EG studies for engines

[Nordlandia]

Difficult EG studies:

[d]k7/6K1/1P6/Bp1p3p/3pbp1P/8/3P1P2/8 w - - 0 0

[d]3q4/5kpp/pr1P3Q/p3p3/b1p5/6P1/3PR3/5BK1 w - - 0 0

[d]1N1R4/p5pk/PPPr3p/Q2b4/3b4/1q6/1N3PPP/4K2R w K - 0 0

[d]8/p2kp2p/Pp4pP/rPp1ppP1/2P2P2/2PKP3/Q7/3n4 w - - 0 0

Anti Houdini puzzle

[d]2RN1q2/3pp1p1/5p1p/1pp4k/5P1P/1B6/6K1/8 w - - 0 1

Oleg Pervakov

[d]8/b3p3/4p3/n1r5/2P1N1k1/1q4p1/3BRP2/1BR2K2 w - - 0 0

Draw or Fortress?

[d]8/4p3/4Rb2/n5k1/8/6P1/2B3K1/8 w - - 0 10

White wins

[d]q6B/r3p3/pRP5/P1R2p2/k4K2/8/1P6/N4b2 w - - 0 0

White wins

[d]8/2N2R1p/5nk1/N3p3/1n1P4/7K/7b/8 w - - 0 0

[Vinvin]

Difficult EG studies:

Please, gives the solutions with the studies.

[Nordlandia]

Part II

[d]4b3/1PK5/P7/k1B3p1/4B1qp/8/P7/8 w - - 0 0

[d]4NN2/8/2r5/1q2b2k/2pp1p1P/5P2/1PPP4/1K1Q1B2 w - - 0 0

[d]3k2b1/p6p/3K2B1/1pp1pN2/8/4N3/2P5/2n5 w - - 0 0

Semi-Difficult

[d]3b1K2/2pr4/1Pp5/Rp1k4/8/8/2P5/8 w - - 0 0

R. Becker | Special Price | White wins

[d]5q1n/p3p2K/p4p2/p2B4/8/5p2/3Q4/5k2 w - - 0 0

5kBN/Br6/8/7r/6R1/8/8/5K2 w - - 0 0

White wins

[d]8/6Pp/7p/3p4/5K1P/6R1/kP2rp1P/8 w - - 0 0

3rd price (bronce medal) | J. Timman | Draw | 8/7

[d]8/3p1p2/3p1P2/2P1P2K/r2Pp1P1/6P1/6b1/1k4B1 w - - 0 0

1. Kc8 | White wins

[d]1K6/pP3B2/1kp5/p7/5p2/1P1P2P1/pB6/4rrq1 w - - 0 0

1. Bb7+ | White wins | R. Becker | 3rd Honourable Mention

[d]6r1/r3p3/k1B1p3/3nR3/1PRK1NQ1/8/2pN1n2/2q1b3 w - - 0 0

1. Kc5 | Draw | D. Hlebec

[d]2k5/8/8/1Q4K1/p3B3/1p4p1/b6q/8 w - - 0 0

White wins

[d]8/7p/5R2/2k3p1/2Pp2Pb/8/4Kp2/8 w - - 0 0

Funny one

[d]kN5B/2PPPPR1/B1PPPP2/1K6/3q4/8/8/r7 w - - 0 0

[peter]

Hi Jon!

Draw or Fortress?
[d]8/4p3/4Rb2/n5k1/8/6P1/2B3K1/8 w - - 0 10

Neither nor, I'd say.

After some Forward- Backward:

[d]8/4p3/4Rb2/n5k1/8/6P1/2B3K1/8 w - - 0 1

Analysis by asmFishW_2017-10-25_popcnt:

10.Bd3 Nb7 11.Rc6 Nd8 12.Rc5+ Kh6 13.Kf3 Ne6 14.Rd5 Nc7 15.Rd7 Ne6 16.Kg4 Nf8 17.Rd5 Ne6 18.Rb5 Ng5 19.Rb8 Nf7 20.Bc2 Kg7 21.Kh5 Be5 22.Rb6 Bf6 23.Rb5 Nd6 24.Rb8 Be5 25.g4 Bf4 26.Ra8 Be3 27.Ra6 Kf6 28.Ra3 Bg1 29.Ra1 Bd4 30.g5+ Ke6 31.Ra8 Kd5 32.Ra5+ Kc6 33.Ra4 Be5 34.Ra8 Kd5 35.Ra7 Bg7 36.Rxe7 Bf8 37.Rc7 Nb5 38.Rd7+ Ke6 39.Bf5+ Ke5 40.Rf7 Ba3 41.g6
+- (114.11) Depth: 42/78 00:03:31 7867MN, tb=23137503

Postion after output- move nr.37 is #24 according to Lomonosovs

[Eelco de Groot]

Difficult EG studies:

[d]k7/6K1/1P6/Bp1p3p/3pbp1P/8/3P1P2/8 w - - 0 0

Is #1 winnable? Kaissa likes the White position, but so far not with a really high score. The bonus for the passed pawn is high more than +2. If score is more than a full pawn there is no scaling down in the code for the opposite bishops.

Code: Select all

    // If we don't already have an unusual scale factor, check for certain
    // types of endgames, and use a lower scale for those.
    if (sf == SCALE_FACTOR_NORMAL || sf == SCALE_FACTOR_ONEPAWN)
    {
        if (     pos.opposite_bishops()
			   && !(more_than_one(pe->passed_pawns(strongSide)))
		      && !(abs((eg - VALUE_DRAW)/ PawnValueEg) > 1))

But it would be nice if then it could actually win this position


k7/6K1/1P6/Bp1p3p/3pbp1P/8/3P1P2/8 w - -

Engine: Kaissa HT (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

32/43 0:01 +1.41 1.Kh6 Kb7 2.Kxh5 f3 3.Bb4 Kxb6 4.Kg5 Kc6
5.Kf6 Bc2 6.h5 Be4 7.Kg7 d3 8.h6 d4
9.h7 Bxh7 10.Kxh7 Kd5 11.Kg6 Kc4
12.Bd6 Kb3 13.Kf5 Kc2 14.Bb4 (26.932.821) 17443

33/44 0:01 +1.46 1.Kh6 Kb7 2.Kxh5 f3 3.Bb4 Kxb6 4.Kg5 Kc7
5.Kf6 Bd3 6.h5 Kc8 7.Ke6 Be4 8.h6 Kc7
9.Kf6 Kd7 10.Ke5 Kc7 11.Kxd4 Kc6
12.Ke5 Kd7 13.d4 Kc6 14.Kf6 (32.257.515) 17417

34/51 0:02 +1.67 1.Kh6 Kb7 2.Kxh5 f3 3.Bb4 Kxb6 4.Kg5 Kc7
5.Kf6 Bd3 6.h5 Be4 7.h6 Kd7 8.Ba3 Bc2
9.Be7 Bb1 10.Bc5 Kc6 11.Bb4 Kd7
12.Ke5 Be4 13.Bc5 Kc6 14.Kxd4 (40.901.027) 17629

35/48 0:02 +1.58 1.Kh6 Kb7 2.Kxh5 f3 3.Bb4 Kxb6 4.Kg5 Kc7
5.Kf6 Bd3 6.h5 Be4 7.h6 Kd7 8.Ba3 Bc2
9.Bc5 Kc6 10.Bf8 Kd7 11.Ba3 Be4
12.Bb4 Bc2 13.Kg7 Ke6 14.h7 (42.592.271) 17665

36/56 0:03 +1.73++ 1.Kh6 (54.267.836) 17945

36/56 0:03 +1.85++ 1.Kh6 (61.800.202) 18080

36/62 0:03 +1.95 1.Kh6 Kb7 2.Kxh5 f3 3.Bb4 Kxb6 4.Kg5 Kc7
5.Kf6 Kc6 6.h5 d3 7.h6 d4 8.Ke5 Bg6
9.Kxd4 Kd7 10.Ke5 Kc6 11.Ke6 Bh7
12.Bd6 Bg8+ 13.Ke5 Bh7 14.Ba3 (68.923.004) 18175

37/47 0:03 +1.95 1.Kh6 Kb7 2.Kxh5 f3 3.Bb4 Kxb6 4.Kg5 Kc7
5.Kf6 Kc6 6.h5 d3 7.h6 d4 8.Ke5 Bg6
9.Kxd4 Kd7 10.Ke5 Kc6 11.Ke6 Bh7
12.Bd6 Bg8+ 13.Ke5 Bh7 14.Ba3 (69.490.096) 18181

38/57 0:04 +2.02++ 1.Kh6 (80.928.429) 18326

38/57 0:04 +2.10++ 1.Kh6 (91.159.464) 18419

38/62 0:05 +2.21++ 1.Kh6 (105.672.505) 18532

38/62 0:06 +2.37++ 1.Kh6 (123.685.608) 18585

38/62 0:07 +2.41 1.Kh6 Bf3 2.Kg5 Be2 3.Kxf4 Kb7 4.Ke5 Bf3
5.Kxd4 b4 6.Bxb4 Kxb6 7.Ke3 Bg2 8.f4 Kb5
9.Ba3 Kc6 10.f5 Kd7 11.Kf4 Ke8
12.Kg5 Be4 13.Kf6 d4 14.Ke5 (142.679.608) 18660

39/51 0:07 +2.41 1.Kh6 Bf3 2.Kg5 Be2 3.Kxf4 Kb7 4.Ke5 Bf3
5.Kxd4 b4 6.Bxb4 Kxb6 7.Ke3 Bg2 8.f4 Kc6
9.f5 Kd7 10.Kf4 Ke8 11.Kg5 Be4
12.Kf6 d4 13.Ke5 Bc2 14.Bc5 (143.044.979) 18657

40/54 0:07 +2.41 1.Kh6 Bf3 2.Kg5 Be2 3.Kxf4 Kb7 4.Ke5 Bf3
5.Kxd4 b4 6.Bxb4 Kxb6 7.Ke3 Bg2 8.f4 Kc6
9.f5 Kd7 10.Kf4 Ke8 11.Kg5 Be4
12.Kf6 d4 13.Ke5 Bc2 14.Bc5 (144.355.471) 18652

41/56 0:07 +2.34-- 1.Kh6 Bf3 (146.299.610) 18651

41/56 0:07 +2.41++ 1.Kh6 (148.775.737) 18645

41/56 0:08 +2.52++ 1.Kh6 (160.714.602) 18633

41/56 0:10 +2.68++ 1.Kh6 (187.602.973) 18687

41/72 0:11 +2.90++ 1.Kh6 (221.382.065) 18718

41/72 0:13 +3.19++ 1.Kh6 (256.702.907) 18756

41/72 0:15 +3.19 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bg2 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc7 7.f5 Kd7 8.d3 Bf1
9.f6 Ke8 10.Ke6 Bh3+ 11.Kxd5 Kf7
12.Be7 Bg2+ 13.Kxd4 Ke6 14.Ke3 (281.704.691) 18729
.
.
.
68/89 2:55 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (3.590.944.252) 20415

69/90 28:32 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.259.382.428) 31104

70/93 28:32 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.267.605.331) 31104

71/93 28:32 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.275.749.238) 31104

72/93 28:33 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.283.917.668) 31104

73/96 28:33 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.294.408.536) 31104

74/95 28:33 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.303.629.661) 31103

75/95 28:34 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.314.782.107) 31103

76/98 28:34 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.327.276.483) 31103

77/99 28:34 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.340.722.574) 31103

78/100 28:35 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.355.482.690) 31104

79/103 28:40 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (53.501.380.914) 31103

80/103 39:27 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (67.516.330.297) 28518

81/103 39:35 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (67.621.863.882) 28470

82/111 39:48 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (67.803.283.087) 28392

83/110 41:12 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (69.049.832.060) 27928

84/109 85:46 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (110.568.401.992) 21482

85/107 92:44 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (116.181.525.737) 20878

[Eelco de Groot]

(continued from above #1) Ancalagon (not sure which version precisely this is) also could not find anything better than 1. Kh6 (best five moves Multi PV)


[d]k7/6K1/1P6/Bp1p3p/3pbp1P/8/3P1P2/8 w - -

Engine: Ancalagon HT (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G.

46 110:38 +2.83 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.Kf6 Bg4
9.Kg6 Ke8 10.d3 Bh3 11.f6 Be6 12.Kg7 Bf7
13.Bc5 Be6 14.Bxd4 (43.039.952.743) 6483

45 110:38 +0.53 1.Kf6 Kb7 2.Kg5 b4 3.Bxb4 Kxb6
4.Kxf4 Bh1 5.Ke5 Bf3 6.Kxd4 Kc6
7.Ke3 Bg2 8.f4 Be4 9.d3 Bg6 10.Kd4 Bf5
11.Ke5 Bxd3 12.Bc3 Kd7 13.Kxd5 Ke7
14.Ke5 (43.039.952.743) 6483

45 110:38 +0.51 1.Kf7 Kb7 2.Kf6 b4 3.Bxb4 Kxb6 4.Kg5 Kc6
5.Kxh5 f3 6.Kg5 Kb5 7.Be7 Kc4 8.h5 Kd3
9.Bf6 Kxd2 10.Bxd4 Ke2 11.Be3 Kd3
12.Bb6 Ke2 13.h6 Kd3 14.Kf4 (43.039.952.743) 6483

45 110:38 +0.37 1.Bb4 Kb7 2.Bc5 Bf3 3.Kf6 b4 4.Bxd4 b3
5.Ke5 b2 6.Bxb2 Kxb6 7.Kxf4 Bg2
8.Bd4+ Kc7 9.Kf5 Bf3 10.d3 Kc6
11.Ke5 Kd7 12.Kf6 Kc6 13.Bc3 Kd6
14.Bb2 (43.039.952.743) 6483

45 110:38 0.00 1.Kf8 Kb7 2.Ke7 b4 3.Bxb4 Kxb6 4.Bd6 Kb5
5.Bxf4 Kc4 6.Kf6 Bf3 7.Kg5 Kd3 8.Kf6 Kc4 (43.039.952.743) 6483

best move: Kg7-h6 time: 110:38.672 min n/s: 6.483.223 nodes: 43.039.952.743

And after some more time given to Kaissa:

84/109 85:46 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (110.568.401.992) 21482

85/107 92:44 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (116.181.525.737) 20878

86/107 99:36 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (122.231.446.484) 20452

87/107 155:53 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 (172.528.604.277) 18445

88/109 359:40 +3.57 1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4
5.Bxb4 Kxb6 6.f4 Kc6 7.f5 Kd7 8.d3 Ke8
9.Ke6 Be2 10.f6 Bg4+ 11.Kxd5 Kf7
12.Be7 Be2 13.Kxd4 Ke6 14.Ke4 {[d]8/4B3/4kP2/7p/4K2P/3P4/4b3/8 b - -} (337.434.084.497) 15635

The final position of the PV in Shredder GUI which only gives the first 14 moves, looks optically good with two passed pawns, but it's just a fortress:

[d]8/4B3/4kP2/7p/4K2P/3P4/4b3/8 b - -

Engine: Kaissa HT (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

38 2:25 +3.06 1...Bd1 2.d4 Bc2+ 3.Kf4 Bg6 4.Kg5 Bf7
5.Bd8 Kd6 6.Ba5 Ke6 7.Bd2 Be8 8.Bc3 Bf7
9.Ba5 Kd6 10.Bd2 Ke6 11.Bc3 Kd6
12.Kh6 Kd7 13.Bd2 Ke8 14.Kg5 Kd8 (1.096.254.688) 7519

38 2:25 +3.06 1...Bf1 2.d4 Bc4 3.Kf4 Kd7 4.Ke5 Kc6
5.Bb4 Bb3 6.Bc5 Bc4 7.Kf4 Kd7 8.Kg5 Bf7
9.Kh6 Ke8 10.Kg7 Bd5 11.Kg6 Bf7+
12.Kh6 Kd7 13.Kg7 Ke8 14.Ba7 Bd5 (1.096.254.688) 7519

38 2:25 +63.07 1...Bg4 {this is lost for Black} 2.Kf4 (1.096.254.688) 7519

best move: Be2-d1 time: 2:25.813 min n/s: 7.519.615 nodes: 1.096.254.688

Black is holding this? So my rule is here not good enough I think, two passed pawns, sf (scalefactor) switched off automatically is not enough (here) and a single pawn that has a bonus of 3 where black has an extra but doubled pawn, is also not enough to win here. Unless I'm missing something. Two passed pawns is reasonably safe (measured in Fishtest) but a single passed pawn advantage (bonus) of +2 full pawns to switch off scaling down opposite color bishop may cost some Elo in my measurements. Not a lot and it's within the error margins. But maybe I should increase to 3 full pawns for a single passer, to switch to normal scaling..

[Paloma]

For #9 i get this

1.Re7 Bg1
2.dxe5 Ne4
3.Kg2

there is probably no clear winning,
just a change in front

[peter]

Hi Eelco!

Is #1 winnable?

I think it is, but I'm not sure, if it is without help from Black.

I have a 30'+15" game won by SugaR with switched off nullmove against asmFish:

[pgn]
[Event "Nr.1 CCC"]
[Site "?"]
[Date "2017.10.28"]
[Round "?"]
[White "DEGS, Jon"]
[Black "Winnable?, Eelco"]
[Result "*"]
[SetUp "1"]
[FEN "k7/6K1/1P6/Bp1p3p/3pbp1P/8/3P1P2/8 w - - 0 1"]
[PlyCount "52"]

1. Kh6 Bf3 2. Kg5 Kb7 3. Kxf4 Bd1 4. Ke5 b4 5. Bxb4 Kxb6 6. f4 Kc6 7. Kxd4 Bc2
8. Ke5 Bh7 9. Bc3 Bg8 10. Kf6 Bh7 11. Ke7 Bf5 12. Be5 Bg4 13. Kf6 Kd7 14. d4
Bf3 15. f5 Be4 16. Bf4 Bb1 17. Bc1 Kd6 18. Ba3+ Kd7 19. Ke5 Bc2 20. Bd6 Bb1 21.
f6 Bg6 22. Kxd5 Bf7+ 23. Kc5 Ke6 24. Be7 Bg6 25. d5+ Kd7 26. Kd4 Be8 *
[/pgn]

Adjudicated, SugaR had the tbs- Bingo- eval of 132+ already after move 15., asmFish a few moves later on too.
SugaR- output at position after 15th move of White:

[d]8/3k4/5K2/3pBP1p/3P3P/5b2/8/8 b - - 0 1

Analysis by S_XPrO 241017 x64 POPCNT:

15...Be4 16.Bb8 Bd3 17.Ba7 Bb1 18.Bc5 Bd3 19.Ke5 Bb1 20.Bd6 Bd3 21.f6 Bg6 22.Kxd5 Bf7+ 23.Kc5 Be8 24.d5 Bg6 25.Bf8 Bf5 26.Kd4 Bh7 27.Ke5 Bg8 28.Be7 Bf7 29.d6 Bg6 30.Kf4 Ke6 31.Kg5 Be8 32.Kh6 Kf7 33.Kxh5 Bc6 34.Kg5 Bb5 35.h5 Bd7 36.h6 Ke6 37.f7 Ba4 38.f8Q Kd5 39.Qa8+ Bc6 40.Qxc6+
+- (132.58) Depth: 42/53 00:04:08 10082MN, tb=9818116

But I'm not able to take the high score back to root position with default Fish.
Trying now with nullmove- less one, maybe that could be better because of necessary triangle- und body- check- maneuvers, couldn't it

[peter]

But I'm not able to take the high score back to root position with default Fish.

Switching off tbs brings more than nullmove.

Went backward with H6 default from the end of the game above, taking back the mating- score from there to 12...Bg4 (?!), from there 12...Bh7 instead takes some moves forward to raise score to 8.00 again.

k7/6K1/1P6/Bp1p3p/3pbp1P/8/3P1P2/8 w - - 0 1

1. Kh6 Bf3 2. Kg5 Kb7 3. Kxf4 Bd1 4. Ke5 b4 5. Bxb4 Kxb6 6. f4 Kc6 7. Kxd4 Bc2 8. Ke5 Bh7 9. Bc3 Bg8 10. Kf6 Bh7 11. Ke7 Bf5 12. Be5 Bh7 13. d4 Bf5 14. Kf6 Bh7 15. Kg5 *

Going this backward again to starting position doesn't show any more eval- jumps and leads to this output with full hash from Backward:

[d]k7/6K1/1P6/Bp1p3p/3pbp1P/8/3P1P2/8 w - - 0 1

Analysis by Houdini 6.02 Pro x64-popc:

1.Kh6 Bf3 2.Kg5 Kb7 3.Kxf4 Bd1 4.Ke5 b4 5.Bxb4 Kxb6 6.f4 Kc6 7.Kxd4 Bc2 8.Ke5 Bh7 9.Bc3 Bg8 10.Kf6 Bh7 11.Ke6 Bg8+ 12.Ke7 Bh7 13.Bd4 Bf5 14.Be5 Bh7 15.d4 Bf5 16.Bd6 Bh3 17.Bb8 Bf5 18.Be5 Bh7 19.Ke6 Bg8+ 20.Kf5 Bh7+ 21.Kg5 Kd7 22.f5 Ke7 23.f6+ Kf7 24.Kxh5 Bd3 25.Kg5 Bh7 26.Bc7 Be4 27.Ba5 Bd3 28.Be1 Bb1 29.h5 Ke6 30.Bg3
+- (6.70) Depth: 42/95 00:04:24 13573MN

[Vinvin]

I put the 2 parts together and ran it under McBrain 2.7, Stockfish_171011 and Sting-sf-89.
Condition : 10 minutes on a i7-6*4GHz. 6 Pieces Syzygy.
Please tell me who is wrong or right. I hold the log if you want to see the eval and timing.

Code: Select all

       McBrain-27         Stockfish_171011   Sting-sf-89
 1)    Kg7-h6                Kg7-h6           Kg7-h6     
 2)    Re2-f2                Re2-f2           Re2-f2     
 3)    O-O                   O-O              O-O        
 4)    Qa2-d2                Qa2-d2           Qa2-d2     
 5)    Kg2-h3                Kg2-h3           Kg2-h3     
 6)    Bb1-c2                Bb1-c2           Bb1-c2     
 7)    Kg2-h3                Kg2-h3           Kg2-h3     
 8)    Bh8-c3                Bh8-e5           Bh8-c3     
 9)    Rf7-e7                Rf7-e7           Rf7-e7     
10)    Bc5-b4                Bc5-b4           Bc5-b4     
11)    Bf1-d3                Bf1-d3           Bf1-d3     
12)    Bg6-h5                Bg6-h5           Bg6-h5     
13)    b6-b7                 b6-b7            b6-b7      
14)    Bd5-c4                Qd2-d1           Qd2-d1     
15)    Ba7-d4                Ba7-d4           Ba7-d4     
16)    Rg3-a3                Rg3-a3           Rg3-a3     
17)    c5-c6                 c5-c6            c5xd6      
18)    Kb8-c8                Kb8-c8           Kb8-c8     
19)    Kd4-c5                Kd4-c5           Kd4-c5     
20)    Qb5-c6                Qb5-c6           Be4-b7     
21)    Ke2-d3                Ke2-d3           Ke2-d3     
22)    Ba6-b7                Ba6-b7           Ba6-b7     

22 positions :

Code: Select all

k7/6K1/1P6/Bp1p3p/3pbp1P/8/3P1P2/8 w - - 0 0
3q4/5kpp/pr1P3Q/p3p3/b1p5/6P1/3PR3/5BK1 w - - 0 0
1N1R4/p5pk/PPPr3p/Q2b4/3b4/1q6/1N3PPP/4K2R w K - 0 0
8/p2kp2p/Pp4pP/rPp1ppP1/2P2P2/2PKP3/Q7/3n4 w - - 0 0
2RN1q2/3pp1p1/5p1p/1pp4k/5P1P/1B6/6K1/8 w - - 0 1
8/b3p3/4p3/n1r5/2P1N1k1/1q4p1/3BRP2/1BR2K2 w - - 0 0
8/4p3/4Rb2/n5k1/8/6P1/2B3K1/8 w - - 0 10
q6B/r3p3/pRP5/P1R2p2/k4K2/8/1P6/N4b2 w - - 0 0
8/2N2R1p/5nk1/N3p3/1n1P4/7K/7b/8 w - - 0 0
4b3/1PK5/P7/k1B3p1/4B1qp/8/P7/8 w - - 0 0
4NN2/8/2r5/1q2b2k/2pp1p1P/5P2/1PPP4/1K1Q1B2 w - - 0 0
3k2b1/p6p/3K2B1/1pp1pN2/8/4N3/2P5/2n5 w - - 0 0
3b1K2/2pr4/1Pp5/Rp1k4/8/8/2P5/8 w - - 0 0
5q1n/p3p2K/p4p2/p2B4/8/5p2/3Q4/5k2 w - - 0 0
5kBN/Br6/8/7r/6R1/8/8/5K2 w - - 0 0
8/6Pp/7p/3p4/5K1P/6R1/kP2rp1P/8 w - - 0 0
8/3p1p2/3p1P2/2P1P2K/r2Pp1P1/6P1/6b1/1k4B1 w - - 0 0
1K6/pP3B2/1kp5/p7/5p2/1P1P2P1/pB6/4rrq1 w - - 0 0
6r1/r3p3/k1B1p3/3nR3/1PRK1NQ1/8/2pN1n2/2q1b3 w - - 0 0
2k5/8/8/1Q4K1/p3B3/1p4p1/b6q/8 w - - 0 0
8/7p/5R2/2k3p1/2Pp2Pb/8/4Kp2/8 w - - 0 0
kN5B/2PPPPR1/B1PPPP2/1K6/3q4/8/8/r7 w - - 0 0