A complex endgame

[Arpad Rusz]

[d]8/1Kbk1p2/8/Nb5p/1Bp2p2/1p5P/3P4/R7 w - - 0 1

1.d3! seems to be the only move which keeps the advantage. Is that sufficient to win?

[zullil]

[d]8/1Kbk1p2/8/Nb5p/1Bp2p2/1p5P/3P4/R7 w - - 0 1

1.d3! seems to be the only move which keeps the advantage. Is that sufficient to win?

Here what the latest Stockfish gives after 13 hours at 25 million nodes per second:

Code: Select all

info depth 54 seldepth 79 multipv 1 score cp 367 nodes 1082076365052 nps 25660862 hashfull 999 tbhits 3449289232 time 42168355 pv d2d3 f4f3 a1d1 c4c3 a5b3 b5c6 b7a6 c3c2 b3c5 d7d8 d1f1 c7f4 b4a3 f4e3 a6b6 c6d5 a3c1 e3d4 b6b5 f3f2 c5e4 d8d7 e4f2 d5g2 b5c4 g2f1 c4d4 d7c6 d4c3 c6d5 c3c2 d5d4 c2d2 f1g2 c1b2 d4d5 d2c3 h5h4 c3b4 f7f5 b4c3 d5e6 b2c1 e6e5 c3c4 e5f6 c4d4 g2c6 d4c5 c6d7 d3d4 d7a4 d4d5 a4c2 c5c6 c2a4 c6b6 f6e5 b6c5 a4b3 f2d3 e5e4 d5d6 b3e6 d3f4 e6c8 f4d5 e4f3
info depth 55 currmove d2d3 currmovenumber 1
info depth 55 currmove b4e7 currmovenumber 6
info depth 55 seldepth 79 multipv 1 score cp 367 nodes 1202644866861 nps 25572159 hashfull 999 tbhits 4073235927 time 47029460 pv d2d3 f4f3 a1d1 c4c3 a5b3 b5c6 b7a6 c3c2 b3c5 d7d8 d1f1 c7f4 b4a3 f4e3 a6b6 c6d5 a3c1 e3d4 b6b5 f3f2 c5e4 d8d7 e4f2 d5g2 b5c4 g2f1 c4d4 d7c6 d4c3 c6d5 c3c2 d5d4 c2d2 f1g2 c1b2 d4d5 d2c3 g2f1 h3h4 f1g2 b2c1 g2f3 c1d2 d5e6 d2g5 e6d7 c3c4 d7d6 g5d2 f3g2 d2c1 g2f3 c4d4 d6e7 d4c5 e7f6 d3d4 f6g7 c5c4 g7g6 c1g5 f3e2 c4b4 g6f5 b4c5 e2f1 g5c1 f7f6 c1b2

[Arpad Rusz]

Great! Thank you for your long analysis!

[Arpad Rusz]

All right, probably for you this is a simple endgame. For me, when white is a rook ahead, and only the 1.d3! pawn sacrifice wins (if it does!), and even Stockfish needs 13 hours at 25 million nodes per second the show some conclusive result, then I think there must be some complexity in the endgame...
Could you be so kind to show us some analysis? Or at least a win or drawn verdict.

[Jack Lad]

If you add a black pawn to h6 then black may be able to hold the draw.
[d] 8/1Kbk1p2/7p/Nb5p/1Bp2p2/1p5P/3P4/R7 w - - 0 1