Endgame Lessons From Crete

ONLY two of my eight games at the Amateur Chess Organization's world senior championship in Crete (I had a compulsory bye in round two) reached an ending.

I am using the 13pt rule to define what is an endgame, ie an ending starts when the value of each side's pieces has fallen below 13 points, using the traditional 9-5-3 scale (pawns are excluded).

Here is the first such ending from Crete.

Black has just captured on d6 in Suresh Jhunjhnuwala (1870 Fide) - Spanton (1814 Fide/1998 ECF) in round seven 

*****

*****

*****

*****

*****

White will soon capture the isolani, and so is winning, according to Stockfish13, although Komodo12.1.1 only gives White the upper hand.

24.Rd3 Rc6 25.Nb5 Nc5 26.Rxd6 Rxd6

Black can keep rooks on with 26...Rc8!?, but White's rook will be more active than Black's.

27.Nxd6 Kf8

It makes sense to activate the king to drive away the white knight before trying to create weaknesses among the white queenside pawns.

28.Kf2 Ke7 29.Nb5 a4 30.Nd4 Ne4+ 31.Ke3 Nc3

Black wins after 31...axb3!? 32.Kxe4?? b2, but I rejected it because 32.axb3 leaves White with a pair of connected queenside pawns.

32.bxa4

The engines reckon even stronger is 32.Ne2!? Nxa3 33.bxa4.

32...Nxa4 33.Nb3 Kd6 34.Kd4 b6 35.a3!?

Before pushing this pawn, according to the engines, White should reposition the knight with 35.Nd2, intending Ne4+, when the white king is threatening to advance.

35...h5?!

Missing a chance to make White's task considerably more difficult.

The engines point out 35...f5, which takes away the e4 square from the white king. White will soon be obliged to move his knight, at which point Black can play ...Nc5, intending to drive the white king off the fourth rank with ....Ne6+.

36.g4!?

It is a well-known rule in such endings that the player ahead on material should in general seek to exchange pieces while keeping on pawns. However SJ may have been worried about the g2 pawn becoming fixed after ...h4, although the black h pawn would then be weak.

36...hxg4 37.hxg4 Nb2?

Much better is 37...f5, which I rejected because I feared the weakness of the g6 pawn after 38.g5, although at least then I could set a not-completely obvious trap with 38...Nc5!?, hoping for 39.Nxc5? bxc5, when the position is a dead draw.

38.c5+ bxc5 39.Nxc5

White has created a passed a pawn, which is almost as effective in a knight-and-pawn ending as in a pawn-ending.

The rest is simple, and I quickly manage to make it even simpler.

39...Nd1 40.a4 Nf2? 41.Ne4+ 1-0