Elephantiasis: reality or myth?

[Lyudmil Tsvetkov]

As I initiated the thread about imbalances, but Harm contributed with a major swing, a bit unrealistic, true, but very interesting, I decided to post a continuation of the topic in a new thread. I know that when threads start getting elephantine in size, most users simply ignore them and even the persons that take active part in the discussion, have a hard time finding where and what had been posted recently. I hope the moderators, as well as Harm and the other contributors and readers will not object to such a step.

I think this thread could again focus not only on elephantiasis, but on imbalances in general.

[Lyudmil Tsvetkov]

And my first contribution to this thread.

I played some more games in the 3Qs vs 7Bs imbalance, and really, I think it favours the bishops much more than the knights. It seems I am able to draw only against Queeny. 2 games, one where Q vs 4Bs of the same colour is a draw, but Queeny does not seem to recognize that. Hey, Harm, does Queeny really think that 4 same-colour bishops are still worth an elephantiasis bonus? Because I would think that those are really redundant pieces. The other game a draw by full-scale closing of the game.

[pgn][Event "Computer chess game"]
[Site "OWNER-PC"]
[Date "2013.10.28"]
[Round "?"]
[White "owner"]
[Black "QN"]
[Result "1/2-1/2"]
[BlackElo "2000"]
[Time "15:41:21"]
[WhiteElo "2400"]
[TimeControl "300"]
[SetUp "1"]
[FEN "bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1"]
[Termination "normal"]
[PlyCount "149"]
[WhiteType "human"]
[BlackType "human"]

1. c4 c5 {(c7c5 e2e3 g7g6 a1c1 e7e5 d2d3 h8g7 e3e4 b8d6 d1a4 d8b6 f2f3)
+4.94/14 12} 2. e4 g6 {(g7g6 a1c1 h8d4 e1f1 f8g7 b2b3 e7e5 c1a3 b7b6 f2f3
a8b7 d2d3) +5.44/15 8} 3. Qac1 b6 {(b7b6 f2f3 h8d4 a2a3 a8b7 b2b4 e7e5 b4b5
b8d6 d1a4 d8h4 e1e2) +5.54/15 9} 4. d3 Bab7 {(a8b7 h2h4 f8g7 b2b3 g7b2 c1b1
h8d4 h4h5 b8e5 e1e2 b2c3 h5h6) +5.66/15 8} 5. f3 Bd4 {(h8d4 d1c2 f8g7 b2b3
e7e5 h2h4 b8d6 h4h5 d8e7 h5h6 g7f6 c1a3) +5.79/15 7} 6. b3 e5 {(e7e5 h2h4
f7f5 h4h5 g8f7 h5g6 h7g6 g2g3 d7d6 f3f4 b8c7 c1a3 g6g5 f4g5 d8g5 a3a7)
+6.05/15 5} 7. g4 Bh4+ {(d8h4 e1f1 f7f6 c1a3 g8e6 d1e2 f8d6 e2g2 h4g5 h2h4
g5f4 h4h5 a7a6 h5g6 h7g6 a3a4) +6.28/17 6} 8. Kf1 f6 {(f7f6 c1a3 g8e6 d1c2
a7a6 c2g2 b8c7 f1e2 f8d6 a3c1 h4g5 c1e1 b6b5 g2h3 h7h6) +6.35/18 6} 9. Qg2
Be6 {(g8e6 c1a3 h4g5 g2h1 f8e7 h2h4 g5e3 h4h5 g6g5 h5h6 d7d6 h1h5 e8f8 a3a4
c8d7 a4a3 b6b5 c4b5 d7b5) +6.38/17 5} 10. a4 Bc7 {(b8c7 a4a5 h4g5 c1a3 f8e7
a5b6 a7b6 f1e1 g5f4 d1c2 d7d6 g2h3 h7h5 a3a4 c8d7) +6.59/16 6} 11. Qde2 d6
{(d7d6 a4a5 h4g5 c1d1 f8e7) +6.68/16 7} 12. Qed1 Bg5 {(h4g5 c1b1 c8d7 a4a5
b6a5 d1e1 f8e7 g2h1 a5a4 b3a4 b7c6 a4a5 g5f4) +6.82/15 3} 13. Qcc2 Bcd7
{(c8d7 c2b1 g5f4 a4a5 f8e7 a5b6 a7b6 b3b4 h7h5 h2h3 h5g4 h3g4 f6f5)
+6.90/14 3} 14. Qa2 Be7 {(f8e7 h2h3 a7a5 d1e2 g5f4 a2a3 f6f5 a3a2) +6.99/14
6} 15. Qac2 f5 {(f6f5 g4f5 g6f5 a4a5 b6a5 c2a2 e8d8 g2h1 f5e4 f3e4 a5a4
b3a4 g5f4) +7.10/14 4} 16. h3 fxg4 {(f5g4 h3g4 h7h5 g2h1 h5h4 d1b1 g5f4
a4a5 g6g5 a5b6 a7b6) +7.76/15 4} 17. fxg4 Bbc8 {(b7c8 g2h1 h7h5 d1b1 g5e3
c2h2 h5g4 h3g4 e6g4 h2h8 e7f8 h8f6 g6g5 h1h8 g4h3 f1e1) +8.01/16 5} 18.
Qce2 h5 {(h7h5 g4h5 e6h3 h5g6 h3g2 f1g2 e8f8 a4a5 c8b7 a5b6 a7b6 e2h5 e7f6)
+9.30/16 5} 19. Qef3 hxg4 {(h5g4 h3g4 e7f6 g2h1 e6g4 h1h7 g4f3 d1f3 g5f4
h7g6 e8e7 g6h7 e7f8 a4a5 f6g7 a5b6 a7b6) +11.12/16 3} 20. hxg4 Bxg4 {(e6g4
f3g4 d7g4 g2g4 c8g4 d1g4 e8f8 g4d7 c7d8 d7a7 g5f4 a4a5 b6a5) +11.09/16 5}
21. Qgxg4 Bxg4 {(d7g4 f3g4 c8g4 d1g4 e8f8 g4d7 c7d8 a4a5 b6a5 d7a7 g5f4
a7b8 g6g5 b8c8 f4e3 c8f5 f8g8 f5g6 g8h8 g6f7) +11.13/18 3} 22. Qxg4 Bxg4
{(c8g4 d1g4 e8f8 g4d7 c7d8 a4a5 b6a5 d7a7 g5f4 f1g2 g6g5 g2h3 f4e3 a7b8
f8f7 b8a7 e3f4 a7b8 f7f6 b8a7 f4e3) +11.15/21 4} 23. Qxg4 Kf8 {(e8f8 g4d7
c7d8 a4a5 b6a5 d7a7 g5f4 a7b7 g6g5 b7d7 f8f7) +11.21/21 6} 24. Kg2 a5
{(a7a5 g4c8 c7d8 c8e6 f8g7 g2g3 g5f4 g3g4 f4e3 g4g3 g6g5 g3g4 e3f4 g4h5
g7f8 e6f5 f4e3 g6e6) +11.71/20 3} 25. Kh3 Kf7 {(f8f7 g4c8 c7d8 c8b7 g5f4
b7d5 f7g7 h3g4 f4d2 d5c6 d2e3 c6e8 e3f4 e8d7 g7f7 g4f3 f7f6 d7h3 f6g5)
+11.70/22 3} 26. Qd7 Bd8 {(c7d8 d7b7 g5f4 b7d5 f7f8 h3g4 f4g5 d5c6 g5e3
c6d5 e3f4 d5c6 g6g5 g4f5 f4d2 c6d5 d2c3 d5e6 f8e8 f5g6) +11.71/23 5} 27.
Kg4 Bf4 {(g5f4 d7c6 f4e3 c6d5 f7g7 d5c6 g7f6 c6e8) +11.75/21 4} 28. Qc8
Bfe3 {(f4e3 c8c6 f7e6 c6c8 e6f6 c8c6 f6f7) +11.74/21 3} 29. Qd7 Bg5 {(e3g5
d7c6 f7f6 c6e8 f6g7 e8b5 g5f4 b5e8 f4e3 e8d7) +11.68/20 5} 30. Qc8 Bf4
{(g5f4 c8c6 f7g7 c6a8 f4e3 a8c6) +11.71/20 2} 31. Qd7 Bfe3 {(f4e3 d7c6 f7f6
c6e8 f6g7) +11.63/22 7} 32. Qc8 Kf6 {(f7f6 c8b7 e3f4 b7d5 f6g7 d5c6 f4e3)
+11.70/21 2} 33. Qd7 Bf4 {(e3f4 d7e8 f6g7 e8b5 f4e3 b5c6 g7f6) +11.70/22 3}
34. Qc8 Bd2 {(f4d2 c8b7 d2g5 b7d5 f6g7 d5c6 g5f4 c6d7 f4e3 d7e8 e3f4)
+11.70/21 2} 35. Qd7 Bb4 {(d2b4 d7e8 f6g7 g4g3 b4d2 e8c6 d2f4 g3g4 f4e3
c6e8 h6f4) +11.71/21 2} 36. Qc8 Bbc3 {(b4c3 c8c6 d4e3 c6e8 f6g7 g4f3 c3d4)
+11.73/21 2} 37. Qd7 Bd2 {(c3d2 d7e8 f6g7 e8d7 d2e3 d7e8 e3f4) +11.69/21 3}
38. Qc8 Bf4 {(d2f4 c8d7 f4e3) +11.71/21 2} 39. Qd7 Bde3 {(d4e3 d7e8 f6g7
e8d7) +11.68/22 3} 40. Qc8 Kg7 {(f6g7 c8d7 e3d4 d7e8 f4e3 e8d7) +11.74/21
1} 41. Qd7 Kf7 {(g7f7 d7c6 f7e6 c6c8 e6f6) +11.75/21 2} 42. Qc8 Kf6 {(f7f6
c8e6 g3f2 f5g6 f8e8 e6c8 f2h4 c8c6 e8f8 c6c8) +11.71/18 1} 43. Qd7 Bg1
{(e3g1 d7e8 f6g7 e8c6 g1d4 c6d7) +11.70/20 1} 44. Qc8 Bg5 {(f4g5 g4h3 g1d4
c8d7 g5c1 h3g4 c1e3 d7e8 f6g7) +11.68/20 1} 45. Qd7 Bd4 {(g1d4 d7c8 f6g7
c8b7 g5f4 b7d7) +11.73/21 1} 46. Qc8 Bc1 {(g5c1 c8c6 f6f7 c6d5 f7g7 d5b7
g7f8 b7c6 c1f4 c6d5 f4d2 e8f7 c6d5 f7g7 d5e6 d2f4 e6g6 g7f8 g6e6) +11.70/21
2} 47. Qd7 Bh6 {(c1h6 d7e8 f6g7 e8d7 h6f4 d7e8 f4e3) +11.71/20 2} 48. Qc8
Kg7 {(f6g7 c8b7 h6f4 b7c8 f4e3 c8c6) +11.67/19 1} 49. Qd7 Bhe3 {(h6e3 d7e6
e3f4 e6d7 f4g5 d7e8) +11.68/21 1} 50. Qc8 Bf4 {(e3f4 c8d7 f4g5 d7e8 g5f4)
+11.71/21 2} 51. Qd7 Bde3 {(d4e3) -M434/100 7} 52. Qc8 Bg1 {(e3g1)
+11.70/20 1} 53. Qd7 Bf2 {(g1f2 d7c8 f4d2 c8a6 f2d4 a6c8 g6g5 g4f5 d2b4
c8e6 g7f8 e6h6 f8e8 h6e6) +11.71/19 1} 54. Qc8 Bc1 {(f4c1 c8e6 f2d4 e6d7
c1f4) +11.71/19 1} 55. Qd7 Bd4 {(f2d4 d7b7 c1d2 b7c8 d2e3 c8c6 e3f4)
+11.66/21 1} 56. Qc8 Bd2 {(c1d2 c8b7 d2g5 b7c8) +11.72/20 1} 57. Qd7 Kf7
{(g7f7 d7b7 f7g7) +11.66/21 2} 58. Qc8 Bb4 {(d2b4 c8d7 b4c3 d7c8 d4e3 g4f3
e3f4 f3g4 f7f6) +11.70/19 5} 59. Qd7 Be1 {(b4e1 d7c8 e1d2 c8b7) +11.62/19
1} 60. Qc8 Bef2 {(e1f2 c8b7 f7f6 b7c8 f2e1) +11.67/18 1} 61. Qd7 Bc3 {(d4c3
d7c6 f2d4 c6d5 f7g7 d5e6 c3b4 e6d5 d4e3) +11.71/18} 62. Qc8 Bfd4 {(f2d4
c8b7 f7g7 b7c8 c3b4 c8e6 b4e1 e6d7) +11.66/19 1} 63. Qd7 Bb4 {(c3b4 d7c8
f7f6 c8d7 b4c3 d7e8 f6g7 e8d7) +11.62/20 3} 64. Qc8 Be3 {(d4e3 g4f3 b4d2
c8h3 g6g5 h3f5 f7e8 f3e2 e3f4 f5g4 d2c3 g4c8 c3d4 e2f3) +11.68/19 1} 65.
Qd7 Bf4 {(e3f4 g4f3 f7f6 d7c8 b4c3 f3g4 f4e3 c8b7 e3f4) +11.74/19} 66. Qc8
Kf6 {(f7f6 c8b7 f4e3 b7d5 f6g7 d5e6 e3d4 e6c8 g7f6) +11.71/20} 67. Qd7 Be3
{(f4e3 d7e8 f6g7 g4f3 e3f4 f3g4 b4c3) +11.67/19} 68. Qc8 Bc3 {(b4c3 c8b7
f6e6 b7c8 e6f7) +11.69/19 1} 69. Qd7 Bh6 {(e3h6 d7e8 h6g7 g4g3 c3d4 e8d7
g7h6 g3h3 h6g5 h3g3 f6g7 g3g4 g5f4) +11.68/20 1} 70. Qc8 Bc1 {(h6c1 c8b7
c1d2 b7d5 d2h6 g4h3 c3d4 d5g8 h6g7 h3g4 d4c3 g4h3 c3b4 g8h7 d8c7 h3g4 b4c3)
+11.69/20 1} 71. Qd7 Bf4 {(c1f4 d7e8 f6g7 e8b5 c3b2 g4f3 b2d4 b5e8 g7f6
f3g4 f6g7) +11.66/20 1} 72. Qc8 Bg5 {(f4g5 c8b7) +11.65/17} 73. Qd7 Kf7
{(f6f7 d7c6 g5e3 c6d5 f7g7 d5e6 e3f4 e6c8 g7f6) +11.69/18} 74. Qc8 Bf4
{(g5f4 c8d7 f4e3 g4f3) 50 +11.67/19 1 50 moves rule} 75. Qd7 {50 moves rule
50 moves rule} 1/2-1/2

[Event "Computer chess game"]
[Site "OWNER-PC"]
[Date "2013.10.28"]
[Round "?"]
[White "owner"]
[Black "QN"]
[Result "*"]
[BlackElo "2000"]
[Time "15:50:52"]
[WhiteElo "2400"]
[TimeControl "300"]
[SetUp "1"]
[FEN "bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1"]
[Termination "unterminated"]
[PlyCount "56"]
[WhiteType "human"]
[BlackType "human"]

1. e4 b6 {(b7b6 g2g4 c7c5 c2c3 e7e5 d2d4 c5d4 c3d4 f8b4 e1e2 a8b7) +4.65/14
9} 2. f3 d5 {(d7d5 d1e2 c7c5 c2c3 d5e4 f3e4 e7e5 b2b4) +4.93/15 10} 3. d3
e5 {(e7e5 a1c1 c8e6 c2c4 d5d4 f3f4 e5f4 c1f4 f8b4 e1f1 a8b7 e4e5) +5.05/13
7} 4. c4 d4 {(d5d4 h2h4 a8b7 a1c1 h7h6 g2g4 c7c5 d1a4 c8d7 a4a3 f8d6 h4h5)
+5.48/14 9} 5. g4 g5 {(g7g5 h1g2 a8b7 g2d2 c8d7 a2a4 h8g7 a4a5 f8c5 a5a6
b7c6) +5.65/14 4} 6. a4 Bb4+ {(f8b4 e1e2 c7c6 d1c2 b8d6 c2b3 a8b7 h1g2 h8g7
e2f1 c8e6 a4a5 d8e7 a5b6 a7b6) +6.49/16 4} 7. Ke2 c6 {(c7c6 a1b1 b8d6 b1c2
a8b7 h1f1 a7a6 c2b3 h8g7 b3c2 c8d7 a4a5 b4a5) +6.56/17 4} 8. b3 Bab7 {(a8b7
a1a2 b8d6 d1f1 h8g7 f1f2 d8e7 h1g2 c8e6 g2f1 c6c5 f1d1 f7f6 h2h4 g8f7 h4g5
f6g5) +6.91/17 5} 9. Qc2 Bde7 {(d8e7 a1a2 c8e6 a2b1 h8g7 b1a2 b8d6 c2b1
c6c5 e2d1 b4c3 a2g2 e8f8 g2g3) +6.90/17 5} 10. h3 c5 {(c6c5 a1g1 b8d6 g1h2
h8g7 h2g1 c8e6 c2b1 h7h6 e2d1 f7f6 h3h4 b4c3 h4g5 h6g5) +7.04/17 4} 11. Qg2
Bg7 {(h8g7 a1b2 b8d6 b2a1 c8e6 e2f2 f7f6 h3h4 h7h6 c2d1 g8h7 h4h5 b4c3 a1a2
c3b4) +7.08/17 7} 12. Qgh1 Bbd6 {(b8d6 a1b1 c8e6 b1b2 h7h6 e2d1 g8h7 b2a2
h7g6 c2e2 b4c3) +7.19/16 4} 13. Qg2 Bd7 {(c8d7 a1a2 d7e6 a2a1 h7h6 a1h1
g8h7 h1h2 h7g6 g2g3 b4c3 h2g1) +7.12/17 9} 14. Qgh1 f6 {(f7f6 h1h2 g8e6
e2f1 h7h6 c2f2 a7a6 a1c1 b6b5 a4b5 a6b5 c1b2 b4c3 b2a3) +7.20/16 13} 15.
Qg2 h6 {(h7h6 a1h1 g8e6 c2d1 e8d8 h3h4 d8c7 h4g5 h6g5 h1h7 e7f8 h7g6 b4c3)
+7.21/14 4} 16. Qgh1 Bge6 {(g8e6 a1g1 e8f8 c2a2 b4c3 e2f1 a7a6 h3h4 c3b4
g1h2 b6b5 h4g5 h6g5 a4b5 a6b5) +7.25/14 3} 17. Qg2 Bf7 {(e6f7 a1g1 e8f8
g2g3 f7e6 c2b1 a7a6 b1c2 e6f7 g1f1 b4c3 e2f2) +7.23/14 3} 18. Qgh1 Bc3
{(b4c3 a1a2 h6h5 a2a3 h5g4 h3g4 f7e6 a3c1 c3b4 c1b2 e8f8 e2f2 b4c3 b2a3
c3b4) +7.35/17 4} 19. Qaa2 Bde6 {(d7e6 e2f1 e8f8 c2b1 f8g8 a2g2 h6h5 g2g3
h5g4 h3g4 c3b4 g3f2 g8f8 f2g3) +7.32/16 5} 20. Qg2 Bbc8 {(b7c8 e2d1 c8d7
a2b1 c3b4 b1a2 e8f8 a2b1 b4c3 b1a2 f8g8) +7.34/16 3} 21. Qh1 h5 {(h6h5 a2b1
h5g4 h3g4 c8b7 b1a2 e8f8 a2b1 e6d7 b1a2 f7e6 e2f2 c3b4 a2b1 f8f7 b1a2)
+7.39/17 4} 22. Qg2 Bb7 {(c8b7 c2b1 h5g4 h3g4 e6d7) +7.40/17 3} 23. Qh1 Bd7
{(e6d7 e2d1 h5h4 c2h2 f7e6 a2b1 c3b4 b1c2 e8f8 c2b1 b4c3 h2e2 f8f7 e2h2
c3b4 b1c2) +7.30/17 6} 24. Qg2 Kf8 {(e8f8 c2c1 h5g4 h3g4 f7e6 g2h3 c3b4
e2f2 f8f7 f2e2 b4c3 h3g2 f7g8 g2h3 c3b4 a2b1 b4c3) +7.42/17 2} 25. Qh1 h4
{(h5h4 e2d1 f7e6 a2b1 c3b4 b1a2 f8f7 c2f2 b4c3 d1c1 f7f8 c1d1 f8g8 f2c2
c3b4 a2b1) +7.42/17 3} 26. Qg2 Kg8 {(f8g8 e2d1 f7e6 d1e2 c3b4 a2b1 g8h8
b1a2 b4c3 g2h1 h8h7 h1g2 c3b4 e2f2 h7h6 f2e2 b4c3) +7.42/18 4} 27. Qh1 Kh8
{(g8h8 a2a3 f7e6 a3c1 c3b4 c1b2 h8h7 b2c1 b4c3 e2f2 h7h6 f2e2 c3b4 c1b2
h6g6 b2c1 b4c3) +7.40/18 4} 28. Qg2 Bb4 {(c3b4 e2d1 f7e6 g2f1 b4c3 a2b1
h8h7 b1a2 c3b4 a2b1) +7.41/19 6} *
[/pgn]

[d]2Qb4/4b3/1p1p1kp1/p1p1p3/P1PbP1K1/1P1P4/3b4/8 b - - 27 38
Queeny thinks it has some 12 pawns advantage here, but the position actually is a draw. You see how much complementary control of squares matters. Here elephantiasis would predict wrong, there is a large number of minor pieces, but all of them control squares of same colour, so that complementarity is at its lowest. Probably with bishops a further specification would be required, taking into consideration the availability of bishops of opposite colours.

Interesting, how many engines see the draw here?

[d]7k/pb1bbbb1/1p1b1p2/2p1p1p1/PbPpP1Pp/1P1P1P1P/Q1Q1K1Q1/8 w - - 6 29
This is draw, as the bishop side can not break through. Now change all black bishops with black knights, and black already wins; you can place knights on b4 and f4, for example, capture d3 and proceed to win. This certainly demonstrates once again how detrimental are blocked pawns to bishops, and how stimulating they are to knights. So that, actually, it is difficult to object to the fact that bishops lose in strength with the number of blocked pawns, while knights, quite the opposite, gain strength.

[hgm]

It seems I am able to draw only against Queeny.

That is a very poor result. QueeNy is a weak engine (the search was ripped out of an engine that would be 2400 Elo on the CCRL list), yet in self-play it always wins with 3 Queens against 7 Bishops. So you are doing worse than a weak engine would do.

2 games, one where Q vs 4Bs of the same colour is a draw, but Queeny does not seem to recognize that. Hey, Harm, does Queeny really think that 4 same-colour bishops are still worth an elephantiasis bonus?

Of course. As I explained before QueeNy is a basic search plus an evaluation that uses Q=9.5, R=B=N=5. And 7*5 = 35. You can trust it knows how to multiply.

Note that QueeNy was constructed to play 'Charge of the Light Brigade', i.e. 7 Knights against 3 Queens. That it even knows how Bishops or Knights move is entirely coincidental, a side effect from the fact the search and move generator were taken from a Chess engine under construction. And at that time the Chess engine in question did not have material tables. So it considers KNK a +3 score, and KNNK a +6, etc.

All extremely simplistic, but good enough to beat engines like Houdini, Stockfish etc. with 7 Knights (and the somewhat weaker like Fruit and Crafty even with 6 Knights) against 3 Queens.

Queeny thinks it has some 12 pawns advantage here, but the position actually is a draw.

The absolute value of QueeNy's score should never be taken seriously. According to the elephantiasis effect each successive Knight is worth more (i.e. its marginal value increases), such that the first two are worth 325 each, but number 6 and 7 are worth over 500 each. QueeNy, with its purely additive eval, counts all Knights for 500 (because in the typical games it plays the Knights hardly get traded to below 5.) So that is already 350 cP too much for the first two knights, and some other 175 for the nex two.

Further notice that the value for the Bishop was just set equal to that of the Knightas a blind guess for its marginal value when it would cooperate with many Knights, not with other Bishops. So it is likely to be completely off.

Probably with bishops a further specification would be required, taking into consideration the availability of bishops of opposite colours.

Indeed, Bishops are much more complex pieces than Knights, due to color binding and pair bonuses. Empirical results generated with programs that do not take thes into account are quite meaningless. Unfortunately normal Chess programs would not calculate these sensibly for more than two Bishops. (And even Fruit assumes that two Bishops would always be on different colors, i.e. form a pair.)

[Eelco de Groot]

Even though the side with the bishops has four bishop pairs, I think it should be won in all cases for the side with the Queens. So playing for a draw by closing the position is not really good enough Lyudmil. I made a new Scid Serpent version with increased redundancy for same squared bishops, but that does not really apply here, with four bishop pairs the redundancy factor will not be awarded unless the bishop pairs will be broken up, only redundancy of the three Queens applies. But I think the Queens still win. I started a small match of Scid Serpent agains Stockfish 4 in SCID that allows these kind of illegal positions luckily. I predict that White will win eight out of eight...

[d]bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1

It is a timecontol 1 minute no Fischer bonus match so it does not take long.

Eelco

[Lyudmil Tsvetkov]

Even though the side with the bishops has four bishop pairs, I think it should be won in all cases for the side with the Queens. So playing for a draw by closing the position is not really good enough Lyudmil. I made a new Scid Serpent version with increased redundancy for same squared bishops, but that does not really apply here, with four bishop pairs the redundancy factor will not be awarded unless the bishop pairs will be broken up, only redundancy of the three Queens applies. But I think the Queens still win. I started a small match of Scid Serpent agains Stockfish 4 in SCID that allows these kind of illegal positions luckily. I predict that White will win eight out of eight...

[d]bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1

It is a timecontol 1 minute no Fischer bonus match so it does not take long.

Eelco

It seems I am the only person that has actually played this particular imbalance, so I can tell you without any doubt, no predictions at all, because the position is very simple, that the bishop side will win all the games.
I would be happy if you post the results after the match is finished.

[Lyudmil Tsvetkov]

It seems I am able to draw only against Queeny.

That is a very poor result. QueeNy is a weak engine (the search was ripped out of an engine that would be 2400 Elo on the CCRL list), yet in self-play it always wins with 3 Queens against 7 Bishops. So you are doing worse than a weak engine would do.

2 games, one where Q vs 4Bs of the same colour is a draw, but Queeny does not seem to recognize that. Hey, Harm, does Queeny really think that 4 same-colour bishops are still worth an elephantiasis bonus?

Of course. As I explained before QueeNy is a basic search plus an evaluation that uses Q=9.5, R=B=N=5. And 7*5 = 35. You can trust it knows how to multiply.

Note that QueeNy was constructed to play 'Charge of the Light Brigade', i.e. 7 Knights against 3 Queens. That it even knows how Bishops or Knights move is entirely coincidental, a side effect from the fact the search and move generator were taken from a Chess engine under construction. And at that time the Chess engine in question did not have material tables. So it considers KNK a +3 score, and KNNK a +6, etc.

All extremely simplistic, but good enough to beat engines like Houdini, Stockfish etc. with 7 Knights (and the somewhat weaker like Fruit and Crafty even with 6 Knights) against 3 Queens.

Queeny thinks it has some 12 pawns advantage here, but the position actually is a draw.

The absolute value of QueeNy's score should never be taken seriously. According to the elephantiasis effect each successive Knight is worth more (i.e. its marginal value increases), such that the first two are worth 325 each, but number 6 and 7 are worth over 500 each. QueeNy, with its purely additive eval, counts all Knights for 500 (because in the typical games it plays the Knights hardly get traded to below 5.) So that is already 350 cP too much for the first two knights, and some other 175 for the nex two.

Further notice that the value for the Bishop was just set equal to that of the Knightas a blind guess for its marginal value when it would cooperate with many Knights, not with other Bishops. So it is likely to be completely off.

Probably with bishops a further specification would be required, taking into consideration the availability of bishops of opposite colours.

Indeed, Bishops are much more complex pieces than Knights, due to color binding and pair bonuses. Empirical results generated with programs that do not take thes into account are quite meaningless. Unfortunately normal Chess programs would not calculate these sensibly for more than two Bishops. (And even Fruit assumes that two Bishops would always be on different colors, i.e. form a pair.)

Again, it seems you have not tested the 3Qs vs 7Bs imbalance, because if you had, you would certainly know that the imbalance is absolutely hopeless for the queen side.
7 Bs are much stronger than 7Ns, but really much much stronger.

[Eelco de Groot]

Even though the side with the bishops has four bishop pairs, I think it should be won in all cases for the side with the Queens. So playing for a draw by closing the position is not really good enough Lyudmil. I made a new Scid Serpent version with increased redundancy for same squared bishops, but that does not really apply here, with four bishop pairs the redundancy factor will not be awarded unless the bishop pairs will be broken up, only redundancy of the three Queens applies. But I think the Queens still win. I started a small match of Scid Serpent agains Stockfish 4 in SCID that allows these kind of illegal positions luckily. I predict that White will win eight out of eight...

[d]bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1

It is a timecontol 1 minute no Fischer bonus match so it does not take long.

Eelco

It seems I am the only person that has actually played this particular imbalance, so I can tell you without any doubt, no predictions at all, because the position is very simple, that the bishop side will win all the games.
I would be happy if you post the results after the match is finished.

Well the results are a bit inconclusive. Maybe the opportunities for both sides are a bit more balanced than the resident elephantologists predicted

I got 3 wins with White, one draw and four wins with Black of which one was on time in an equal position, the other wins look more regular with mate. It is possible I have overlooked something as I have never played a match in SCID before and have no idea how to access the clipbase for instance. Stockfish 4 won the match so I think the Queen elephantiasis rule has to be modified a bit, it seems Stockfish plays this better without, maybe not suprising against bishops instead of knights

Eelco

[hgm]

Well QueeNy has no problem winning this for the Queens against itself:

[pgn][Event "Computer Chess Game"]
[Site "MAKRO-PC"]
[Date "2013.10.28"]
[Round "-"]
[White "QueeNy 0.16"]
[Black "QueeNy 0.16"]
[Result "1-0"]
[TimeControl "40/300"]
[FEN "bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1"]
[SetUp "1"]
1. e4 {-4.57/14} b6 {+4.61/14 5} 2. f3 {-4.81/14 7} d5 {+4.90/14 4} 3. exd5
{-4.98/14 7} Bxd5 {+4.95/15 6} 4. d4 {-5.01/14 6} c6 {+5.17/14 5} 5. b3
{-5.23/14 7} b5 {+5.37/14 6} 6. Qdc1 {-5.23/13 7} Ba5+ {+5.57/14 13} 7. Kd1
{-5.58/14 6} f6 {+5.75/15 8} 8. h4 {-5.78/13 7} e5 {+6.06/15 11} 9. c3
{-5.96/13 5} Bb6 {+6.26/14 6} 10. Qab2 {-6.33/13 5} exd4 {+6.59/15 8} 11.
cxd4 {-6.57/14 8} a6 {+6.62/13 5} 12. h5 {-6.67/13 8} B8a7 {+6.70/13 8} 13.
Qf4 {-6.69/12 4} Bge6 {+6.87/13 6} 14. Qb1 {-6.24/13 6} f5 {+6.08/14 5} 15.
g4 {-6.06/14 4} Bxd4 {+5.91/14 8} 16. gxf5 {-5.96/14 6} Bg8 {+5.97/13 6}
17. Qd3 {-5.59/13 6} Kd7 {+5.41/13 8} 18. Qhh2 {-3.84/15 6} Bdb6
{+3.86/14 4} 19. h6 {-3.83/15 8} Bf7 {+2.90/14 9} 20. Qde2 {-1.03/14 11}
Bfc5 {+0.39/13 11} 21. Qee5 {+0.40/13 4} Be7 {-1.77/13 10} 22. Qc7+
{+5.90/14 7} Ke8 {-5.47/14 4} 23. Qxc8+ {+10.69/14 8} Bbd8 {-11.49/15 15}
24. Qd6 {+12.64/14 16} Bxb3+ {-320.00/100 26} 25. axb3 {+19.54/13 4} Bxb3+
{-319.93/14 22} 26. Kd2 {+319.91/14 9} Bb6 {-319.92/14 5} 27. Qdxc6+
{+319.92/13 4} Kf8 {-319.94/14 2.6} 28. Qxb6 {+319.94/14 5} Bg5+
{-319.95/17 6} 29. Ke1 {+319.95/17 5} gxh6 {-319.96/16 6} 30. Qcxd8+
{+319.96/18 4} Bxd8 {-319.97/25 3} 31. Qxd8+ {+319.97/26 5} Kg7
{-319.98/31 2.9} 32. Qe7+ {+319.98/33 6} Bf7 {-319.99/44 2.9} 33. Qg3#
{+319.99/48 5}
{Xboard adjudication: Checkmate} 1-0
[/pgn]

Against Stockfish it draws with the Queens. Which is also remarkable, considering that Stockfish is some 600 Elo stronger.

[pgn][Event "Computer Chess Game"]
[Site "MAKRO-PC"]
[Date "2013.10.28"]
[Round "-"]
[White "QueeNy 0.16"]
[Black "Stockfish 4 64 SSE4.2"]
[Result "1/2-1/2"]
[TimeControl "40/300"]
[FEN "bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1"]
[SetUp "1"]
1. e4 {-4.65/14} c5 {-8.76/17 6} 2. d4 {-4.70/14 4} g6 {-8.20/18 5} 3. c3
{-4.72/15 7} cxd4 {-7.15/20 11} 4. cxd4 {-4.75/15 7} b6 {-6.98/20 8} 5.
Qac1 {-5.09/14 8} Bcb7 {-6.86/21 11} 6. d5 {-5.70/13 5} Bfg7 {-6.70/21 8}
7. Qe2 {-5.73/14 9} Bxb2 {-5.29/20 8} 8. Qcxb2 {-6.30/15 5} Bxb2
{-2.14/18 3} 9. Qxb2 {-6.72/17 5} f6 {-2.06/18 6} 10. f3 {-6.82/16 5} e6
{-2.06/19 8} 11. dxe6 {-6.98/16 7} Bxe6 {-2.06/20 10} 12. h4 {-7.20/15 8}
Ba6 {-0.96/19 6} 13. Qd2 {-7.09/15 9} Bg3+ {+0.00/25 9} 14. Kd1
{-7.51/16 6} Bc6 {+0.00/28 5} 15. Qh6 {-7.36/15 6} Bdc7 {+2.54/21 2.2} 16.
Qxh7 {-6.21/15 8} Bce5 {+2.58/21 2.4} 17. Qxg6+ {-8.79/14 5} Kf8
{+2.38/23 9} 18. Qg1 {-7.54/15 5} Ba4+ {+2.34/19 7} 19. Kd2 {-8.50/16 4}
Bgf4+ {+2.34/20 2.7} 20. Qe3 {-8.28/16 5} Bf7 {+2.26/22 7} 21. Qh6+
{-7.21/16 5} Bxh6 {+2.02/22 9} 22. Qxh6+ {-7.19/17 5} Kg8 {+1.33/22 25} 23.
f4 {-7.28/18 9} Bb2 {+1.31/21 10} 24. h5 {-6.15/16 9} d6 {+0.00/19 7} 25.
e5 {-5.56/17 9} fxe5 {-0.06/20 7} 26. f5 {-4.32/16 9} e4 {+0.00/20 3} 27.
f6 {-4.36/17 5} e3+ {+0.00/22 4} 28. Qxe3 {-4.39/18 10} Bxf6 {+0.00/24 4}
29. Qf4 {-4.22/17 8} Kg7 {+0.00/24 10} 30. Qxa4 {-4.07/16 9} Bac4
{+0.00/24 5} 31. Qxa7 {-4.22/18 5} Bxa2 {+0.00/25 6} 32. Qxb6 {-3.78/20 7}
Bxh5 {+0.00/28 7} 33. Qxd6 {-3.76/21 7} Bhf7 {+0.00/29 6} 34. Qg3+
{-3.49/21 6} Kh6 {+0.00/30 7} 35. Qh3+ {-0.12/25 6} Kg5 {+0.00/28 6} 36.
Qg3+ {-0.10/23 9} Kf5 {+0.00/28 6} 37. Qh3+ {-0.05/23 11} Kg5 {+0.00/39 20}
38. Qg3+ {+0.01/28 7} Kf5 {+0.00/54 5} 39. Qh3+ {+0.04/25 7} Kg5
{+0.00/48 5}
{XBoard adjudication: repetition draw} 1/2-1/2
[/pgn]

So 'absolutely hopeless for the Queens' from your mouth seems again to mean: 'it is not proven convincingly the Queens always win'!

[Lyudmil Tsvetkov]

Well QueeNy has no problem winning this for the Queens against itself:

[pgn][Event "Computer Chess Game"]
[Site "MAKRO-PC"]
[Date "2013.10.28"]
[Round "-"]
[White "QueeNy 0.16"]
[Black "QueeNy 0.16"]
[Result "1-0"]
[TimeControl "40/300"]
[FEN "bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1"]
[SetUp "1"]
1. e4 {-4.57/14} b6 {+4.61/14 5} 2. f3 {-4.81/14 7} d5 {+4.90/14 4} 3. exd5
{-4.98/14 7} Bxd5 {+4.95/15 6} 4. d4 {-5.01/14 6} c6 {+5.17/14 5} 5. b3
{-5.23/14 7} b5 {+5.37/14 6} 6. Qdc1 {-5.23/13 7} Ba5+ {+5.57/14 13} 7. Kd1
{-5.58/14 6} f6 {+5.75/15 8} 8. h4 {-5.78/13 7} e5 {+6.06/15 11} 9. c3
{-5.96/13 5} Bb6 {+6.26/14 6} 10. Qab2 {-6.33/13 5} exd4 {+6.59/15 8} 11.
cxd4 {-6.57/14 8} a6 {+6.62/13 5} 12. h5 {-6.67/13 8} B8a7 {+6.70/13 8} 13.
Qf4 {-6.69/12 4} Bge6 {+6.87/13 6} 14. Qb1 {-6.24/13 6} f5 {+6.08/14 5} 15.
g4 {-6.06/14 4} Bxd4 {+5.91/14 8} 16. gxf5 {-5.96/14 6} Bg8 {+5.97/13 6}
17. Qd3 {-5.59/13 6} Kd7 {+5.41/13 8} 18. Qhh2 {-3.84/15 6} Bdb6
{+3.86/14 4} 19. h6 {-3.83/15 8} Bf7 {+2.90/14 9} 20. Qde2 {-1.03/14 11}
Bfc5 {+0.39/13 11} 21. Qee5 {+0.40/13 4} Be7 {-1.77/13 10} 22. Qc7+
{+5.90/14 7} Ke8 {-5.47/14 4} 23. Qxc8+ {+10.69/14 8} Bbd8 {-11.49/15 15}
24. Qd6 {+12.64/14 16} Bxb3+ {-320.00/100 26} 25. axb3 {+19.54/13 4} Bxb3+
{-319.93/14 22} 26. Kd2 {+319.91/14 9} Bb6 {-319.92/14 5} 27. Qdxc6+
{+319.92/13 4} Kf8 {-319.94/14 2.6} 28. Qxb6 {+319.94/14 5} Bg5+
{-319.95/17 6} 29. Ke1 {+319.95/17 5} gxh6 {-319.96/16 6} 30. Qcxd8+
{+319.96/18 4} Bxd8 {-319.97/25 3} 31. Qxd8+ {+319.97/26 5} Kg7
{-319.98/31 2.9} 32. Qe7+ {+319.98/33 6} Bf7 {-319.99/44 2.9} 33. Qg3#
{+319.99/48 5}
{Xboard adjudication: Checkmate} 1-0
[/pgn]

Against Stockfish it draws with the Queens. Which is also remarkable, considering that Stockfish is some 600 Elo stronger.

[pgn][Event "Computer Chess Game"]
[Site "MAKRO-PC"]
[Date "2013.10.28"]
[Round "-"]
[White "QueeNy 0.16"]
[Black "Stockfish 4 64 SSE4.2"]
[Result "1/2-1/2"]
[TimeControl "40/300"]
[FEN "bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1"]
[SetUp "1"]
1. e4 {-4.65/14} c5 {-8.76/17 6} 2. d4 {-4.70/14 4} g6 {-8.20/18 5} 3. c3
{-4.72/15 7} cxd4 {-7.15/20 11} 4. cxd4 {-4.75/15 7} b6 {-6.98/20 8} 5.
Qac1 {-5.09/14 8} Bcb7 {-6.86/21 11} 6. d5 {-5.70/13 5} Bfg7 {-6.70/21 8}
7. Qe2 {-5.73/14 9} Bxb2 {-5.29/20 8} 8. Qcxb2 {-6.30/15 5} Bxb2
{-2.14/18 3} 9. Qxb2 {-6.72/17 5} f6 {-2.06/18 6} 10. f3 {-6.82/16 5} e6
{-2.06/19 8} 11. dxe6 {-6.98/16 7} Bxe6 {-2.06/20 10} 12. h4 {-7.20/15 8}
Ba6 {-0.96/19 6} 13. Qd2 {-7.09/15 9} Bg3+ {+0.00/25 9} 14. Kd1
{-7.51/16 6} Bc6 {+0.00/28 5} 15. Qh6 {-7.36/15 6} Bdc7 {+2.54/21 2.2} 16.
Qxh7 {-6.21/15 8} Bce5 {+2.58/21 2.4} 17. Qxg6+ {-8.79/14 5} Kf8
{+2.38/23 9} 18. Qg1 {-7.54/15 5} Ba4+ {+2.34/19 7} 19. Kd2 {-8.50/16 4}
Bgf4+ {+2.34/20 2.7} 20. Qe3 {-8.28/16 5} Bf7 {+2.26/22 7} 21. Qh6+
{-7.21/16 5} Bxh6 {+2.02/22 9} 22. Qxh6+ {-7.19/17 5} Kg8 {+1.33/22 25} 23.
f4 {-7.28/18 9} Bb2 {+1.31/21 10} 24. h5 {-6.15/16 9} d6 {+0.00/19 7} 25.
e5 {-5.56/17 9} fxe5 {-0.06/20 7} 26. f5 {-4.32/16 9} e4 {+0.00/20 3} 27.
f6 {-4.36/17 5} e3+ {+0.00/22 4} 28. Qxe3 {-4.39/18 10} Bxf6 {+0.00/24 4}
29. Qf4 {-4.22/17 8} Kg7 {+0.00/24 10} 30. Qxa4 {-4.07/16 9} Bac4
{+0.00/24 5} 31. Qxa7 {-4.22/18 5} Bxa2 {+0.00/25 6} 32. Qxb6 {-3.78/20 7}
Bxh5 {+0.00/28 7} 33. Qxd6 {-3.76/21 7} Bhf7 {+0.00/29 6} 34. Qg3+
{-3.49/21 6} Kh6 {+0.00/30 7} 35. Qh3+ {-0.12/25 6} Kg5 {+0.00/28 6} 36.
Qg3+ {-0.10/23 9} Kf5 {+0.00/28 6} 37. Qh3+ {-0.05/23 11} Kg5 {+0.00/39 20}
38. Qg3+ {+0.01/28 7} Kf5 {+0.00/54 5} 39. Qh3+ {+0.04/25 7} Kg5
{+0.00/48 5}
{XBoard adjudication: repetition draw} 1/2-1/2
[/pgn]

So 'absolutely hopeless for the Queens' from your mouth seems again to mean: 'it is not proven convincingly the Queens always win'!

Look how Queeny scores jump: they jump more than a kangaroo!

[hgm]

Kanguru is another engine... I am not sure what you refer too. To me the eval graph looks pretty smooth.

Second game ended in a draw as well. Stockfish must really be extremely incompetent at this, that it cannot win from a 600 Elo weaker engine from such a won position, or the position is not nearly as won as you claim.

[pgn][Event "Computer Chess Game"]
[Site "MAKRO-PC"]
[Date "2013.10.28"]
[Round "-"]
[White "QueeNy 0.16"]
[Black "Stockfish 4 64 SSE4.2"]
[Result "1/2-1/2"]
[TimeControl "40/300"]
[FEN "bbbbkbbb/pppppppp/8/8/8/8/PPPPPPPP/Q2QK2Q w - - 0 1"]
[SetUp "1"]
1. c3 {-4.66/14} b6 {-8.82/18 10} 2. d3 {-4.78/13 4} c5 {-8.66/17 6} 3. d4
{-4.95/13 5} g6 {-8.10/18 7} 4. f3 {-5.04/13 4} cxd4 {-7.57/19 6} 5. cxd4
{-5.60/15 8} Bdc7 {-7.25/20 16} 6. Qd2 {-5.26/13 7} Bfg7 {-6.14/19 7} 7. e3
{-5.92/15 4} f5 {-5.93/21 8} 8. f4 {-5.97/15 5} Bcb7 {-5.39/21 8} 9. Qg1
{-6.13/15 5} Bxg2 {-4.78/22 24} 10. Qdxg2 {-6.18/16 8} Bxg2 {-2.84/18 1.6}
11. Qxg2 {-6.16/16 6} Bxd4 {-2.82/23 7} 12. exd4 {-5.91/15 5} Bxd4
{-2.88/24 20} 13. Qc1 {-6.21/15 6} Be6 {-2.76/24 6} 14. Qh3 {-6.47/17 6} h5
{-2.76/23 7} 15. Qd2 {-6.94/17 7} Bf6 {-2.70/21 5} 16. Qdg2 {-6.91/16 5}
Bf7 {-2.74/22 7} 17. Qb7 {-7.08/14 5} Kf8 {-2.64/22 6} 18. Qd3 {-7.17/15 9}
d5 {-2.38/19 5} 19. Qc2 {-7.02/16 6} Bxf4 {-2.40/20 6} 20. Qcc8+
{-6.85/16 5} Kg7 {-1.87/19 1.5} 21. Qbxb8 {-6.79/16 5} Bxb8 {+0.34/19 5}
22. Qxb8 {-6.66/21 7} Bd4 {+0.24/20 8} 23. Qxa7 {-6.55/21 9} e5
{+0.00/21 11} 24. Qc7 {-6.48/20 6} f4 {+0.00/24 10} 25. Ke2 {-6.18/21 6} h4
{+0.00/26 7} 26. b4 {-5.19/22 7} b5 {+0.00/26 5} 27. Qd6 {-4.89/20 5} g5
{+0.00/25 4} 28. Qd8 {-4.57/21 5} Kg6 {+0.00/26 5} 29. Qd6+ {-3.94/23 7}
Kf5 {+0.00/28 5} 30. Qf8 {-3.74/24 11} Kf6 {+0.00/28 7} 31. Qd8+
{-3.44/23 5} Kf5 {+0.00/30 5} 32. Qf8 {-0.04/28 10} Kg6 {+0.00/28 6} 33.
Qd6+ {+0.07/30 9} Kf5 {+0.00/34 10} 34. Qf8 {-0.01/26 7}
{XBoard adjudication: repetition draw} 1/2-1/2
[/pgn]