Why Retreat Your Bishop in This Common Position?

The following comes up very often it seems.  My bishop pins their knight, after 1.d4 Nf6 2.c4 e6 3.g5 h6,  and they are bringing out their pawn at h6 to attack my bishop.  Instead of going ahead and taking the knight with my bishop, I see in most higher level games the bishop retreats to h4.  But then they can hastle my bishop by g5 and make it move yet again to g3. 

   Why is it good to let them keep making me move the same piece multiple times while they advance and develop?  (Is it possibly because it gets their Queen out early?  or destroys their king side castling protection?)

Thanks for any answers from highly rated players.  Ipren?  et al?   

Because g3 is a good place for the bishop, the pawn moves aren't really tempo gaining in as much as they aren't getting any new pieces out and they are weakening their kingside.

The first retreat is simply maintaining the pin. Beyond that, playing g5 and g4 weakens the kingside, and usually that's where the black player would usually like to castle, so in general, it's not a bad thing to keep your valuable bishop and let your opponent create structural weakness in their position.

There will of course be situations when chopping off the knight is best, but quite often it's worth spending the tempi to preserve your bishop and entice your opponent into creating the aforementioned structural weaknesses. So yeah, you were on the right track when you said that it destroys their kingside castling protection :)

I'm sure some players with higher ratings than I will come along and give you thorough answers, hopefully with examples!

The pin can be troublesome (or simply doing a good job) so by going to h4 you're keeping the pin.  If they go "all out" to break the pin with g5 then  yes, they've loosened their kingside while your bishop (presumably) wont be badly placed on g3.  In any case weaknesses pawns leave behind are permanent features of the game while your bishop can go somehwere else eventually if it really needs to.

continuing with g4 just to kick the knight... there's no such thing as a one many army in chess.  Don't let moves like this trouble you unless your opponent has a follow up.  If there aren't any other pieces contributing to the attack then your opponent is just hurting himself for the sake of one move threats.

But of course it depends on the position.  Sometimes it's worth it (or at least not bad) to play the knight pawn out and kick the bishop again.  The point is it has to be part of an overall idea though... the plan can't be "oh, I get to feel safe for 1 move because he has to move his bishop"  You need to plan to make use of the kingside space while deciding your opponent can't use it against you.

Ahh, thanks for the replies so far.  This has bothered me for the last 3 months.  Before I have always gone ahead and taken their knight to double up their pawns or get the Queen out early.  Bad beginner's thinking I suppose.  Yay!  Yet another leak taken out of my game!

Not IMO, you're giving up the bishop pair for no compensation. The queen will be on a safe square, there's no advantage to getting it out early on f6, the only advantage is if it comes out early to a square you can attack while developing.

If taking the knight gives them doubled pawns (e.g. Ruy Lopez exchange variation), then it's a trade off between doubled pawns and bishop pair, which I think most people would say was fairly even.

Capturing in the opening like that usually leads to a loss of time.  If you spend a move capturing and they spend a move bringing out their queen (or a knight or bishop) you've lost time.  (whether their piece is well posted on the re-capture square or your bishop was more important is worth looking at, but just talking basics here).

Keeping the pin also limits blacks development options.  Not that he'll have only bad moves to make, but, for example, he'll have to decide he's ok with doubled pawns before moving the queen away.

That said, the doubled pawns don't offer white much in a middlegame.  The good points for black are an open file (e or g) and better control over e5.  If it goes into an endgame, and you're damn good, then you might prove it's a tangible/winning advantage :)

You can also consider the bishop pair is worth a bit usually, so giving up a bishop so willingly, you'll want it to be part of a larger idea... like doubling his pawns or as a way to fight for the e4/d5 squares (the squares a knight on f6 guards).  Something like that.

Sorry for no examples... maybe someone can sum it up better for you... in any case chess is a complicated game with lots of exceptions, so it's hard to give a really straightforward answer.

Oh, didn't notice your notation at first:

Capturing is fine here I think.  The queen isn't so great on f6, and after gxf6... in both cases to me, the point is black doesn't have space so things like the queen coming out (or after gx the bishop pair) don't credit black's position like they would if the pawn structures were reversed for example.

And even playing g5 after the retreat doesn't look bad... reminds me of trompowsky opening when e6 is played.

Thanks again Scott, WaffleMeister and Ireland.   For the first time ever, I think I'll just try the bishop retreat and see how it hampers black's king side castling.  (Before, it was just a very common thing I did, bringing my bishop out to pin his knight to his queen and then to double his pawns.  But if I can keep taking out any minor leaks in my game, I will continue to improve.  Yes, I'm also working fervently on my middle game and end game.   )

Ha!  I play a computer at 1600 and beat it 3 out of 4 times.  But this time I retreated my bishop twice to g3 . . . and lost the game.  I guess it was due to other factors.  Need to try it again.  Yep, like Firebrand said, seems black had a strong attack on my Queen side and threw everything he had at me there.

It shouldn't be such a strong attack.  Sure that's a general idea, but by no means should it be one sided, black simply has some pressure/play in that area.  For more specific info it would be useful to post a game.

Ah yes, Estragon brings up a point I'd forgotten - develop knights before bishops and according to Lasker, 'don't pin the knight before opponent has castled.'   And the general rule, bring out your knights and then a bishop straightaway to f3/c3 or f6/c6 (but probably not both, as White). . . in general of course. 

Thanks for the link.

Of course, Paul.  Makes me wonder though, by pinning his knight early, breaking this bishop out early rule,  and retreating twice to g3, WOULD destroy his king side castling, thus causing much positional damage. 

thats good opening.