Shirov's Bh3, Houdini solves it

[Lyudmil Tsvetkov]

just checked briefly with SF.

black plays, for example, Be4 Kf2 f5 (and not Kf5, as Dvoretsky would suggest) g3 Kd7:

[d]8/3k4/6p1/3p1p2/p3b2P/2B3P1/5K2/8 w - - 0 4

and white can not hold the draw here

white further plays Kc6-b5-c4, puts one passer on d3, another on a2, returns with the king to h7, and white crumbles.

[Lyudmil Tsvetkov]

most convincing black win however comes after Be4 Kf2 f5 g3 f4:

[d]8/8/4k1p1/3p4/p3bp1P/2B3P1/5K2/8 w - - 0 4

and the most white can get out of this is an ending with a and g pawns, easily won.

[Eelco de Groot]

just checked briefly with SF.

black plays, for example, Be4 Kf2 f5 (and not Kf5, as Dvoretsky would suggest) g3 Kd7:

[d]8/3k4/6p1/3p1p2/p3b2P/2B3P1/5K2/8 w - - 0 4

and white can not hold the draw here

white further plays Kc6-b5-c4, puts one passer on d3, another on a2, returns with the king to h7, and white crumbles.

This is about your first diagram Lyudmil. Just checking with the engine. I think White only has to keep control of all the dark green squares, and Black just can not get through it. f and g pawn are controlled by White's h and g pawn, Bishop stops the a pawn and King has to worry only about the d pawn for which the black bishop has the right colour. Black bishop is like a big pawn that can't promote...


8/3k4/6p1/3p1p2/p3b2P/2B3P1/5K2/8 w - -

Engine: Rainbow Serpent 20160903_029 HT (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

25/37 0:01 -3.70 1.Ke1 Kc6 2.Kd2 Kb5 3.Bf6 Kc4 4.Be7 Kb3
5.Bf6 a3 6.Be5 Bf3 7.Bf6 Bg2 8.Bg7 Kc4
9.Ke3 Be4 10.Bd4 Bb1 11.Bg7 Kb4
12.Bf6 Be4 13.Be5 Kc5 14.Bf6 (11.070.796) 6729

26/37 0:01 -3.70 1.Ke1 Kc6 2.Kd2 Kb5 3.Bf6 Kc4 4.Be7 Kb3
5.Bf6 a3 6.Be5 Bh1 7.Kc1 Bf3 8.Kd2 Be4
9.Bd4 Bb1 10.Bg7 Kc4 11.Ke3 Be4
12.Be5 Kb4 13.Bd4 Bb1 14.Kd2 (12.288.356) 6755

27/37 0:02 -3.70 1.Ke1 Kc6 2.Kd2 Kb5 3.Bf6 Kc4 4.Be7 Kb3
5.Bf6 a3 6.Be5 Kc4 7.Ke3 Bg2 8.Bg7 Kb4
9.Kd2 Bf3 10.Be5 Be4 11.Ke3 Bb1
12.Bf6 Kb3 13.Be5 Kc4 14.Bg7 (14.331.956) 6789

28/41 0:03 -3.54 1.Ke1 Kc6 2.Bb4 Kb5 3.Bf8 Kc4 4.Kd2 Kb3
5.Bg7 a3 6.Bd4 Bb1 7.Bg7 Kc4 8.Ke3 Be4
9.Bd4 Kb4 10.Be5 Kb3 11.Kd2 Kc4
12.Ke3 Bh1 13.Bd4 Bg2 14.Bg7 (23.471.948) 6985

29/41 0:04 -3.54 1.Ke1 Kc6 2.Bb4 Kb5 3.Bf8 Kc4 4.Kd2 Kb3
5.Bg7 a3 6.Bd4 Bb1 7.Bg7 Kc4 8.Ke3 Be4
9.Be5 Kc5 10.Bf6 Kb4 11.Bg7 Bb1
12.Bd4 Bc2 13.Be5 Kc4 14.Kd2 (28.721.522) 7036

30/41 0:05 -3.46++ 1.Ke1 (38.513.193) 7044

30/41 0:05 -3.46 1.Ke1 Kc6 2.Bb4 Kb5 3.Bf8 Kc4 4.Kd2 Kb3
5.Bg7 a3 6.Bd4 Bf3 7.Be5 Kc4 8.Ke3 Bg4
9.Bg7 Be2 10.Bf6 Bd1 11.Bg7 Ba4
12.Be5 Bc2 13.Bg7 Kb3 14.Kd2 (39.194.076) 7055

31/41 0:06 -3.46 1.Ke1 Kc6 2.Bb4 Kb5 3.Bf8 Kc4 4.Kd2 Kb3
5.Bg7 a3 6.Bf6 Kc4 7.Ke3 Bc2 8.Kd2 Ba4
9.Ke3 Bd1 10.Bd4 Bg4 11.Be5 Bh3
12.Bg7 Bg2 13.Bd4 Kb3 14.Kd2 (48.714.949) 7065
.
.
.
43/52 4:28 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Kb3 6.Kd2 Bg2 7.Bd4 Bf3 8.Bf6 Kc4
9.Ke3 Bg2 10.Bd4 Bh3 11.Bf6 Kb3
12.Kd3 Bf1+ 13.Kd2 Bb5 14.Be5 (1.709.926.857) 6373

44/52 5:28 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Kb3 6.Kd2 Bg2 7.Bd4 Bf1 8.Bf6 Ba6
9.Be5 Bc4 10.Bf6 Bb5 11.Bd4 Kc4
12.Ke3 Ba4 13.Be5 Bd7 14.Bd4 (2.091.142.368) 6375

45/53 7:30 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Kb3 6.Kd2 Bg2 7.Bd4 Bf1 8.Bf6 Ba6
9.Be5 Bc4 10.Bf6 Bb5 11.Bd4 Kc4
12.Ke3 Ba4 13.Bf6 Bc2 14.Kd2 (2.860.171.212) 6345

46/54 9:13 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Kb3 6.Kd2 Bg2 7.Bd4 Bf1 8.Bf6 Ba6
9.Be5 Bc4 10.Bf6 Bb5 11.Bd4 Kc4
12.Ke3 Ba4 13.Bf6 Bc2 14.Be5 (3.517.435.842) 6351

47/54 11:59 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Bf6 Bb1 6.Bd4 Bc2 7.Bf6 Be4 8.Be5 Kb3
9.Kd2 Kb4 10.Bd4 Bb1 11.Be5 Kc4
12.Ke3 Kb3 13.Kd2 Be4 14.Bg7 (4.560.118.286) 6339

48/56 14:12 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Bb1 6.Bd4 Bc2 7.Bf6 Ba4 8.Be5 Bd1
9.Bd4 Bg4 10.Bg7 Bh3 11.Be5 Kb3
12.Kd3 Bf1+ 13.Kd2 Bb5 14.Bd4 (5.397.277.443) 6329

[peter]

most convincing black win however comes after Be4 Kf2 f5 g3 f4:

[d]8/8/4k1p1/3p4/p3bp1P/2B3P1/5K2/8 w - - 0 4

and gxf4 a3 and now White can even give back the f-Pawn with f5+ and no matter how Black takes it, with King, Pawn or Bishop, Game is over and drawn.

[Lyudmil Tsvetkov]

just checked briefly with SF.

black plays, for example, Be4 Kf2 f5 (and not Kf5, as Dvoretsky would suggest) g3 Kd7:

[d]8/3k4/6p1/3p1p2/p3b2P/2B3P1/5K2/8 w - - 0 4

and white can not hold the draw here

white further plays Kc6-b5-c4, puts one passer on d3, another on a2, returns with the king to h7, and white crumbles.

This is about your first diagram Lyudmil. Just checking with the engine. I think White only has to keep control of all the dark green squares, and Black just can not get through it. f and g pawn are controlled by White's h and g pawn, Bishop stops the a pawn and King has to worry only about the d pawn for which the black bishop has the right colour. Black bishop is like a big pawn that can't promote...


8/3k4/6p1/3p1p2/p3b2P/2B3P1/5K2/8 w - -

Engine: Rainbow Serpent 20160903_029 HT (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

25/37 0:01 -3.70 1.Ke1 Kc6 2.Kd2 Kb5 3.Bf6 Kc4 4.Be7 Kb3
5.Bf6 a3 6.Be5 Bf3 7.Bf6 Bg2 8.Bg7 Kc4
9.Ke3 Be4 10.Bd4 Bb1 11.Bg7 Kb4
12.Bf6 Be4 13.Be5 Kc5 14.Bf6 (11.070.796) 6729

26/37 0:01 -3.70 1.Ke1 Kc6 2.Kd2 Kb5 3.Bf6 Kc4 4.Be7 Kb3
5.Bf6 a3 6.Be5 Bh1 7.Kc1 Bf3 8.Kd2 Be4
9.Bd4 Bb1 10.Bg7 Kc4 11.Ke3 Be4
12.Be5 Kb4 13.Bd4 Bb1 14.Kd2 (12.288.356) 6755

27/37 0:02 -3.70 1.Ke1 Kc6 2.Kd2 Kb5 3.Bf6 Kc4 4.Be7 Kb3
5.Bf6 a3 6.Be5 Kc4 7.Ke3 Bg2 8.Bg7 Kb4
9.Kd2 Bf3 10.Be5 Be4 11.Ke3 Bb1
12.Bf6 Kb3 13.Be5 Kc4 14.Bg7 (14.331.956) 6789

28/41 0:03 -3.54 1.Ke1 Kc6 2.Bb4 Kb5 3.Bf8 Kc4 4.Kd2 Kb3
5.Bg7 a3 6.Bd4 Bb1 7.Bg7 Kc4 8.Ke3 Be4
9.Bd4 Kb4 10.Be5 Kb3 11.Kd2 Kc4
12.Ke3 Bh1 13.Bd4 Bg2 14.Bg7 (23.471.948) 6985

29/41 0:04 -3.54 1.Ke1 Kc6 2.Bb4 Kb5 3.Bf8 Kc4 4.Kd2 Kb3
5.Bg7 a3 6.Bd4 Bb1 7.Bg7 Kc4 8.Ke3 Be4
9.Be5 Kc5 10.Bf6 Kb4 11.Bg7 Bb1
12.Bd4 Bc2 13.Be5 Kc4 14.Kd2 (28.721.522) 7036

30/41 0:05 -3.46++ 1.Ke1 (38.513.193) 7044

30/41 0:05 -3.46 1.Ke1 Kc6 2.Bb4 Kb5 3.Bf8 Kc4 4.Kd2 Kb3
5.Bg7 a3 6.Bd4 Bf3 7.Be5 Kc4 8.Ke3 Bg4
9.Bg7 Be2 10.Bf6 Bd1 11.Bg7 Ba4
12.Be5 Bc2 13.Bg7 Kb3 14.Kd2 (39.194.076) 7055

31/41 0:06 -3.46 1.Ke1 Kc6 2.Bb4 Kb5 3.Bf8 Kc4 4.Kd2 Kb3
5.Bg7 a3 6.Bf6 Kc4 7.Ke3 Bc2 8.Kd2 Ba4
9.Ke3 Bd1 10.Bd4 Bg4 11.Be5 Bh3
12.Bg7 Bg2 13.Bd4 Kb3 14.Kd2 (48.714.949) 7065
.
.
.
43/52 4:28 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Kb3 6.Kd2 Bg2 7.Bd4 Bf3 8.Bf6 Kc4
9.Ke3 Bg2 10.Bd4 Bh3 11.Bf6 Kb3
12.Kd3 Bf1+ 13.Kd2 Bb5 14.Be5 (1.709.926.857) 6373

44/52 5:28 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Kb3 6.Kd2 Bg2 7.Bd4 Bf1 8.Bf6 Ba6
9.Be5 Bc4 10.Bf6 Bb5 11.Bd4 Kc4
12.Ke3 Ba4 13.Be5 Bd7 14.Bd4 (2.091.142.368) 6375

45/53 7:30 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Kb3 6.Kd2 Bg2 7.Bd4 Bf1 8.Bf6 Ba6
9.Be5 Bc4 10.Bf6 Bb5 11.Bd4 Kc4
12.Ke3 Ba4 13.Bf6 Bc2 14.Kd2 (2.860.171.212) 6345

46/54 9:13 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Kb3 6.Kd2 Bg2 7.Bd4 Bf1 8.Bf6 Ba6
9.Be5 Bc4 10.Bf6 Bb5 11.Bd4 Kc4
12.Ke3 Ba4 13.Bf6 Bc2 14.Be5 (3.517.435.842) 6351

47/54 11:59 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Bf6 Bb1 6.Bd4 Bc2 7.Bf6 Be4 8.Be5 Kb3
9.Kd2 Kb4 10.Bd4 Bb1 11.Be5 Kc4
12.Ke3 Kb3 13.Kd2 Be4 14.Bg7 (4.560.118.286) 6339

48/56 14:12 -3.46 1.Ke1 Kc6 2.Ke2 Kc5 3.Ke3 a3 4.Bd4+ Kc4
5.Be5 Bb1 6.Bd4 Bc2 7.Bf6 Ba4 8.Be5 Bd1
9.Bd4 Bg4 10.Bg7 Bh3 11.Be5 Kb3
12.Kd3 Bf1+ 13.Kd2 Bb5 14.Bd4 (5.397.277.443) 6329

sorry Eelco, I can not vouch that it is either lost or drawn.
I checked a bit with SF, and, in more than half of the lines, I get to a black win (+700 cps), but in a range of lines some (very strange) fortresses also arise.

As I am overworked and overstressed, I can not judge beyond any reasonable doubt if there is a win apart from Bh3. Neither have the time to analyse in-depth.

I have seen this position posted a couple of times here, but, as I judged it too simple, never really paid attention to it. Bh3 is indeed elegant, but I do not see why engines should see it, if there are (possible) other wins.

If there is no win apart from Bh3, I think this will be very unrepresentative of the balance of forces, as black has 2 pawns more with pawns on both sides of the board, eval at root should be no less than 200cps, and in such conditions I do not see why eval should be scaled down at all. Maybe some fortresses hold after all, but if the end position is KBP vs bare K with wrong bishop, making 400cps black edge, the save has been a miracle.

Sometimes, a spare tempo, or a single additional mobility square, worth some 10-15cps, decide the outcome of the game, and in this case it is very difficult to call the game.

migth be back to this position at a later stage again, but currently no more time available to spare on it. maybe people who analysed it have done so extensively, but it is also possible they made some mistake.

would not quite wish to join the likes of Shirov, Sutovsky and Kasimdzhanov, who also were not quite certain if there is another winning move apart from Bh3.

[MikeB]

i remember when this thread came about, thinking this was about to be a dead test position...sadly, not the case. the new engines prune out Bh3 without tricks, even houdini 5 with 50 move distance set to 15 doesnt see it after 15 minutes and 20BN searched. SF, komodo suffer similarly.

interestingly, SF with multipv on finds a tablebase (6 man) mate in just over two minutes on my machine, with multipv turned off , it searched over 10 billion nodes still hanging on to move "a3" with small advantage for black - about 2+ pawns

Code: Select all

info depth 35 seldepth 58 multipv 1 score cp 12816 nodes 4927641503 nps 36287890 hashfull 999 tbhits 18333644 time 135793 pv f5h3 g1h2 h3g4 h2g1 e6f5 c3d2 g4h5 g1f2 f5e4 d2c1 f6f5 f2g3 e4d3 g3f4 d3c2 c1e3 a4a3 e3d4 a3a2 f4g5 c2b1 d4e5 a2a1q e5f4 b1c2 f4h2 d5d4 h2e5 c2b1 e5f4 d4d3 f4e3 b1c2 e3b6 a1c1 g5f6 c1b2 f6f7 b2b6 f7e8 d3d2

[kgburcham]

most convincing black win however comes after Be4 Kf2 f5 g3 f4:

[d]8/8/4k1p1/3p4/p3bp1P/2B3P1/5K2/8 w - - 0 4

and the most white can get out of this is an ending with a and g pawns, easily won.

convincing 40 seconds easily won

[D] 8/8/4kpp1/3p1b2/p6P/2B5/6P1/6K1 b - -

Nalimov 6 men EGTB available - 32 MB cache
Engine: Houdini 5.01 Pro x64-pext (8192 MB)
by Robert Houdart
8 threads used

35/65 0:36 -2.71-- 47...Bd3 48.Kf2 Kf5 49.Bb2 Ke4 50.g4 Kf4 (1.167.002.856) 31929 TB:77.576
35/65 0:40 -2.85++ 47...Bh3 (1.281.428.710) 32005 TB:90.554
35/65 0:51 -2.95++ 47...Bh3 (1.631.827.907) 31945 TB:142.635
35/65 1:09 -3.10++ 47...Bh3 (2.214.629.699) 31989 TB:222.130


[Event "Linares"]
[Site "Linares ESP"]
[Date "1998.03.04"]
[Round "10"]
[White "Veselin Topalov"]
[Black "Alexey Shirov"]
[ECO "D85"]
[Result "0-1"]

1. d4 Nf6 2. c4 g6 3. Nc3 d5 4. cxd5 Nxd5 5. e4 Nxc3
6. bxc3 Bg7 7. Bb5+ c6 8. Ba4 O-O 9. Ne2 Nd7 10. O-O e5
11. f3 Qe7 12. Be3 Rd8 13. Qc2 Nb6 14. Bb3 Be6 15. Rad1 Nc4
16. Bc1 b5 17. f4 exd4 18. Nxd4 Bg4 19. Rde1 Qc5 20. Kh1 a5
21. h3 Bd7 22. a4 bxa4 23. Ba2 Be8 24. e5 Nb6 25. f5 Nd5
26. Bd2 Nb4 27. Qxa4 Nxa2 28. Qxa2 Bxe5 29. fxg6 hxg6
30. Bg5 Rd5 31. Re3 Qd6 32. Qe2 Bd7 33. c4 Bxd4 34. cxd5
Bxe3 35. Qxe3 Re8 36. Qc3 Qxd5 37. Bh6 Re5 38. Rf3 Qc5
39. Qa1 Bf5 40. Re3 f6 41. Rxe5 Qxe5 42. Qa2+ Qd5 43. Qxd5+
cxd5 44. Bd2 a4 45. Bc3 Kf7 46. h4 Ke6 47. Kg1 Bh3 48. gxh3
Kf5 49. Kf2 Ke4 50. Bxf6 d4 51. Be7 Kd3 52. Bc5 Kc4 53. Be7
Kb3 0-1

[yanquis1972]

May be it was a lucky run:

Code: Select all

FEN: 8/8/4kpp1/3p1b2/p6P/2B5/6P1/6K1 b - - 0 1
St-mod-16-12-02:
Found 510 tablebases
 50/56	02:12	 173.020k	1.306k	+2,26	a4-a3 Kg1-f2 Bf5-c2 Kf2-e2 Bc2-e4 g2-g3 Ke6-f5 Bc3-d4 Kf5-g4 Bd4xf6 Kg4xg3 Ke2-d2 Kg3-f4 Bf6-d4 a3-a2 Bd4-f6 Kf4-g4 Kd2-e3 Kg4-f5 Bf6-c3 Be4-h1 Ke3-d2 Bh1-f3 Kd2-c2 Kf5-f4 Kc2-d3 Bf3-g2 Kd3-d2 Kf4-g3 Bc3-f6 Kg3-g4 Kd2-c1 Bg2-e4 Kc1-d2 Kg4-h3 Kd2-e2 Kh3-g3 Ke2-d2 Kg3-g2 Kd2-e3 Be4-c2 Ke3-d2 Bc2-b1 Bf6-a1 Kg2-h3 Ba1-f6 Kh3-g4 Kd2-c1 Kg4-f4 Kc1-d2 Bb1-f5 Bf6-d4 Kf4-g4 Bd4-f6 Kg4-f3 Bf6-a1
 51/56	02:17	 182.220k	1.323k	+2,26	a4-a3 Kg1-f2 Bf5-c2 Kf2-e2 Bc2-e4 g2-g3 Ke6-f5 Bc3-d4 Kf5-g4 Bd4xf6 Kg4xg3 Ke2-d2 Kg3-f4 Bf6-d4 a3-a2 Bd4-f6 Kf4-g4 Kd2-e3 Kg4-f5 Bf6-c3 Be4-h1 Ke3-d2 Bh1-f3 Kd2-c2 Kf5-f4 Kc2-d3 Bf3-g2 Kd3-d2 Kf4-g3 Bc3-f6 Kg3-g4 Kd2-c1 Bg2-e4 Kc1-d2 Kg4-h3 Kd2-e2 Kh3-g3 Ke2-d2 Be4-b1 Kd2-e3 Bb1-f5 Ke3-d2 Kg3-f3 Bf6-a1 Bf5-e4 Ba1-c3 Kf3-g2 Bc3-f6 Kg2-g3 Kd2-c1 Be4-b1 Kc1-d2 Kg3-g2 Bf6-a1
 52/57+	04:39	 436.623k	1.563k	+2,33	Bf5-h3
 52/57+	05:20	 538.195k	1.680k	+2,41	Bf5-h3
 52/57	10:53	 1.378.699k	2.109k	+3,53	Bf5-h3

but stockfish found it in 4:39 min on a amd laptop with 4 threads 1024 mb hash and 6-piece syzygy.

@Ludmil: As you can see it is a mate in 4 or 5 moves

Kind regards
Bernhard

i think its just lucky, but its difficult to know for sure; hash size, TB probe depth + limit, SSD/HDD are all variables that seem so flexible (minus SSD being superior to HDD ofc) the engine authors themselves only really seem to be able to guess at optimal settings.

re Bh3, it is decisively best & thus perfectly valid as a test position, humans & machines alike agree on this, end of story there.

[Lyudmil Tsvetkov]

most convincing black win however comes after Be4 Kf2 f5 g3 f4:

[d]8/8/4k1p1/3p4/p3bp1P/2B3P1/5K2/8 w - - 0 4

and gxf4 a3 and now White can even give back the f-Pawn with f5+ and no matter how Black takes it, with King, Pawn or Bishop, Game is over and drawn.

gf4 Kf5, with nice winning chances. (do not look at what engines suggest)

the objective outcome of the game migth indeed be a draw (which I am still not certain), but, out of curiosity, I ran a 10 games shootout with Komodo 10.1, and black won 3 games, with 7 draws. Maybe drawing chances are better after all, but any guess what that means in terms of winning probability, game stats and tuning of engines in terms of large number of random games?

games were at 1 min. bullet at 4 cores, and black won most of the games after an f5-f4 push, which unfortunately Komodo delays a lot, until the 50-moves come closer.

[Lyudmil Tsvetkov]

interesting, that Komodo followed the same plan: Be4, f5, but then delays f4.