6 5.Nc3 Nc6 6.cxd Nxd5 7.Nf3 Bg4 8.Qb3 Bxf3 9.gx Nb6 However, White should have the better chances in this middlegame due to his bishop pair.


I hope you're not trolling, as that's not taken too kindly on this discussion board by the moderator. Otherwise, perhaps a calmer approach will help any further discussion.

I beg your pardon??
I do not agree with your views or comments on this variation, I have a completely different opinion about it. I'm sorry that you regard this as trolling and that you need to get the admins to review my posts because I simply disagree with you.
I have not attacked the person, but the ideas, and I have backed them with evidence from both books and GM practice. I just think people who read your post could be missled about this variation, and they deserve to read a different view so that they can make up their own minds.
5...e6 can lead to one of the most fascinating type of positions (IQP) that the chess world has seen over the last 50 years, positions which GMs still don't agree on despite the use of computers. To compare that to the boring endgame that you can get in the 5...Nc6 variation and still call 5...e6 "unspirited" is, to my eyes, both wrong and utterly unfair.

And then you say " against the less spirited 5...e6 and 5...g6 lines. A bit of additional work, but at least White can be rewarded with his share of fun by playing the Panov.". "a bit of additional work"?? Really?? It has taken Masters years to understand those positions and play them well!
Kevin, I know your enthusiasm and contributions to this community have been nothing short of fantastic, I'm not taking away any of the credit that you have so deservedly earned, but in this particular case, you're portraying an unrealistic picture of how hard it is to master those variations.

[Kevin Pacey]

Re-read what I said about 5...g6. Also, the Panov I never said would be entirely light on the homework, if a player is conscientious (personally as a relatively low rated player I try to study at minimum, i.e. enough to know the basic main continuation plus pitfalls and themes needed to at least survive in a given opening - IM N. Noritsyn has commented much the same elsewhere long ago). Most popular (by how much?) does not necessarily mean something's the only 'mainline' (especially these days, at the top - plus many databases include lots of games of low rated players pitted against each other). The books you quoted have old opinions of 5...Nc6. The Endgame Variation is recommended by Schandorff (2010 admittedly, but not as old as the books you quote) - he quotes quite a number of high level games from it. Otherwise, I didn't like the strong tone (and hasty generalities) of some of your earlier remarks. Now, your tone is even stronger. When you're not 100% sure you're seeing all the main points of the other side of an issue, ask questions, and gently, for God's sake. I wrote I personally was unclear about something, and it set you off because to you it seemed I disagreed with you, which you couldn't tolerate in the least, if your reply is to be taken at face value. Let's get that straight. What's worse, the unrelated issues you brought up kept spiralling from there. The signs are now most certainly there that you are acting as a troll (the heading [plus its punctuation] of your post is one giveaway), but I replied against my better judgement to your last post, if only to bring up points that might interest or inform others. You are going on my Ignore list. I hope you are banned as quickly as possible.

https://en.wikipedia.org/wiki/Internet_troll

[Javier Gil]

Yes, some of those books are old, but it's not like they publish new books on the Caro or the Panov every year. I think the last one was Schandorff's, and whilst it probably is a great piece of work, Schandorff has only played against the Panov twice in his whole life, and he chose 5...e6 in the first one, dating back to 2002, and in 2016 5...Nc6 in the second, winning both). (according to chessbase's 2018 MegaDatabase + Twic),
Nikolai Noritsen has never played against a Panov in his whole life. (same sources)

I'm an International Master and have played the Caro-Kann all my life, spending hundreds of hours on the Panov, so of course I'm going to defend the 5...e6! variation vehemently, because like other GMs and IMs, I believe it's the most reliable move, particularly if you're playing for a win.

I have only debated the argument, never attacked the person, whom I respect. Unfortunately, I think you have both attacked my tone and called me names (Troll), which is sad. :(

"argumentum ad hominem", is a fallacious argumentative strategy whereby genuine discussion of the topic at hand is avoided by instead attacking the character.

392

https://en.wikipedia.org/wiki/Ad_hominem

[Kevin Pacey]

I happened to see a friendship request, and guessed it was you, so I was curious about your post here afterwards. Maybe I'm no longer used to seemingly hot conversation on message boards, such as often happens on chesstalk (Canada's de facto main message board, still). I'd read the wiki link I last gave before posting it, and thought many of the main description points matched up, but I suppose coincidences can happen even in bunches when it comes to style, especially as you didn't initially reveal you wanted a big debate, moving from one point to another and not sometimes pausing to acknowledge when the other fellow may have any sort of a point even if minor.

Debating, if you'd call it that, forever would seem to have been almost always mild, short-lived and orderly on this message board (e.g. not much stuff off the main topic of a thread seems preferred), except maybe at times in the Canadian chess politician's meetings forums when they're happening. This message board has for ages not been very active, in that few people post in any forum for long periods, especially between political meetings. A problem is I don't fully trust that other message boards are not seriously infected at times, as I don't have great anti-virus stuff. I also don't much trust the internet in general.

I'm an aging guy without much stamina, e.g. for bending over a machine for extra hours on end for days in a row, or for feeling agitated for long periods, justifiably or not. Often brain fog and brain lock, perhaps due to medication, is a handicap when playing chess at my club, when I still do, though strangely a single beer before a game doesn't always seem to hurt my play much. Currently it's more chess variants that I play, by email correspondence, on chessvariants.com, where inventing and commenting on chess variants are possible things I've done too (the website's database of chess variants alone is huge). I was into Canadian chess politics for a while, but eventually all my ideas dried up or were left to rot, possibly rightly in a lot of cases. Things are slowly improving for our chess federation nowadays, after a low point about 10 years ago.

Jouska has two books on the Caro, the last one 2015, I think. There's also a 2010 Dangerous Weapons series book on the Caro, in which strangely there are no chapters concerning 5...e6, but rather three on 5...Nc6, one rooting for the Black side of a line with an early ...g6, so in a way I was being generous to say all the books I have show respect for both 5...e6 and 5...Nc6, as some of the newer ones don't even mention 5...e6 (though clearly it's a reliable, solid, very good, and very natural move, which does go without saying). Noritsen (sp?) I was writing of in terms of his general approach to studying openings, without knowing his actual repertoire, which I may not have made clear enough.

I apologize that I misread the situation and overreacted badly. Apparently I am getting old and cranky, which is sad. I hope you might even start threads of your own for a while. I tried to liven things up here more a while back by posting many threads in a relatively short time, when I was feeling a lot better physically than lately, but few of those threads were responded to. Maybe you'll have better luck, if that's ever a goal for you. Whenever I check relatively few people seem to visit most of the threads I'd begun this year, so I often may not be too interesting a poster for the average reader.

[Javier Gil]

Thanks for the explanations Kevin, it's all good. I guess I can a be a little too stubborn at times :(, so I take my share of the blame...

One of the reasons why I never took up the ...Nc6 line is that after 6.Nf3, (6.Bg5 being considered as "risky" for both sides, but probably more so for White), black's best move is 6...Bg4, which leads to an endgame where neither side has much, and it's practically impossible to play for a win for Black... or White!.
There's been a recent try to improve the variation for white with an in between check on c5:



[Event "Analysis"]
[Site "Canadian forum"]
[Date "2018.12.12"]
[Round "0"]
[White "Not important"]
[Black "Not importan"]
[Result "?"]
[PlyCount "30"]

1. e4 c6 2. d4 d5 3. exd5 cxd5 4. c4 Nf6 5. Nc3 Nc6 6. Nf3 {This is the most
solid line here, after which ...Bg4 is thought to be best. The main line now
leads to a completely equal position, but there's a recent new idea which is
causing black some problems.} Bg4 7. cxd5 Nxd5 8. Qb3 Bxf3 9. gxf3 e6 (9...
Nxd4 $4 10. Bb5+ {And White wins.}) 10. Qxb7 Nxd4 11. Bb5+ Nxb5 12. Qc6+ (12.
Qxb5+ Qd7 13. Qxd7+ (13. Nxd5 exd5 {Unlike the line with 12...Qc6+, here
Black's King hasn't moved and can therefore still castle, so any move other
than Qxd7, which tranposes into the endgame, is just bad.}) 13... Kxd7 14. Nxd5
exd5) 12... Ke7 13. Qc5+ $5 {This was first tried by GM Grischuk in a rapid
game. The move looks absurd because Black can retreat his Knight to d6.} (13.
Qxb5 Qd7 14. Nxd5+ Qxd5 15. Qxd5 exd5 {This has been played hundreds of times.
Unlike the other line, Black's King is here on e7 rather than d7, but the
position is so equal that it doesn't really make much of a difference.}) 13...
Ke8 (13... Qd6 14. Nxd5+ (14. Qxb5 {This has been tried also.}) 14... exd5 15.
Qxb5 Rb8 16. Qd3 $5 {With an unbalanced position.}) (13... Nd6 14. Nxd5+ exd5 {
And here both 15.Qxd5 or, better, 15.0-0 give White good compensation for the
piece, with Black's King being stuck in the center for a while.}) 14. Qxb5+ Qd7
15. Nxd5 exd5 {And here White can retreat his Queen to b3 or d3. Since Black's
already moved his King, he can no longer castle, which could cause him some
headaches in the long run.} *

[Vladimir Drkulec]

I did not see this conversation as anything remotely approaching something that administrators might intervene in. As an occasional Caro-Kann player (something like seven games with five wins and two draws as black if my memory serves) I found it very interesting. When I played 1.e4 the Panov attack was my preferred way of dealing with the Caro-Kann. At the time my play was full of gambits and it seemed a bargain to play white and get all that play without even giving up a pawn. The state of theory and research was much simpler in the 1970s.

Even though my record with the Caro-Kann would suggest that I should continue to play it, I put it on the shelf for a while pending actually learning some theory. I do occasionally still play it in online blitz with reasonable results. I was miserably lost in almost every one of those tournament games, even the ones I won mainly because my entire foray into understanding the opening was about one hour of training. I even played it against a life time Caro-Kann player in hopes of learning the line he feared most. It seemed a very realistic fear based on the game we played though he lost his way in the complications and I eventually won from a lost position.

I have a much better theoretical knowledge of 1.e4 e5 or 1.e4 c5 or even the Scandinavian or French defense as openings that I have played and studied more extensively.

[Kevin Pacey]

Grischuk's move I was unaware of. As Black, if I ever were to play 5...Nc6, the Endgame Variation (or now Grischuk's refinement) is something I would possibly side-step, as it does not suit my style, at least at the moment. There is the apparently risky 9...Nb6!?, if home analysis with an engine works out in a certain variation (but my engines are older ones now), and there is the Dangerous Weapons book's idea of 6...g6 (again after 6.Nf3), though the author of that chapter admits it's not exactly perfect - it looks to me that White might gain a slight edge in case(s).

As an aside, it seems I may have taken a quote from the Well's book out of context earlier regarding 5...Nc6 (i.e. that it may have been sounder than 5...e6), as there was some possible ambiguity involved (i.e. the author likely actually meant to compare 5...Nc6 to 5...g6 instead). Improperly checked things happen in the course of internet conversations, at times. It's hard for me personally to tell if modern books recommending 5...Nc6 (not at all implying 5...e6 is worse) is simply a reflection of a short-lived fad, without seeing if a considerable number of GMs are still playing 5...Nc6 these days at the top level. I don't have much in the way of 2018 databases. Otherwise, I have a fairly modern book on the Panov by D'Costa (2013) where so far I've found no clue about the popularity of any of the major continuations.

[Kevin Pacey]

I did not see this conversation as anything remotely approaching something that administrators might intervene in. As an occasional Caro-Kann player (something like seven games with five wins and two draws as black if my memory serves) I found it very interesting. When I played 1.e4 the Panov attack was my preferred way of dealing with the Caro-Kann. At the time my play was full of gambits and it seemed a bargain to play white and get all that play without even giving up a pawn. The state of theory and research was much simpler in the 1970s.

Even though my record with the Caro-Kann would suggest that I should continue to play it, I put it on the shelf for a while pending actually learning some theory. I do occasionally still play it in online blitz with reasonable results. I was miserably lost in almost every one of those tournament games, even the ones I won mainly because my entire foray into understanding the opening was about one hour of training. I even played it against a life time Caro-Kann player in hopes of learning the line he feared most. It seemed a very realistic fear based on the game we played though he lost his way in the complications and I eventually won from a lost position.

I have a much better theoretical knowledge of 1.e4 e5 or 1.e4 c5 or even the Scandinavian or French defense as openings that I have played and studied more extensively.
Nowadays I have a 'repertoire' with both sides that's wide, even when I'm not occasionally deviating from it with previouly abandoned or suspect stuff. I don't study nearly as much as I think I should (even minimally), nor always keep up with the very latest stuff. Otherwise you might say I have 6 basic choices around move one with both colours, regardless of their move, assuming a mainline opening is played. In the case of 1.e4 c5 or 1...e5 I have two choices with each, against Open Sicilians and Ruy Lopez, so you might say I have 8 regular defences against 1.e4, though in the past I've played 1...e6 (intending the Winawer almost always) about a third of the time. It's part of the core that I know best, and often save for the toughest opponents.

When it comes to the Caro I often avoid the most ultra-solid stuff as Black, though there are exceptions. Against the Classical (3.Nc3) I lean on 4...Bf5 (often preferring to castle long, but still hope for some activity, in a way) or else I play the more suspect Bronstein-Larsen after 4...Nf6. I don't know if 4...Nd7, which I've never played yet, is so much in vogue as 4...Bf5, but they seem to be both mainlines. The Advance so far I've played 3...Bf5 against, but would sidestep any attempt to get into the wildest lines of the Shirov line, where there's way too much theory to be absorbed just to survive as Black, I would think (just like I don't allow the Marshall as White in the Lopez, but play it as Black since most of the traps and pressures work against White).

One time when I transposed to a Panov as White (via 1.c4) against NM (and Newsletter editor, and TD) John Upper, he played 5...Nc6, and I played one of the few ways to avoid the Endgame Variation (though I played 6.Nf3 instead of 6.Bg5, even). It was a way mentioned by Gallagher. I lost, but probably not the fault of the opening. The few other times I've played the Panov I've had better success, though still not as much as with the offbeat 2.Nc3 d5 3.Qf3 (or even more normal 3.Nf3, thus far), which somehow has suited me even better so far, though I haven't faced the Caro nearly as often as even the French, which happens historically about 20% of the time to me when I play 1.e4 (though with Halldor, or formerly John, it happens 100% of the time). I also haven't yet faced too many master+ strength players who play the Caro as Black.

[Vladimir Drkulec]

There was a Jovanska Houska 2015 book on the Caro-Kann. There are also three new 2018 books from Andrew Martin, Daniel Fernandez and Victor Bologan. I am sure that Larry Bevand and Strategy and Games has most if not all of them. The Fernandez book is from Thinkers Publishing and is the only one that might be tricky though Larry usually has a good selection of such publishers.

[Javier Gil]

Vlad: glad to know you're a member of the Caro-Kann lovers club! The caro will always be a realiable and faithful defence, but true, some effort on our side to keep up with modern theory is essential.
Kevin, good point about the ...Nb6 variation. I did mention it in passing when I talked about Jacob Aagard's book on the caro-kann, and his opinion is that "white should have the advantage", but as you pointed out, the book is rather old.
I did a quick search in Mega2018, with games where at least one of the players was 2300. This is the statistics that I got:

393

So, that looks like bad news, and the fact that so few people chose it is also a bad sign...

I did another search in the main line of the ...Nc6 variation, also for games where at least one player was 2300+. The results prove that it is incredibly solid (but difficult to win for either side).

394

Kevin, regarding your choice of subvariations in the Caro, I´m glad to hear you try the Bronstein variation from time to time. I've been playing it consistently for many years.

[Vladimir Drkulec]

Hi Javier,

I don't see your stats as bad news at all. Black plays pretty close to his rating and white does not play much above his rating. I think if you pick your spots the Caro-Kann will get the job done for most players. I have found that young players do not always know how to deal with the middlegames that arise. I may make a push to look at the theory and try to play it in some tournament games in the new year. Resources like chesspublishing.com are useful in prep work.