In my game with a player equal to my strength, I decided to do an experiment with a rook sac for a dark squared bishop. I saw that all his pieces will be scrambling for their life to protect the king, and my double bishops might have proven superior to his passive rooks. In the end, my opponent got in confusion and also broke down mentally as my pieces kept pressuring the queen and the light squared bishop, his only active pieces on the board. Please tell me, all you experts. Is this a legitimate sacrifice? I would like all of your helpful opinions. Thank you in advance. Here is the game:

I would also like to know my mistakes and my opponent's mistakes. Thank you again.

My analysis : (please note I am no expert for the English opening with KID for Black, so there might be more flaws in the opening than elsewhere)

As for your question : the sacrifice is worse than dubious, especially with the hidden tactical line I found.

thank you. I see that there are many holes in our part. Your analysis was accurate and clear. The problem was that black was in time trouble after the dubious sacrifice.

agree with everything except for I am not so happy about q-e4 after your cxd4 line...eg e5, e3 and the black dark square bishop is on the wrong side of the board and the white bishops are going to be monsters with no compensation for the poor structure by black...

Oh and 11..b-f5?? 12 nxf5 and 13 qxf5 is pretty horrible for black no?

Agreed : 11...Bf5 is not good at all, thanks.

As for the ...cxd4 thing, I thought of ...Be5 rather (it's not great, but it's playable). More analysis :

bf6 is suicide...f5 has to be winning (eg n-e5 then g4 with the idea that after gxf5 Rxf5!  If you don't take then r-h5 comes with devastating effects.  ANd if you offer to exchange queens then white gladly accepts with the bishop pair,  a pawn up, better pawn structure and no use of the "extra exchange as the files are not useful...if n-c5 then nxf6 followed by q-h4 sacking the exchange on f5 if white likes positional suicide.  If black plays g5 then q-h6!? with ideas like h4 etc...But, much more subtle and dangerous are the bxg5 and f6 idea winning the queen) Have to look at th e other line some other time when I have time

When you want to sac the exchange for the bishop the queen and bishop need to use the long diagonal as soon as possible. For example, in the grunfeld the queen lands on a1!

Oh I see now...there is only one type of position where U can sac the exchang :/  Now I have to post a line damn you

No, what I mean is that the idea of sacking the exchange for fianchetto bishop is to use the long diagonal. The OP never really gained control of the diagonal in his game, so black was better.

There are plenty of other times to sac the exchange, for passed pawns, active pieces, etc, but in this case the idea is to destroy black's best piece and access the diagonal.

Also, in the line I gave black doesn't have the best coordination, this probably justifies the sacrifice.

well, you posted that right after I posted about sacking the exchange so I (mistakenly?) thought you were talking about my line?

And I am having trouble posting a line :(

No I was talking about the OP, sorry.

"Is this a legitimate sacrifice?"

 Legitimately bad.

PS Get some chess software 

man....I spent all that time postiing the diagram(and probably greatly flawed diagram at that lol)

@erikido :

First of all, in your diagram, I prefer Black in the final position. Personal taste, I am not at all sure of what Houdini would say, but it's not desperately losing and might even be winning. Black will probably have to sacrifice a rook vs. bishop and pawn, but that does not make him worse off. For instance 10...Rfc8 11.b3 (maybe not the best) Kg7 12.Bb2 (?) Nxc4 13.Bxc4 Rxc4 14.bxc4 (the desperado Bxf6+ may be better, but it takes a unimportant pawn at the cost of a tempo) 14...Rxb2 and Black's structure is more than compensated by his rook activity.

Second, instead of 4...gxf5 after what 5.gxf5 should be enough to win, I would play 4...g5 ; for instance 5.h4 h6 6.hxg5 hxg5 and Black gets to the h file first (...Kg7 and ...Rh8).

Same for your sideline with 2....Nc5, 4...g5 maybe works (less sure though) followed if 5.Qh6 by ...Kh8 and ...Rg8. You can at least admit that 4...Bxf5 is better than ...gxf5 (creating an ugly tripled pawn for no reason while delaying development).

In the sideline ...Re8, 6...Ng4 may be better than ...Nd7, it gives the additional possibility of ...Ne3.

Of course you like black better when you offer pawns as white to black....b-b2?????(I don't think I can give enough question marks) as opposed to b-a3 after rc to d8 seems forced and with the rooks passive a pair of raking bishops, mobile queenside pawn majority for the exchange(that knight in the center looks pretty but it can only really be a defensive piece

4 g5 is also just (at least slightly) worst for black.  Don't be so obsessed with the h file.  Whites slight advantage lies in the fact that black has more than one weakness(on the kingside and in the center).  So instead of immediately ramming the pawn white will play r-d1 with annoying pressure(the h pawn can be pushed at some point to after the black pieces start trying to defend the lowly central pawn.  (maybe b3-ba3, maybe b4 and c5)  What are blacks plans?  Try and take advantage of the queen in the center maybe?  That doesn't work,  1. rd1, r-d8 2. b3, and now if black uses his discovered retreat (n-c6) then white responds with q-d5 picking up the d6 pawn while that f6 pawn becomes weak. 

I can only think of one sacrifice I've done which didn't immediately win back the material.

My pieces were getting cramped and I'd get 2 pawns for the knight and an open king. Played a few mistakes and never really took advantage of the exposed king but won a nice pawn endgame.

http://www.chess.com/echess/game.html?id=54291654

Usually my sacrifices win back material: like this greek gift! It's not often you get to play a greek gift sacrifice in the QG, pretty much my first one ever, although I know the theory.