Caro-Kann with 3.Nc3

Recently, I started playing the Caro-Kann defence as black.

I studied some of the main-line variations, and I came upon a small curiosity.

And now I question myself. 5...gxf6 weakens the kingside pawn cover, but 5...exf6 could eventually lead to 6.Qe2+ Be7 and now black's bishop is "kind of" pinned there!

Looking on the Chess 365 opening explorer, it says me black wins a little more often with gxf6, but it says me exf6 is more drawish.

Which would you choose, and why?

Thanks for the help!

gxf6 I imagine would be played more in blitz time controls or faster.

4...Nd7?! delays developpement of the c8 bishop.

4...Nf6?! is not the best, but after 5.Nxf6+ exf6 6.Qe2+!?, Qe7!? might be an alternative (probably better) to Be7.

4...Bf5 is met by 5.Ng3 and now black must move his bishop twice in a row.

Talking of popularity, 4...Nf6 gets black to win more often(almost 3% more than Bf5, and 5% more often than Nd7), according to Chess 365. Also, Nf6 is the move played by my computer chess software. By the way, it says that black's best move after 5.Nxf6+ is exf6. And I don't like playing out too long theoritical lines. A last look at the opening database: 6.c3 Bd6 7.Bd3 O-O 8.Qc2 Re8+ 9.Ne2 g6 10.h4 Be6 (the most popular moves, except the last, which is much better than the most popular(Nd7, which reduces black winning chances to 17%)) and now there is equal counterplay on both sides.

Thanks for the help. Is there something I'm missing?  

the gx capture, Bronstein-Larsen variation it's called, is really, really fun.   I play it a lot, win a lot.   In a fast game, white is often caught off-guard by the open g file and fails to see the attack potential there.   In a slower game, with care, black can quell central/queenside stuff (having castled 0-0-0 or remained central with king) and nurture the pawns along.  

Against a computer, I probably wouldn't play it.   I'd guess it's not 100% sound.   But against humans, it's well worth trying.  

Thanks for the tip. But would you rather play the exf or the gxf capture?

I would rather play exf. black is likely to suffer after gxf (6.c3 Bf5 7.Nf3 e6 8.g3 and black's play is no fun). exf gives more chances to hold an inferior endgame

Black has lot of fun here : Qc7, Bd6, then h5, h4 or f5-f4, Rg8 and so on. White attack is slower.

This is also my experience.  It's one thing for white to play some sound prophylactic moves, but another to say that Black has a bad game.  Rather, while White takes care of problems in his own formation, black is developing pieces and aiming at immediate counterplay.  Perhaps this is over-ambitious for the black pieces, but I prefer this style of thinking to simply waiting around for a double piece sac and mate combo from white, or moving toward a depressing and chance-less endgame (French defense anyone?).  

You are insane!  4...Nd7 does not deserve the "Dubious" tag (?!).

4...Nd7 is perfectly sound, as is 4...Bf5.  Your moron comments like "4...Bf5 is met by 5.Ng3 and black must move his bishop twice in a row", implying it's bad for Black because he moves a piece twice in the opening, just go to show what little you know about the game.

What is your basis for this BS?  The old adage "Don't move a piece more than once in the opening"?  It's a general comment for ABSOLUTE BEGINNERS so that they don't just dance the knights around the board and do nothing else (typical beginner play).

Did you also consider that WHITE MOVED THE KNIGHT THREE TIMES IN A ROW?  HELLO?????????????????????

I guess you must think that the Scandinavian Defense with 2...Qxd5 is an absolute bust as well as it develops the queen early, and involves moving a piece twice in a row in the opening.

Shoot!  You must think the Ruy Lopez is no good because WHITE moves the same piece on back-to-back moves!

Speak when you know what you are talking about, and not until then!  I've been a long time advocate of the French (formerly, quit in 2007), Caro-Kann, 1...e5, Modern, and Taimanov and Scheveningen Sicilians as Black against 1.e4.  So before you go spouting out how wrong you think I am, I know what I'm talking about here!  You don't see me spouting out garbage assessments about the Najdorf anywhere, do you?  Again, talk when you know what you are talking about!

if you are playing gxf6 the idea is 

- castle queenside

- attack the kingside thru the g file

this allows more prospects than exf6 but is more dangerous for black too

4... Bf5 is the main move, 4...Nd7 is a very sound variation, often played by Smyslov and Karpov, 5...gxf6 is playable but inferior for black and 5...exf6 was played more often lately

gxf6 , sad to say is kind of bogus. Aftr a simple Nf3 Be2 0-0 bf4-g3 c4-d5 white is better pretty easily. Note the moves do not have to played ain a particular order. exf6 is solid, but maybe not th best. I think Bf5 is hard to equalise but if you know theory its good to play.

Nd7 is good, but the pieces ar akwardly placed, so white has to play something lik ng5 or whatever (I kind of forgot th theory). I think white has a strond attack after, say 10 moves. I remember a game where I played Ng5 the a few moves later i had to play ne4 becausse I forgot theory

Well, I think in the Bd3 line white does return his knight to e4.

White is better, but black's attack can become vicious quickly if white is not careful.
Larsen played it a lot. Obviously this doesn't signify the variation is sound. 

Well put pfren.  I also love how the baffoon goes to some stupid database that probably includes amateur games (365chess.com has games of mine against other non-masters, and I'm only an Expert, while newinchess.com does have about 10 of my games or so, but they are all against masters, so at least one master is involved in each) to get these "17% winning chances" statistics.

He probably has no clue that Black winning 17% of the time, in whatever database he is looking at, doesn't mean that Black scores 17% with that opening variation.

I'm going to wager that if he were to use 365chess.com, and looked up an old line in the French Tarrasch that is highly drawish, namely 1.e4 e6 2.d4 d5 3.Nd2 c5 4.exd5 exd5 5.Ngf3 Nc6 (I prefer 5...Nf6 for Black) 6.Bb5 Bd6 7.dxc5 Bxc5 8.O-O Nge7 9.Nb3 Bd6 10.Re1 O-O 11.Bg5 Bg4, he'd probably be clueless and think that Black only scores 11.9% (as that's the percentage of games that Black "wins") when in reality, it's 46.2% (68.6% of the games are draws), and that also doesn't mean that this line scores exactly 46.2% across all of chess, which he probably is also clueless about.

Chess.com should have users enter their FIDE, BCF, USCF, or whatever other "over the board" ID number into their profile and track rating.  If you aren't at least 1800, you can't post into the "Openings" forum, but this is serious where I see the worst comments in all of the chess.com message board!

Tubebender - As for the Knicks, they are no different than anybody else in New York sports for 2013.  2013 was a HORRIBLE year for New York in professional sports.  I've gone thru the agony of 162 games of Mets baseball, an NFL team that starts out 0 and 6, and now this garbage with the Knicks.

That said, in response to your message, if he had phrased it in the form of a question, asking about how to read database statistics and how trustworthy they are, that's fine, and I'd be more than glad to explain it to him, but just spouting out cr*p like he did, trying to make it sound like he knows everything when he really knows nothing, does nothing but make him look ignorant, and I hate ignorant people.  Keep in mind, there's a huge difference between "ignorant" and "lacking knowledge in a certain area".  A person that lacks knowledge in anything, speaking Spanish, cooking, tying your shoes, or playing the Sicilian Najdorf, who is "not ignorant", would have enough brains to ask for information.  An ignorant one will spout out whatever nonsense just to try to sound good.  It's like a company I worked for in May of 1999 thru early 2000 when a large consulting firm overtook our small one, where the office in Charlotte had 3 or 4 managers when the large firm took over.  NONE of the managers knew a lick of technology, but instead of having the developers (what I was then) explain to them technical terms, and how to use the application, they would have these "buzzwords" that they would just go spouting out over and over again, sounding once again, outright ignorant.  Needless to say that's the worst company I've worked for in my 15+ year career thus far!  So yes, I have very little respect for ignorant people!

YOU ALL SEEM TO HAVE GOT ALL TOPIC!

My question, reworded, was:

After 1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Nxf6+, what was the best move for Black between the 2 following:

-5...exf6

-5...gxf6

Could we please go back to that?

Oh, and ThrillerFan, I know exactly what you mean by "scoring" 46%. But I am looking for an opening(other than the sicilian) that gives black the most WINNING chances, not "scoring chances". That being siad, let's go back on topic.

I can tell you from personal experience that finding out that 5...exf6 is a real opening over the board is a bit embarrassing, it has been relegated to a footnote, but it is playable.

From the perspective of a USCF 1950 player with a knack for openings:

5... gxf6 (The Bronstein-Larsen Attack):

An attempt to be aggressive by Black. Against pedestrian play, there can be some chances of success by castling queenside and blasting down the g-file. This variation has been fairly well refuted, though, by a sequence of moves involving White fianchettoing his bishop on g2, then castling kingside.

The g3-pawn, firmly supported by h2 and f2, presents a pretty firm barrier against g-file assaults, while castling queenside is now dangerous for black, as a b-pawn advance can tear out the c6 pawn and lay bare the h1-a8 diagonal for the king's bishop.

Given these dire circumstances, Black often tries to survive by castling kingside and putting his bishop on g7, but his shattered kingside structure leaves him with very few realistic winning chances.

Conclusion: unless you are playing blitz or feel suicidal, don't touch 5... gxf6.

5... exf6 (The Tarkatower Variation):

1. Attack on the kingside. With a greater share of central space, this is a natural idea for White; it is quite often successful.
2. Trade pieces. Since Black's kingside pawn majority is "crippled" by the doubled pawns, it will be impossible for him to create a passed pawn in that area without a major mistake from White. White, however, has a healthy 4-3 queenside majority, from which he can create a central passed pawn - another detractor of the removal of the e-pawn from its home file.

Often it is possible for White to combine these plans, attacking and scaring Black into trading pieces, which is really what White wanted all along.

Black's one trump card in this variation is the ability to advance his second f-pawn without repercussions, but this is rarely enough to outweigh the weak suits of his opening choice. At its core, the Tarkatower Variation is something of an "accept a disadvantage and try to hold on" opening, which does not appeal to most people.

Conclusion: The Tarkatower Variation is not an opening solution either. While Black is solid, he accepts a significant disadvantage that is likely to translate into an endgame loss.

SO! If I had to advise whether to pick 5... gxf6 or 5... exf6, I would say 5... exf6. Black can struggle for a while, at least, but it's still likely to be futile. My real recommendation is to play one of the main lines.

I prefer to play 4... Bf5, but many people prefer 4... Nd7 just as strongly. Either way, you'll get better chances than in either 4... Nf6 variation.

So long,

~Yohan Saboba