In some games of chess, often in the early middlegame, you will likely reach a point where there are a number of plausible moves, each one instigating a slightly different plan which will frame the remainder of your play in the game. In some cases this may be a simple choice of move order, in which certain responses by your opponent must be considered. In others, it may be about creating a target for your pieces, whether that be a forward outpost for your Knights, the creation of an open file for your Rooks, or saddling your opponent with a weak backward pawn. The decision is not often an easy one as it involves projecting your thoughts forward several moves and ensuring there are no tactical flaws in your chosen idea. There are also other considerations, such as remaining time for both players, and a knowledge of the style of position that suits your own play best, or perhaps given a familiar opponent, a pawn structure that they have struggled with in the past. Active moves are always better than passive ones in this context, they force your opponent to give your play due consideration, whereas passive defensive play allows them to dictate terms. Sometimes if you have a particularly poorly placed bishop the investment of a pawn in order to ensure its freedom is worth it. With the advent of highly advanced engines it is often revealed to be the case that several moves are broadly similar in evaluation in a given position. This allows a player more freedom to steer the game along a path of their choosing. The position below comes after black’s 13th move, in the Cambridge Springs variation of the Queen’s Gambit Declined (Full game and notes attached). Black’s last move was Nd7xe5, completing a pawn exchange on the e5 square. In the position white has a number of possible candidate moves, examples of which follow the diagram:

Garys diagram of the game position]

  1. The occupation of the d file with one of the rooks, the King’s rook being preferable while the Queens rook maintains a close watch on the a pawn.

  2. Playing pawn to a3, forcing the b4 bishop into a decision.

  3. Playing pawn to c5, interfering with the Queen’s protection of the e5 Knight. The exact position has occurred before, from the games I could find, white had either played a3 with some success, or the rather slow looking Nd1 (leading to defeat), possibly with the idea of shoring up the c pawn with b3 followed by Nb2. After some minutes of consideration, and at this point a considerable amount of time behind on the clock, white found a very interesting continuation: 14.Nd5! Putting the position into an engine afterwards, the engine initially preferred the moves a3 and Rfd1, but once Nd5 was entered this started to change. Even after a long period of assessment the move is not rated much better than the previously mentioned moves (and at a similar level to the companion move 14.Nb5), but engines don’t take into account the additional psychological impact of such a move on the second player. Up until this point he had assessed his position as relatively comfortable. The purpose of the move is to put a difficult question to both the b4 Bishop and e5 Knight, the former of which is under threat of exchange and the latter now being en prise due to the interference in the Queen on a5’s defence. In addition there are possibilities of playing Ne7+ advantageously in some lines. In addition the a2-g8 diagonal to the black King is severely weakened, such that taking the Knight with the pawn in exchange for the e5 Knight will lead to a great deal of pressure being exerted by white’s long ranged pieces against the vulnerable black Kingside. In the actual game black fell apart in very few additional moves due to a miscalculation, starting with …Nxf3+?, losing at least a pawn. Had he found the strongest reply, …Bd6, white would still have had the better game, and some momentum. Should white have opted for one of the quieter but equally good moves from an engine perspective such as a3 or Rfd1 in the position, black would have had the luxury of choosing whether to exchange on c3 to saddle white with weakened pawns, or to complete his development naturally and perhaps consider an assault on the white King’s position later in the game.

PGN

[pgn height=400 layout=horizontal autoplayMode=none] Date "2014.12.3"] [White "Sharp, G"] [Black "Saunders, R"] [WhiteElo "146"] [BlackElo "150"] [Result "1-0"] [ECO "D52"] 1. Nf3 d5 2. d4 Nf6 3. c4 e6 4. Nc3 Nbd7 5. Bg5 c6 6. e3 Qa5 {I knew this was the Cambridge Springs variation but was unsure how to continue. After a long think I came up with the Nd2 Qc2 plan which turns out to be a reasonable setup} 7. Nd2 Bb4 8.Qc2 Ne4 9. Ndxe4 dxe4 10. Bf4 {a slight inaccuracy} (Bh4 O-O 11.Be2 e5 12.O-O {with a small plus}) f5 11. Be2 O-O 12. O-O e5 13. dxe5 Nxe5 {this position has been reached before. Previously white played 14.a3 and 14.Nd1. Also possible are 14.Rfd1 and 14.c5} 14. Nd5! {my opponent had been playing confidently until this move and had gained around a half hour lead on the clock but was visibly shaken by it. The engines for the most part do not rate the move much more highly than a3 or Rfd1, but the psychological effect of the move in an over the board situation meant it turned out to be the more correct. My opponent had his first long think of the game and immediately started going wrong} (Nb5! {an interesting alternative sacrifice} cxb5 15. Bxe5 Be6 16.a3 Be7 17.b4 Qa4 18.Qb2 {with advantage}) Nf3+? {black cracks under the pressure. The following alternative variations would have given white an advantage but black would still be in the game} (14... Bd6 15. Rfd1 cxd5 16. Rxd5) (14...cxd5 15.Bxe5 Be6 16.a3) 15. gxf3 cxd5 16. cxd5 Qxd5?? {a pawn grabbing miscalculation played instantly, but with an immediate refutation} (16... Qb6 17.Qb3 Rf6 18.Be5 Rg6+ 19. Kh1 {and white has a clear advantage}) 17. Bc4 {and the Queen is lost} 1-0 [/pgn]

Gary