Studies in Phailure: Episode 3

Studies in Phailure: Episode 3

Jun 4, 2010, 4:31 PM |

I don't have many games that advance to a point where only Kings and pawns remain on the board. Generally my incompetence asserts itself early enough and intensely enough that I've resigned before getting anywhere near a King and pawn endgame. Consequently, I haven't had much experience with the situation. Further consequently, I therefore lose the bulk of those games that do make it that far.

I played White in this game, a Sicilian Najdorf (Amsterdam variation [B93]).  After getting into a sticky situation by the 11th move, I managed to come out of a frantic 9-piece exchange series down by only a pawn after move 17. That one-pawn deficit stayed with me through the rest of the game, and we entered the King and pawn endgame on move 28, with me having 4 pawns and Black having 5. I thought I might have a shot because Black had one pawn isolated on the d-file and doubled pawns on the b-file, but maybe those who know endgames better could say with just a glance that it was a lost position for me all along. Even if it wasn't, I quickly turned it into one.

The diagram at left shows the position as we enter the King and pawn phase, right after I finish a Rook exchange with 28.Kxf2. In replays, watching my King run around aimlessly for the last few moves is pretty pathetic. Worse yet was that I managed to miscount a move sequence, and I let my King get too far to one direction and allowed Black's King to get through to my pawns, eliminating any chance of a draw.


The diagram at right shows the position after Black has taken my pawn with 12...Nxe4 . On the previous move Black played 11...Qb6, by which the Queen targeted my unprotected b2 pawn and also became a second attacker (along with the c6 Knight) on my d4 Knight. I chose to protect my pawn via b4. Now, by playing his Knight to e4 and grabbing my pawn, Black had moved the Knight off his g7 Bishop's diagonal and thus added the Bishop as a 3rd attacker to my Knight, which still had only 2 defenders.

There were four ways to respond to the situation, and I only thought of three of them. One is to remove the target. In this instance that wasn't a viable solution, because removal of the d4 Knight would mean his Bishop now targets the unprotected c3 Knight, from which it forks my King and Rook. Trying to protect the c3 Knight with the other Knight doesn't work: Nde2 or Ndb5 both lose my Bishop to ...Qxe3. 13.Nxc6 results in the loss of both of my Knights and a Bishop after 13...Bxc3+ 14.Bd2 (14.Kf1 leads to checkmate) Bxd2+ 15.Kf1 Bxc6. So toss out option #1.

The second approach is to take his unprotected Knight on e4 with either my c3 Knight or my Bishop, but to do so does not win a Knight for a pawn, because my other Knight still has more attackers than defenders, so leaving that situation means I'll lose that Knight and thus Nxe4 is essentially an exchange of Knights plus probably several other pieces.

A third possibility is to accept the temporary loss of the pawn and add a 3rd protector to my d4 Knight via Nce2. I decided to take this route. I knew that both the 2nd approach and this approach would likely lead to a large series of exchanges (which is what happened), but when I played out the lines I thought I ended up in a better position after the 3rd approach than after the 2nd. The material would be the same either way-- I would be even for everything except that initial pawn.

Rybka preferred a 4th approach, one that I hadn't thought of. Instead of adding a 3rd protector, it would also be effective to decrease the number of attackers to two. This can be done with 13.Nd5, targeting the Queen. Obviously the Black Queen can't lead off an exchange series on her own, or I'd just take her on the first exchange with a lesser piece and then refuse to complete the rest of the series, preserving my own Queen. So the Black Queen has to move off of b6 by retreating. 13...Qa7 might look reasonable at first, staying on the diagonal and maintaining a third attacker on the d4 Knight, but that d4 Knight can now move, because the Bishop on e3 that was previously vulnerable to capture by the Queen is now protected by the Knight that just moved to d5. Thus, 13...Qa7 can be met with 14.Ne6 (discovered attack on the Queen) Qb8 15.Nxf8 Kxf8 16.Rb1 and Black loses a Rook for a Knight (if 15...Bxa1 then 16.Nxd7 Qd8 17.Qxa1 Qxd7 18.Bxe4 and Black is down 2 Bishops). Bottom line, 13.Nd5 leaves only ...Qc7 and ...Qd8 as viable replies. Time to smack my forehead-- why didn't I think of that?!

From about move 17 onward, the Mobility Chart reflects a matched steady decay for both sides, indicative of a game with pretty equally matched positions where both players are losing mobility strictly through piece exchanges, rather than positional disadvantages.