Annotated Game #68: How to deal with the QGD setup in the English?

The following seventh-round tournament game features an old problem: how to deal with the Queen's Gambit Declined (QGD) setup in the English.  White tries yet another approach, this time exchanging on d5 immediately.  Analysis of the game shows that this is not a bad way to play, particularly if White had tried a different approach on move 8; the game included in the notes from Jesse Kraai is interesting to see, in that respect.

Prior to embarking on a comprehensive analysis of my tournament games, I had not realized either the frequency with which I had actually faced the QGD, or the difficulties inherent in playing the English against it, rather than simply transposing with d4 to the main lines.  (There of course are plenty of other difficulties involved in that, including the large body of opening theory.)  As a result, I've now worked out a reasonably consistent approach involving an early e3, which I'm satisfiied with (if not completely happy).  This should have better practical results than essentially randomly picking from the variety of other early move choices (4. g3, 4. b3 and 4. cxd5).  As I noted in Annotated Game #64, the lack of such a consistent approach made it feel like I was playing a new, unfamiliar opening each time.

Going back to the actual game, White makes a number of small errors and one significant one on move 13.  The engines' recommendation of 13...a5 I found instructive, showing how Black can use that type of pawn lever against White's queenside formation when it is left underdefended.  Although Black retains a noticeable advantage, thanks to White's somewhat incoherent strategic play, White is smart enough to realize it and then manages to trade down into a drawn position.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class C"] [Result "1/2-1/2"] [ECO "A17"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "59"] [EventDate "2006.??.??"] {A17: English Opening: 1...Nf6 with ...Bb4} 1. c4 e6 2. Nf3 Nf6 3. Nc3 d5 4. cxd5 {I've played a number of moves at this point against the QGD setup. The most popular way to play this position is to simply transpose to a QGD after d4. The text move is the second most popular.} exd5 5. g3 {again, transposing with d4 (into an Exchange QGD) is the most played here, but I prefer to take an independent route and keep the game in English territory. The text move is again second most popular.} Bd6 {this makes the d5 pawn weaker and doesn't seem to place the bishop on a particularly useful diagonal. c6 is most often played and seems logical to play immediately, thereby reinforcing d5 again and blunting the Bg2 development.} 6. Bg2 O-O 7. O-O c6 8. a3 {Prevents intrusion on b4, notes Fritz. The point was to prepare and then execute the b4 advance.} (8. d3 {would seem to be a more useful move, if one intends to avoid more standard play with d4. Here is an interesting sample game from Jesse Kraai.} Nbd7 9. e4 dxe4 10. dxe4 Qc7 11. Nd4 Rd8 12. Nf5 Bf8 13. Qc2 Ne5 14. Bg5 Bxf5 15. Bxf6 gxf6 16. exf5 Qa5 17. Ne4 Be7 18. Qe2 Rd7 19. Qh5 Rad8 20. Qh6 Ng4 21. Qh5 Ne5 22. g4 Nd3 23. Qh6 Qe5 24. f4 Qxb2 25. g5 Nxf4 26. Rxf4 Qxa1+ 27. Bf1 Rd1 28. g6 Rxf1+ 29. Rxf1 Qd4+ 30. Nf2 fxg6 31. fxg6 hxg6 32. Qxg6+ Kf8 33. Qh6+ Kf7 34. Qh7+ Kf8 35. Qh8+ {1/2-1/2 (35) Kraai,J (2454)-Pilgaard,K (2440) Budapest 2003}) 8... Re8 9. b4 Nbd7 10. d4 {White, having carefully avoided this move until now, plays it at a bad time.} (10. Bb2 {would have been a more logical follow-up for development, taking advantage of the pawn not being on d4 to cut off the bishop's long diagonal.}) (10. Re1 {is Houdini's top choice. The rook is best placed on the e-file.}) 10... Nf8 (10... h6 {is a good example of a useful prophylactic move.}) 11. Bg5 {White's idea is to exchange off the knight, which is well-placed for sallying onto e4 or g4. Having now locked the bishop out of the long diagonal, perhaps it's not such a bad idea.} (11. b5 {however would be a nicely disruptive move on the queenside.}) 11... h6 12. Bxf6 Qxf6 13. Rc1 $6 {the c-file is not a productive place for the rook, which is better placed on the a-file.} (13. Re1 Bf5 $11) 13... Bf5 (13... a5 { is how the engines exploit the rook move. for example} 14. bxa5 Bxa3 15. Rb1 { and the a-pawn will fall.}) 14. Na4 $15 {now ...a5 would still break up White's queenside formation to Black's advantage, but no longer lead to a material loss.} Ne6 (14... a5 15. Nc5 axb4 16. axb4 Bxc5 17. bxc5 Ra2 18. e3 $15) 15. e3 {White sees Black's latent threats down the e-file.} Bg6 {waste of a move.} (15... Bg4 $5) 16. Nh4 Bh7 17. Qf3 {White decides that Black has all the play in the position and attempts to head for a draw.} Qxf3 18. Bxf3 $6 { This leaves things awkward for the Nh4.} Ng5 (18... g5 {would have driven the point home.}) 19. Bg2 Rac8 $11 (19... a5 {again would be a good try for Black to generate play on the queenside.}) 20. Nf3 Nxf3+ {Black would need to preserve the knight in order to attempt to play for an advantage.} (20... Ne4) 21. Bxf3 {White can now successfully block any attempt at progress by Black.} Bf5 22. Rc3 Bd7 23. Nc5 {this forces the exchange of bishop for knight, Black gets to decide which bishop. Without the two bishops, any theoretical plus for Black vanishes.} Bxc5 24. bxc5 Rc7 25. Rb3 Bc8 (25... Bf5 {would be a cleaner defense, controlling the diagonal including b1 and not blocking the 8th rank. However, the position is still drawn.}) 26. Rfb1 Rce7 27. h4 g6 {Covers h5} 28. Kh2 Kg7 29. Bg2 f5 30. Rh1 {neither side will be able to break through.} 1/2-1/2