Queens vs Knights

[hgm]

so what happens when you have an UCI engine correctly showing score from white point of view only?

The same as what happens when you correctly misspell a word.

watching scores has meaning only for humans, and those are accustomed to the anchor of the white side, in this way it is immediately clear who has the advantage.

But humans are not supposed to watch raw engine output. UCI is intended for communication with the GUI, not with humans. This is also why it uses the human-unfriendly long-algebraic move notation.

not to mention the case, when one engine shows score from white point of view, and another one from stm point of view: a total chaos.

This is why it is important to stick to the standard.

[hgm]

with big pawn span, queens always fork something; this is unavoidable, I checked that many times before.

That doesn't matter if all pieces are protected.

[hgm]

I now changed all probing (for new game moves, new PVs, and new chat messages) by the viewer JavaScript to asynchronous calls. Indeed the loss of characters that previously occurred when typing in the chat now has disappeared.

In Firefox on Linux (which seems to use much larger font than on Windows?!) the upper part of the page looks like this:



Below that (not shown, as I could not fit it on the small screen of my Linux laptop) there are buttons to fetch a list with available games from the server, so you can select the one you want to view from that list.

[Evert]

maybe you will share your code with us?

Sure. This is the functional change over the Stockfish master I checked out this morning:

Code: Select all

--- Stockfish/src/material.cpp	2017-07-18 13:21:30.000000000 +0200
+++ Nightfish/src/material.cpp	2017-07-18 23:05:43.000000000 +0200
@@ -121,6 +121,10 @@
     if  (pieceCount[Us][QUEEN] == 1 && pieceCount[Them][QUEEN] == 0)
          bonus += QueenMinorsImbalance[pieceCount[Them][KNIGHT] + pieceCount[Them][BISHOP]];
 
+    // Elephantiasis correction for Q vs N
+    if (pieceCount[Them][KNIGHT] > 2)
+       bonus -= 16*112*(pieceCount[Them][KNIGHT] - 2)*pieceCount[Us][QUEEN];
+
     return bonus;
   }

I had originally missed that Stockfish scales the imbalance correction by a factor 16 (so they are not in... whatever you'd call the evaluation unit in Stockfish). The factor "112" corresponds to 60cp. Note that this is not quite enough to reduce the evaluation in the root position to "balanced". Evaluation still favours the queens by about 100cp.

from what position?

The relevant start position,
[d]nnnnknnn/pppppppp/8/8/8/8/PPPPPPPP/1Q1KQ1Q1 w - -

how many games?

Six or seven at various time controls, but I'll start a longer series to run overnight.

what TC?

40/0:30. I ran a couple at 40/1:00. At longer time controls the Queens hang in there better (or perhaps it's the correction to the score, since otherwise the queens will think themselves superior enough to play recklessly), meaning they will draw when they can successfully close the position completely.

SF-revised playing both sides?

Of course.

can we look at some of the games?

Sure. Here's a couple; I adjudicated the first.
[pgn]
[Event "Computer Chess Game"]
[Site "vivaine.local"]
[Date "2017.07.18"]
[Round "1"]
[White "Nightfish 180717 64"]
[Black "Nightfish 180717 64"]
[Result "1/2-1/2"]
[TimeControl "40/60"]
[FEN "nnnnknnn/pppppppp/8/8/8/8/PPPPPPPP/1Q1KQ1Q1 w - - 0 1"]
[SetUp "1"]

{--------------
n n n n k n n n
p p p p p p p p
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
P P P P P P P P
. Q . K Q . Q .
white to play
--------------}
1. f4 {+1.59/17} e6 {-1.29/16 2.1} 2. e4 {+1.39/20 2.0} Nhg6 {-1.28/17 1.4}
3. f5 {+1.29/15 0.6} N6e7 {-1.37/18 1.6} 4. b4 {+1.24/17 1.4} d6
{-1.42/15 0.3} 5. a4 {+1.09/18 4} Nbd7 {-1.34/16 1.7} 6. d3 {+1.28/16 1.2}
h6 {-1.14/16 1.2} 7. g4 {+1.13/17 2.7} Ngf6 {-1.25/18 1.6} 8. Qb2
{+1.29/17 1.3} g5 {-1.14/17 2.6} 9. Qg2 {+1.23/18 2.4} e5 {-1.01/18 4} 10.
Qh3 {+1.04/18 2.6} Neg8 {-0.95/17 1.0} 11. Kc1 {+1.03/18 1.2} c6
{-0.85/18 1.7} 12. Qhg3 {+0.88/18 4} Nc7 {-0.74/16 1.0} 13. c3
{+0.77/19 2.3} N8h7 {-0.69/18 2.3} 14. Qee2 {+0.75/18 4} b6 {-0.80/17 1.4}
15. d4 {+0.68/16 1.4} Kf8 {-0.63/17 0.9} 16. Qed3 {+0.89/18 1.9} h5
{-0.79/19 2.3} 17. h3 {+0.55/19 3} h4 {-0.63/16 0.6} 18. Qgh2
{+0.50/15 0.4} Nce8 {-0.60/15 0.5} 19. Qbc2 {+0.51/17 4} Nge7
{-0.08/17 1.9} 20. dxe5 {+0.38/17 1.7} Nxe5 {-1.16/21 2.7} 21. Qxe5
{+0.36/16 0.5} dxe5 {-1.05/20 0.7} 22. Qxd8 {+0.83/18 0.6} Kg7
{-1.10/21 1.4} 23. Qh2 {+1.14/18 0.5} Ned6 {-0.84/18 0.6} 24. Qxe5
{+0.98/17 0.6} Nxe4 {-1.14/22 2.8} 25. a5 {+1.24/19 1.0} Nf2 {-0.79/23 2.7}
26. Qe3 {+1.25/20 0.6} N2e4 {-0.84/24 1.1} 27. c4 {+1.25/21 0.8} c5
{-0.65/23 0.4} 28. bxc5 {+1.08/23 1.5} Nxc5 {-0.31/21 0.7} 29. axb6
{+1.05/23 0.6} axb6 {-0.16/24 1.9} 30. Qc3 {+0.98/21 0.5} Ne4
{-0.16/21 0.4} 31. Qa3 {+0.80/23 1.2} Nf8 {+0.00/22 0.8} 32. Qa1
{+0.20/20 1.0} Kg8 {+0.00/25 1.1} 33. Kd1 {+0.00/22 1.1} Nd7 {+0.00/28 1.1}
34. Qd4 {+0.00/24 0.6} Nec5 {+0.00/29 1.1} 35. Ke2 {+0.00/25 0.5} f6
{+0.00/29 0.6} 36. Qe8 {+0.00/23 0.4} Kg7 {+0.00/31 1.5} 37. Qd2
{+0.00/27 0.5} Ne4 {+0.00/29 1.3} 38. Qd3 {-0.07/29 1.3} Ned6
{+0.08/27 0.9} 39. Qxe7+ {-0.17/21 0.7} Nxe7 {+0.05/36 2.4} 40. Qxd6
{+0.00/22 0.6} Ng8 {+0.20/33 0.6} 41. Qc7 {-0.26/26 0.6} Nh6 {+0.00/37 4}
42. Qa7 {-0.08/34 1.2} Kf7 {+0.00/46 1.9} 43. Qb7 {-0.08/40 1.6} Kg7
{+0.14/40 3} 44. Kf2 {-0.08/38 1.8} Kh7 {+0.14/38 1.8} 45. Qc7
{-0.08/39 1.7} Nf7 {+0.00/38 2.6} 46. Ke2 {-0.08/40 1.9} Kg7 {+0.06/34 0.6}
47. Qa7 {-0.08/39 1.4} Nde5 {+0.08/39 5} 48. Qxb6 {-0.12/36 1.8} Nxc4
{+0.08/37 0.5} 49. Qb7 {-0.12/38 1.5} Ne5 {+0.08/33 1.5} 50. Qe7
{-0.14/34 2.7} Nfd7 {+0.39/33 3} 51. Qe8 {-0.11/35 1.6} Nc5 {+0.55/33 0.8}
52. Kf1 {-0.10/31 0.6} Nb7 {+0.55/35 0.6} 53. Qe7 {-0.42/39 10} Nd8
{+0.39/43 11} 54. Kg1 {-0.08/31 0.8} Nec6 {+0.60/34 5} 55. Qc5
{-0.35/38 2.5} Nb8 {+0.60/34 2.2} 56. Qd5 {-0.47/42 1.6} Ndc6
{+0.60/42 0.7} 57. Qe6 {-0.47/41 0.6} Nfe5 {+0.60/40 0.7} 58. Qe8
{-0.47/37 1.0} Na6 {+0.60/42 0.7} 59. Qc8 {-0.47/41 1.0} Nab8
{+0.60/41 0.4} 60. Kf2 {-0.47/41 0.9} Nbd7 {+0.60/43 0.7} 61. Qe8
{-0.47/39 0.6} Nc5 {+0.60/43 0.4} 62. Kf1 {-0.47/43 1.2} Ncd3
{+0.60/45 0.7} 63. Kg2 {-0.47/43 1.0} Nf4+ {+0.60/43 0.5} 64. Kh2
{-0.47/44 0.6} Nd5 {+0.60/45 0.8} 65. Kg2 {-0.47/38 1.2} Nce7
{+0.60/42 0.9} 66. Kg1 {-0.47/42 1.0} Nf3+ {+0.60/43 0.7} 67. Kf1
{-0.47/44 1.1} Ne5 {+0.60/36 0.7} 68. Kg2 {-0.47/40 1.1} N7c6
{+0.60/39 0.7} 69. Qc8 {-0.47/41 1.2} Nf4+ {+0.60/36 0.6} 70. Kh2
{-0.47/39 1.0} Kf7 {+0.60/39 1.0} 71. Qb7+ {-0.47/35 0.9} Ne7
{+0.60/40 0.7} 72. Qa8 {-0.47/35 0.6} N7c6 {+0.60/39 0.7} 73. Qb7+
{-0.47/37 2.3} Ne7 {+0.60/40 0.5} 74. Qa8 {-0.47/36 0.6} Nfd5 {+0.60/40 3}
75. Qd8 {-0.63/36 2.7} Nf3+ {+0.60/38 0.5} 76. Kg2 {-0.63/35 0.9} Ne5
{+0.60/38 0.4} 77. Kg1 {-0.63/36 1.0} N5c6 {+0.60/37 0.8} 78. Qh8
{-0.63/37 1.0} Ne5 {+0.60/36 0.5} 79. Qd8 {-0.63/36 0.3} Kg7 {+0.60/40 1.3}
80. Kh2 {-0.54/37 3} Nf3+ {+0.60/37 0.7} 81. Kg2 {-0.54/33 2.3} Ne5
{+0.59/35 1.7} 82. Qe8 {+0.00/29 2.5}
{User adjudication} 1/2-1/2
[/pgn]

[pgn][Event "Computer Chess Game"]
[Site "vivaine.local"]
[Date "2017.07.18"]
[Round "3"]
[White "Nightfish 180717 64"]
[Black "Nightfish 180717 64"]
[Result "0-1"]
[TimeControl "40/60"]
[FEN "nnnnknnn/pppppppp/8/8/8/8/PPPPPPPP/1Q1KQ1Q1 w - - 0 1"]
[SetUp "1"]

{--------------
n n n n k n n n
p p p p p p p p
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
P P P P P P P P
. Q . K Q . Q .
white to play
--------------}
1. f4 {+1.59/17} e6 {-1.60/17 2.1} 2. e4 {+1.39/20 2.0} Nhg6 {-1.36/16 0.6}
3. f5 {+1.29/15 0.5} N6e7 {-1.33/18 1.7} 4. b4 {+1.24/17 1.4} d6
{-1.41/15 0.5} 5. a4 {+1.09/18 4} Nbd7 {-1.28/16 0.8} 6. d3 {+1.28/16 1.2}
b6 {-1.02/17 3} 7. c3 {+1.20/17 2.2} c5 {-1.08/18 3} 8. g4 {+1.09/17 1.6}
e5 {-1.03/19 3} 9. b5 {+1.08/17 1.3} Nc7 {-0.88/21 4} 10. c4 {+1.03/19 1.3}
Ngf6 {-0.91/20 1.1} 11. Qa2 {+0.74/21 7} Nb7 {-0.84/20 0.8} 12. Qae2
{+0.57/19 2.6} h6 {-0.65/18 0.6} 13. h4 {+0.42/18 1.4} g6 {-0.82/15 0.4}
14. g5 {+0.13/22 6} Nh5 {-0.35/15 0.6} 15. Qf3 {+0.00/22 3} Na5
{+0.02/16 0.7} 16. gxh6 {-0.13/17 1.0} Nhf6 {+0.34/17 0.7} 17. Qgg3
{-0.12/17 0.9} gxf5 {+0.68/19 2.8} 18. exf5 {-0.26/17 1.2} Nb3
{+0.70/15 0.5} 19. a5 {-0.53/20 4} Kd8 {+0.87/17 2.0} 20. Qg7
{+0.04/14 0.4} Nd4 {+0.92/18 2.1} 21. Qff2 {-0.87/16 2.1} d5 {+1.28/18 3}
22. axb6 {-1.34/17 3} axb6 {+1.62/17 1.3} 23. Qxf7 {-1.04/17 1.5} Nd6
{+1.87/17 0.9} 24. Qg7 {-1.46/16 1.2} N6xf5 {+1.83/17 0.6} 25. Qgg1
{-1.46/16 0.8} Nfe6 {+1.83/19 2.8} 26. h7 {-1.82/17 1.2} Nxh7
{+1.90/19 0.6} 27. Kc1 {-1.77/14 0.5} Nf4 {+1.75/17 0.4} 28. Qef1
{-1.75/16 0.6} Nhf6 {+1.93/17 0.9} 29. Kb2 {-1.89/15 0.5} Kc8
{+2.07/17 0.9} 30. Qfh2 {-1.54/14 0.4} Nd6 {+2.28/19 2.2} 31. h5
{-1.77/14 0.6} N4xh5 {+2.76/19 3} 32. Ka1 {-2.02/14 0.6} Nef5
{+3.06/17 1.0} 33. Qd1 {-1.60/10 0.5} Nf4 {+3.56/16 1.0} 34. Qa2
{-1.85/10 0.6} dxc4 {+3.94/16 0.6} 35. dxc4 {-1.58/11 0.6} e4
{+4.51/16 0.6} 36. Qgh2 {-2.99/10 0.4} Nd3 {+4.41/15 0.6} 37. Qdd2
{-3.28/10 0.6} N7e5 {+6.10/16 0.7} 38. Qdg2 {-4.26/9 0.6} Nexc4
{+6.94/17 1.8} 39. Qa4 {-4.40/10 0.5} Na5 {+7.66/16 0.9} 40. Qga2
{-5.80/10 0.6} c4 {+8.99/18 2.0} 41. Qh8+ {-9.60/18 4} Nfe8 {+9.90/19 1.6}
42. Qh7 {-11.45/22 14} Ncxb5 {+11.76/19 1.9} 43. Qb2 {-13.92/23 3} Ndb3+
{+13.84/18 1.5} 44. Kb1 {-16.80/21 4} Nfd4 {+14.84/18 0.6} 45. Qaa2
{-20.08/23 7} Kb8 {+23.96/20 2.8} 46. Qh4 {-64.19/22 4} Nxb2
{+99.53/23 2.4} 47. Qxb2 {-99.58/21 1.6} c3 {+99.65/27 1.3} 48. Qa2
{-99.62/23 0.9} c2+ {+99.73/31 1.4} 49. Qxc2 {-99.72/29 0.7} Na3+
{+99.75/31 0.4} 50. Ka2 {-99.74/29 0.6} Naxc2 {+99.79/37 1.4} 51. Qd8+
{-99.80/32 0.8} Ka7 {+99.81/35 0.6} 52. Qd7+ {-99.82/35 0.6} Ka6
{+99.83/42 1.5} 53. Qd8 {-99.84/35 0.8} Nb4+ {+99.85/41 0.4} 54. Kb1
{-99.86/36 0.6} Ne2 {+99.87/49 1.6} 55. Qa8+ {-99.88/44 0.6} Kb5
{+99.89/1 0.5} 56. Qc8 {-99.90/48 0.6} Nd2+ {+99.91/65 1.4} 57. Ka1
{-99.92/53 0.6} Nxc8 {+99.93/68 0.4} 58. Kb2 {-99.94/1 0.6} Nac4+
{+99.95/127 0.6} 59. Ka1 {-99.96/1 0.6} Nb3+ {+99.97/127 0.6} 60. Kb1
{-99.98/1 0.6} Nc3# {+99.99/127 0.4}
{Xboard adjudication: Checkmate} 0-1[/pgn]

[Adam Hair]

Pawel has an elephantiasis correction for the queen in Rodent III:

If # of own queens > 1
then correction = -constant * (# of opp knights + # of opp bishops)

The default value of the constant is 4, which may be too low for this situation. I used 60 in the game above, though I have no idea if that was a good value.

This doesn't seem to work. I just played Rodent III against Stockfish 8, and they seem to agree that the score of the initial position is around +8 for the Queens.

I am assuming that you used Rodent III 0.172, which is the last official version from Pawel. I am not certain if it has the elephantiasis correction. And if it does, the constant is hardwired and would probably be 4. So the correction in this case would be 28cp.

I am using version 0.176, whose source I grabbed from github. I changed the constant to a parameter that I could access via "setoption".

Here is my compile of Rodent III 0.176 with a parameter "ELEPH" that changes the multiplier in the elephantiasis correction:

http://www.mediafire.com/file/ooviav8pq ... _0.176.zip

[Evert]

so what happens when you have an UCI engine correctly showing score from white point of view only?

By correctly you mean buggy, right?
What do you think happens? The GUI will report the inverted score when the engine is playing black.

watching scores has meaning only for humans, and those are accustomed to the anchor of the white side, in this way it is immediately clear who has the advantage.

Speak for yourself. I never think "geez, white's position is really good here", I think "crap, I'm totally screwed".
Either way, that's up to how you configure the GUI.

not to mention the case, when one engine shows score from white point of view, and another one from stm point of view: a total chaos.

Yeah. So don't do that. Follow the protocol.

probably, it is time to revise that in UCI.

UCI has issues. This is not one of them (and in fact is one of the things it does right compared to CECP, which AFAIK doesn't define the point of view of the score).

[Lyudmil Tsvetkov]

maybe you will share your code with us?

Sure. This is the functional change over the Stockfish master I checked out this morning:

Code: Select all

--- Stockfish/src/material.cpp	2017-07-18 13:21:30.000000000 +0200
+++ Nightfish/src/material.cpp	2017-07-18 23:05:43.000000000 +0200
@@ -121,6 +121,10 @@
     if  (pieceCount[Us][QUEEN] == 1 && pieceCount[Them][QUEEN] == 0)
          bonus += QueenMinorsImbalance[pieceCount[Them][KNIGHT] + pieceCount[Them][BISHOP]];
 
+    // Elephantiasis correction for Q vs N
+    if (pieceCount[Them][KNIGHT] > 2)
+       bonus -= 16*112*(pieceCount[Them][KNIGHT] - 2)*pieceCount[Us][QUEEN];
+
     return bonus;
   }

I had originally missed that Stockfish scales the imbalance correction by a factor 16 (so they are not in... whatever you'd call the evaluation unit in Stockfish). The factor "112" corresponds to 60cp. Note that this is not quite enough to reduce the evaluation in the root position to "balanced". Evaluation still favours the queens by about 100cp.

from what position?

The relevant start position,
[d]nnnnknnn/pppppppp/8/8/8/8/PPPPPPPP/1Q1KQ1Q1 w - -

how many games?

Six or seven at various time controls, but I'll start a longer series to run overnight.

what TC?

40/0:30. I ran a couple at 40/1:00. At longer time controls the Queens hang in there better (or perhaps it's the correction to the score, since otherwise the queens will think themselves superior enough to play recklessly), meaning they will draw when they can successfully close the position completely.

SF-revised playing both sides?

Of course.

can we look at some of the games?

Sure. Here's a couple; I adjudicated the first.
[pgn]
[Event "Computer Chess Game"]
[Site "vivaine.local"]
[Date "2017.07.18"]
[Round "1"]
[White "Nightfish 180717 64"]
[Black "Nightfish 180717 64"]
[Result "1/2-1/2"]
[TimeControl "40/60"]
[FEN "nnnnknnn/pppppppp/8/8/8/8/PPPPPPPP/1Q1KQ1Q1 w - - 0 1"]
[SetUp "1"]

{--------------
n n n n k n n n
p p p p p p p p
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
P P P P P P P P
. Q . K Q . Q .
white to play
--------------}
1. f4 {+1.59/17} e6 {-1.29/16 2.1} 2. e4 {+1.39/20 2.0} Nhg6 {-1.28/17 1.4}
3. f5 {+1.29/15 0.6} N6e7 {-1.37/18 1.6} 4. b4 {+1.24/17 1.4} d6
{-1.42/15 0.3} 5. a4 {+1.09/18 4} Nbd7 {-1.34/16 1.7} 6. d3 {+1.28/16 1.2}
h6 {-1.14/16 1.2} 7. g4 {+1.13/17 2.7} Ngf6 {-1.25/18 1.6} 8. Qb2
{+1.29/17 1.3} g5 {-1.14/17 2.6} 9. Qg2 {+1.23/18 2.4} e5 {-1.01/18 4} 10.
Qh3 {+1.04/18 2.6} Neg8 {-0.95/17 1.0} 11. Kc1 {+1.03/18 1.2} c6
{-0.85/18 1.7} 12. Qhg3 {+0.88/18 4} Nc7 {-0.74/16 1.0} 13. c3
{+0.77/19 2.3} N8h7 {-0.69/18 2.3} 14. Qee2 {+0.75/18 4} b6 {-0.80/17 1.4}
15. d4 {+0.68/16 1.4} Kf8 {-0.63/17 0.9} 16. Qed3 {+0.89/18 1.9} h5
{-0.79/19 2.3} 17. h3 {+0.55/19 3} h4 {-0.63/16 0.6} 18. Qgh2
{+0.50/15 0.4} Nce8 {-0.60/15 0.5} 19. Qbc2 {+0.51/17 4} Nge7
{-0.08/17 1.9} 20. dxe5 {+0.38/17 1.7} Nxe5 {-1.16/21 2.7} 21. Qxe5
{+0.36/16 0.5} dxe5 {-1.05/20 0.7} 22. Qxd8 {+0.83/18 0.6} Kg7
{-1.10/21 1.4} 23. Qh2 {+1.14/18 0.5} Ned6 {-0.84/18 0.6} 24. Qxe5
{+0.98/17 0.6} Nxe4 {-1.14/22 2.8} 25. a5 {+1.24/19 1.0} Nf2 {-0.79/23 2.7}
26. Qe3 {+1.25/20 0.6} N2e4 {-0.84/24 1.1} 27. c4 {+1.25/21 0.8} c5
{-0.65/23 0.4} 28. bxc5 {+1.08/23 1.5} Nxc5 {-0.31/21 0.7} 29. axb6
{+1.05/23 0.6} axb6 {-0.16/24 1.9} 30. Qc3 {+0.98/21 0.5} Ne4
{-0.16/21 0.4} 31. Qa3 {+0.80/23 1.2} Nf8 {+0.00/22 0.8} 32. Qa1
{+0.20/20 1.0} Kg8 {+0.00/25 1.1} 33. Kd1 {+0.00/22 1.1} Nd7 {+0.00/28 1.1}
34. Qd4 {+0.00/24 0.6} Nec5 {+0.00/29 1.1} 35. Ke2 {+0.00/25 0.5} f6
{+0.00/29 0.6} 36. Qe8 {+0.00/23 0.4} Kg7 {+0.00/31 1.5} 37. Qd2
{+0.00/27 0.5} Ne4 {+0.00/29 1.3} 38. Qd3 {-0.07/29 1.3} Ned6
{+0.08/27 0.9} 39. Qxe7+ {-0.17/21 0.7} Nxe7 {+0.05/36 2.4} 40. Qxd6
{+0.00/22 0.6} Ng8 {+0.20/33 0.6} 41. Qc7 {-0.26/26 0.6} Nh6 {+0.00/37 4}
42. Qa7 {-0.08/34 1.2} Kf7 {+0.00/46 1.9} 43. Qb7 {-0.08/40 1.6} Kg7
{+0.14/40 3} 44. Kf2 {-0.08/38 1.8} Kh7 {+0.14/38 1.8} 45. Qc7
{-0.08/39 1.7} Nf7 {+0.00/38 2.6} 46. Ke2 {-0.08/40 1.9} Kg7 {+0.06/34 0.6}
47. Qa7 {-0.08/39 1.4} Nde5 {+0.08/39 5} 48. Qxb6 {-0.12/36 1.8} Nxc4
{+0.08/37 0.5} 49. Qb7 {-0.12/38 1.5} Ne5 {+0.08/33 1.5} 50. Qe7
{-0.14/34 2.7} Nfd7 {+0.39/33 3} 51. Qe8 {-0.11/35 1.6} Nc5 {+0.55/33 0.8}
52. Kf1 {-0.10/31 0.6} Nb7 {+0.55/35 0.6} 53. Qe7 {-0.42/39 10} Nd8
{+0.39/43 11} 54. Kg1 {-0.08/31 0.8} Nec6 {+0.60/34 5} 55. Qc5
{-0.35/38 2.5} Nb8 {+0.60/34 2.2} 56. Qd5 {-0.47/42 1.6} Ndc6
{+0.60/42 0.7} 57. Qe6 {-0.47/41 0.6} Nfe5 {+0.60/40 0.7} 58. Qe8
{-0.47/37 1.0} Na6 {+0.60/42 0.7} 59. Qc8 {-0.47/41 1.0} Nab8
{+0.60/41 0.4} 60. Kf2 {-0.47/41 0.9} Nbd7 {+0.60/43 0.7} 61. Qe8
{-0.47/39 0.6} Nc5 {+0.60/43 0.4} 62. Kf1 {-0.47/43 1.2} Ncd3
{+0.60/45 0.7} 63. Kg2 {-0.47/43 1.0} Nf4+ {+0.60/43 0.5} 64. Kh2
{-0.47/44 0.6} Nd5 {+0.60/45 0.8} 65. Kg2 {-0.47/38 1.2} Nce7
{+0.60/42 0.9} 66. Kg1 {-0.47/42 1.0} Nf3+ {+0.60/43 0.7} 67. Kf1
{-0.47/44 1.1} Ne5 {+0.60/36 0.7} 68. Kg2 {-0.47/40 1.1} N7c6
{+0.60/39 0.7} 69. Qc8 {-0.47/41 1.2} Nf4+ {+0.60/36 0.6} 70. Kh2
{-0.47/39 1.0} Kf7 {+0.60/39 1.0} 71. Qb7+ {-0.47/35 0.9} Ne7
{+0.60/40 0.7} 72. Qa8 {-0.47/35 0.6} N7c6 {+0.60/39 0.7} 73. Qb7+
{-0.47/37 2.3} Ne7 {+0.60/40 0.5} 74. Qa8 {-0.47/36 0.6} Nfd5 {+0.60/40 3}
75. Qd8 {-0.63/36 2.7} Nf3+ {+0.60/38 0.5} 76. Kg2 {-0.63/35 0.9} Ne5
{+0.60/38 0.4} 77. Kg1 {-0.63/36 1.0} N5c6 {+0.60/37 0.8} 78. Qh8
{-0.63/37 1.0} Ne5 {+0.60/36 0.5} 79. Qd8 {-0.63/36 0.3} Kg7 {+0.60/40 1.3}
80. Kh2 {-0.54/37 3} Nf3+ {+0.60/37 0.7} 81. Kg2 {-0.54/33 2.3} Ne5
{+0.59/35 1.7} 82. Qe8 {+0.00/29 2.5}
{User adjudication} 1/2-1/2
[/pgn]

[pgn][Event "Computer Chess Game"]
[Site "vivaine.local"]
[Date "2017.07.18"]
[Round "3"]
[White "Nightfish 180717 64"]
[Black "Nightfish 180717 64"]
[Result "0-1"]
[TimeControl "40/60"]
[FEN "nnnnknnn/pppppppp/8/8/8/8/PPPPPPPP/1Q1KQ1Q1 w - - 0 1"]
[SetUp "1"]

{--------------
n n n n k n n n
p p p p p p p p
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
P P P P P P P P
. Q . K Q . Q .
white to play
--------------}
1. f4 {+1.59/17} e6 {-1.60/17 2.1} 2. e4 {+1.39/20 2.0} Nhg6 {-1.36/16 0.6}
3. f5 {+1.29/15 0.5} N6e7 {-1.33/18 1.7} 4. b4 {+1.24/17 1.4} d6
{-1.41/15 0.5} 5. a4 {+1.09/18 4} Nbd7 {-1.28/16 0.8} 6. d3 {+1.28/16 1.2}
b6 {-1.02/17 3} 7. c3 {+1.20/17 2.2} c5 {-1.08/18 3} 8. g4 {+1.09/17 1.6}
e5 {-1.03/19 3} 9. b5 {+1.08/17 1.3} Nc7 {-0.88/21 4} 10. c4 {+1.03/19 1.3}
Ngf6 {-0.91/20 1.1} 11. Qa2 {+0.74/21 7} Nb7 {-0.84/20 0.8} 12. Qae2
{+0.57/19 2.6} h6 {-0.65/18 0.6} 13. h4 {+0.42/18 1.4} g6 {-0.82/15 0.4}
14. g5 {+0.13/22 6} Nh5 {-0.35/15 0.6} 15. Qf3 {+0.00/22 3} Na5
{+0.02/16 0.7} 16. gxh6 {-0.13/17 1.0} Nhf6 {+0.34/17 0.7} 17. Qgg3
{-0.12/17 0.9} gxf5 {+0.68/19 2.8} 18. exf5 {-0.26/17 1.2} Nb3
{+0.70/15 0.5} 19. a5 {-0.53/20 4} Kd8 {+0.87/17 2.0} 20. Qg7
{+0.04/14 0.4} Nd4 {+0.92/18 2.1} 21. Qff2 {-0.87/16 2.1} d5 {+1.28/18 3}
22. axb6 {-1.34/17 3} axb6 {+1.62/17 1.3} 23. Qxf7 {-1.04/17 1.5} Nd6
{+1.87/17 0.9} 24. Qg7 {-1.46/16 1.2} N6xf5 {+1.83/17 0.6} 25. Qgg1
{-1.46/16 0.8} Nfe6 {+1.83/19 2.8} 26. h7 {-1.82/17 1.2} Nxh7
{+1.90/19 0.6} 27. Kc1 {-1.77/14 0.5} Nf4 {+1.75/17 0.4} 28. Qef1
{-1.75/16 0.6} Nhf6 {+1.93/17 0.9} 29. Kb2 {-1.89/15 0.5} Kc8
{+2.07/17 0.9} 30. Qfh2 {-1.54/14 0.4} Nd6 {+2.28/19 2.2} 31. h5
{-1.77/14 0.6} N4xh5 {+2.76/19 3} 32. Ka1 {-2.02/14 0.6} Nef5
{+3.06/17 1.0} 33. Qd1 {-1.60/10 0.5} Nf4 {+3.56/16 1.0} 34. Qa2
{-1.85/10 0.6} dxc4 {+3.94/16 0.6} 35. dxc4 {-1.58/11 0.6} e4
{+4.51/16 0.6} 36. Qgh2 {-2.99/10 0.4} Nd3 {+4.41/15 0.6} 37. Qdd2
{-3.28/10 0.6} N7e5 {+6.10/16 0.7} 38. Qdg2 {-4.26/9 0.6} Nexc4
{+6.94/17 1.8} 39. Qa4 {-4.40/10 0.5} Na5 {+7.66/16 0.9} 40. Qga2
{-5.80/10 0.6} c4 {+8.99/18 2.0} 41. Qh8+ {-9.60/18 4} Nfe8 {+9.90/19 1.6}
42. Qh7 {-11.45/22 14} Ncxb5 {+11.76/19 1.9} 43. Qb2 {-13.92/23 3} Ndb3+
{+13.84/18 1.5} 44. Kb1 {-16.80/21 4} Nfd4 {+14.84/18 0.6} 45. Qaa2
{-20.08/23 7} Kb8 {+23.96/20 2.8} 46. Qh4 {-64.19/22 4} Nxb2
{+99.53/23 2.4} 47. Qxb2 {-99.58/21 1.6} c3 {+99.65/27 1.3} 48. Qa2
{-99.62/23 0.9} c2+ {+99.73/31 1.4} 49. Qxc2 {-99.72/29 0.7} Na3+
{+99.75/31 0.4} 50. Ka2 {-99.74/29 0.6} Naxc2 {+99.79/37 1.4} 51. Qd8+
{-99.80/32 0.8} Ka7 {+99.81/35 0.6} 52. Qd7+ {-99.82/35 0.6} Ka6
{+99.83/42 1.5} 53. Qd8 {-99.84/35 0.8} Nb4+ {+99.85/41 0.4} 54. Kb1
{-99.86/36 0.6} Ne2 {+99.87/49 1.6} 55. Qa8+ {-99.88/44 0.6} Kb5
{+99.89/1 0.5} 56. Qc8 {-99.90/48 0.6} Nd2+ {+99.91/65 1.4} 57. Ka1
{-99.92/53 0.6} Nxc8 {+99.93/68 0.4} 58. Kb2 {-99.94/1 0.6} Nac4+
{+99.95/127 0.6} 59. Ka1 {-99.96/1 0.6} Nb3+ {+99.97/127 0.6} 60. Kb1
{-99.98/1 0.6} Nc3# {+99.99/127 0.4}
{Xboard adjudication: Checkmate} 0-1[/pgn]

thanks.

I see half of the games you post end in a draw, so where is the knight side advantage?

in my opinion, any TC shorter than 1 min. + small increment is simply detrimental both to engine development, as well as general positional assessment. so, a bit longer TC would do fine.

of course, we need at least 30-40 games, to have even very basic statistical relevance. in my sample, queens also started with couple of losses, but then scored more wins. will be waiting for your new(1000 games?) series.

you picked up a good starting position.
of course, every single game could be drawn, provided the queens close the entire board. that is why I said pawns restrain knights more, because they are more in number.

interesting what happens, if the el.value is set to 90 or 130?

since when did SF integrate queen vs minors imbalance?

no info about it on the dev page.

[Lyudmil Tsvetkov]

I will visit when you get more than 100 simultaneous viewers.

How would you know I have them? How do you know I don't have them now? How would I know that?

BTW, 11-2 for QueeNy, so far. So SF 8 manages to win some games. Often through a recognizable stupidity of QueeNy. Like putting a Knight on h1, and then (blocked) Pawns on f7, g6. Seems a draw is coming (2Q vs 4N, no Pawns left).

that is a big progress, man!

from 'Queeny wins 100%', meaning 100/100, here we already have 11/13, which is only about 80+%.

2Qs vs 4Ns should be won or close to won for queen side.

maybe you adjudicate in some way detrimental to queens?

but most importantly, as I told you, you can not compare apples and oranges.
why match an engine that knows about the imbalance and another one, that is clueless about it?

please, grab Evert's revised SF version, or maybe Evert can grab Queeny, to do more meaningful comparison.

[Lyudmil Tsvetkov]

I now changed all probing (for new game moves, new PVs, and new chat messages) by the viewer JavaScript to asynchronous calls. Indeed the loss of characters that previously occurred when typing in the chat now has disappeared.

In Firefox on Linux (which seems to use much larger font than on Windows?!) the upper part of the page looks like this:



Below that (not shown, as I could not fit it on the small screen of my Linux laptop) there are buttons to fetch a list with available games from the server, so you can select the one you want to view from that list.

that is not replayable, does not concern me.

does not that pawn on h6 promote soon?

[Lyudmil Tsvetkov]

so what happens when you have an UCI engine correctly showing score from white point of view only?

The same as what happens when you correctly misspell a word.

watching scores has meaning only for humans, and those are accustomed to the anchor of the white side, in this way it is immediately clear who has the advantage.

But humans are not supposed to watch raw engine output. UCI is intended for communication with the GUI, not with humans. This is also why it uses the human-unfriendly long-algebraic move notation.

not to mention the case, when one engine shows score from white point of view, and another one from stm point of view: a total chaos.

This is why it is important to stick to the standard.

I am watching raw engine output all the time, comparing evals.

and, with 300cps score from the point of view of white, one engine showing +300cps, while the other one -300cps, or maybe, one 250cps and the other -198cps, when the scores stand to each other in the notation, is really a maddening experience.

your mind should divert a substantial portion of valuable resources for conversion purposes, something the engine or gui should have already done.

showing scores in this way is extremely misleading.

I don't have any issues with SAN or any other algebraic notation, unless figurines are added. both engines show very same standard after all, which is not misleading.

I don't know what you have in mind by mentioning MissPelling?