Benidorm: Typical Mistakes & How To Avoid Them (Part Seven)

I MAY have missed a win with white against Daniel Martínez Dourado (1837) in round seven at the Benidorm (Meliá) U2000 tournament. 

Black has just played 41...h5

*****

*****

*****

*****

*****

*****

*****

*****

*****

White has an extra pawn, but the white kingside pawns are targets for the bishop. My main analysis  engines reckon White is winning after 42.Ke2 or 42.Nc7. However they reckon both moves are of exactly equal value (+6.50, according to Stockfish15; +2.66, according to Komodo13.02), which in an ending is often a sign that such an advantage is not real.

I played 42.h3?, after which the engines reckon White has at best just a slight edge.

It is fairly easy to see why White's advantage might be negligible, so let's take a look at how the engines reckon the game could have been won.

The first point after 42.Ke2 is that 42...Bg1 can be met by 43.Nb4, when 43...Bxh2 44.Kf3 means the black a pawn will fall, giving white a passed pawn.

So instead the engines suggest 42...Bd6 43.Kd3 Bc5 44.Nc7 Bg1 45.Ke2 a5 46.Kf3 Bc5, when it is not completely clear how White makes progress.

One line runs 47.Nd5 Ke6 48.Ne3 Bb6 49.Ke2 Bd8 50.Kd3 Bb6 51.Nc4!? Bg1 52.Ke2 Bxh2 53.Kf3 Bg1 54.Nxa5 Kd5, when White has a passed pawn but the black king is very active.

A possible continuation is 55.Nb7 Kc6 56.Nd8+ Kd7 57.Nf7 Ke6 58.Ne5 Bb6 59.Nc4 Ba7 60.a5 Kd5 61.Ne3+ Ke6 62.a6 Bc5 63.Ke2 Bb6 64.Kd3 Ba7 65.Ng2 Bb6 66.Nh4 Bg1 67.Kc4 Bf2 68.Kb5 Kd5 69.Nxf5 Ke4 70.Nd6+ Kd5 71.Nc8 Ke4 72.Nb6 and wins.

Is this convincing? Perhaps, but there were many alternatives along the way for both sides.

Now let's look at the game continuation.

Black played 42...Bd6, and I created a passed pawn on the kingside by 43.g4?! fxg4+ 44.hxg4 hxg4+ 45.Kxg4, but after 45...Ke6 the position is almost certainly drawn.

A better try may be 43,Ke3 Bc5+ 44.Kd3, but 44...Bf2 shows why my 42nd move was mistaken, ie 43.Nb4 Bxg3 44.Nxa6 Bxf4 gives Black a rival passed pawn and a fairly smooth path to a draw.

I am far from convinced by much of my (engine-assisted) analysis here, and any win after 42.Ke2 or 42.Nc7 is not easy, but 42.h3? followed by 43.g4?! was part of a faulty plan that left too few pawns on the board.

However 42.h3? with following up by pushing the g pawn was wrong anyway, as concrete analysis shows getting the pawn off a dark square weakens the remaining white kingside pawns.