Why h4 h6 in Caro-Kann?

The classical variation for Caro-Kann goes as follows:

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.h4 h6

My two questions:

1. Why h4? The claimed goals of that move are to threaten to trap the black bishop and to strenghten the king side. However, I do not see any serious threat on the bishop and I thought that advancing a pawn on the castle side (since the variation often continues with a castle on the king's side) was a weakening rather than strengthening move. Why isn't a developping move like Nf3 considered a better alternate?

2. Why h6? What is so immediately threatening the bishop that would cause black to reply with h6, which in my mind has the same drawback as h4?

Thank you

Why h4?? 1.basic motive is to trap the Bishop.  2. I think its also cause black to loose tempo, (He moved the same bishop twice, as white also moved his knight twice) (so compensation).  3. By no any means it weaken the kingside of white  

Hope I could help you!   AND about Nf3, you are right it also a good move, (ECO B18)   http://www.chessgames.com/perl/chessgame?gid=1295458   check out this game (you need java to see this game)  It played with Nf3

Actually, I've wondered the same, lbonetto.  I've read that it supposedly causes endgame weaknesses, and doesn't seem to portend any major advantage in the opening or middlegame, yet it's rare to see a mainline Caro-Kann game without these moves.  Why?

Your 2nd question is a simple one. If black plays something like Nf6 or Nd7 it comes h5. The bishop it's trapped.

Then, why h4, i'm not comlpetely sure how, but i belive that if GM plays it, it has to be a good reason. Probably having a slightly more advantage than playing something like Nf3 as you said. I don't think that Nf3 is bad as VINDYVINDY said too. My recomendation for you is to search for GM games with and without h4 (and h5) and study them a bit. In the end i think that it would be better for you to play to do the one you understand more if you are playing to win (or killing some time). But if you are intending to learn, play the thing u wanna learn about. Make mistakes, identify the mistakes and fix them

Good to know ;). Also, i'm not sure if it's true or not (but certainly it's a possibility), white can play Rh4 to go with the rook to the queenside easily. I'm not sure if this is commonly done, or even if normally fails, but an idea is an idea xD

Here are my thoughts. Note that while I do play the Caro-Kann, I play the 4...Nd7 variation, not the 4...Bf5 variation.

• When white plays h4, black could play e6. [6.h4 e6 7.h5 Bf5 8.Nxf5 Qa5+ 9.c3 Qxf5 10.Bd3 and white has a distinct advantage, with the bishop pair, better development, a better center, and more space.]
• For that matter, after 4...Bf5 5.Ng3 black could play a slightly different version of the above (without h4-h5 for white) [4...Bf5 5.Ng3 e6 6.Nxf5 Qa5+ 7.c3 Qxf5 8.Bd3 with much the same position, but perhaps a tick less bad for black because white doesn't have that h5 pawn]
• So given that, black usually tries to preserve this bishop, and doesn't want to play ...h5. This leads to 4...Bf5 5.Ng3 Bg6 6.h4 h6.
• After 6.h4 h6 white usually plays 7.Nf3 with the threat of 8.Ne5. So this forces black to play 8...Nd7.
• Only after the inclusion of those moves does black play h5 and then Bd3.
• This leads us to the typical main line of [4...Bf5 5.Ng3 Bg6 6.h4 h6 7.Nf3 Nd7 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3]
• All that being said, sometimes white for whatever reason prefers to play 6.Nf3 instead of 6.h4, or sometimes white prefers his pawn on h4 instead of h5, by playing 8.Bd3 Bxd3 9.Qxd3 instead of including 8.h5 Bh7.

The typical reasons to have these rather subtle changes of opening are because white is looking for a way to play the position other than the "normal, recommended" way. I'm not sure of the current state of theory in the 4...Bf5 line.

I like the analysis xD. Thanks Ozzie ;)

All. Thank you so much for helping me understand this.

xD, yes, when I don't understand the theory, I go with what you suggest: my instinct. I didn't do h4 on the current game I am playing and will learn something from it I am sure.

Ozzie, thank you, thank you, thank you. Great point by point explanation.

The basic idea of h5 is to take advantage of the position of the black queen's bishop to gain space on the k-side and when this has been done to the maximum extent, with the black bishop pushed back to h7, trade it off for the white king bishop. If White doesn't  do this, then black's queen's bishop will be outside black's light square pawn structure and can be useful rather than being a "bad" bishop (as it is in the French defense), with no compensating space gain for white. The advanced white h pawn is a precondition for a white k-side attack if black tries to castle K side, which is why Q side castling is the traditional main line for Black, although that has been changing recently.  White usually doesn't plan on castling K side in this line, so doesn't mind advancing pawns-the position of the bishop gives him a positional justification to do so. It's not a tactical move and White doesn't really expect to be able to trap the bishop, just to make Black pay for the privelege of getting his bad bishop free.

The idea of advancing the h pawn is credited to Spassky I believe. The purpose is to gain space on the kingside by taking advantage on the black light squared bishop with the purpose of mounting on attack on that side by forcing black to weaken his kingside position by h6. Black in this variation may castle queenside for a quieter game or a more aggresive and modern approach where he castles kingside and meets the attack while mouting his own on the queenside. The white h pawn usually becomes an endgame liability if black is not blown off the board.

I can say from my experience that h5 sometimes (though really rarely) lets white play Rh4 and jump over 4th rank

Its main idea is though gaining space on kingside. In this variation white usually puts the knight on e5, then if black trades this gives e5-f4-g3,h5 structure vs. e6-f7-g7-h6

I found the pawn on h5 useful for breaking through the black pawns in this structure, often play g4, f5, f6 (with bishop on c1-h6 diagonal) which makes h6 very weak and many times give me the passed pawn h5. Also tried e4-e5-g5-h5 and if black takes hxg5 then f6 makes h5 a passed pawn (symmetrically fxg5 g6 clears the way for the e5 pawn).

Contrary to many opinions about endgames being comfortable for black in Caro-Kann I found the endings coming from Spassky variation quite nice for white.

Probably stronger players can prove that is good for black - I have not met anyone stronger than FIDE 2360 in C-K

There is also another move order, namely N1e2 after Bg6 with the idea to play Nf4, followed by h4

if black answers h6 in that situation white simply captures Nxg6 fxg6 and black's pawn structure is completely demolished.

Spassky v Karpov games from 1974 Candidates