On-line engine blitz tourney November

Discussion of chess software programming and technical issues.

Moderators: hgm, Rebel, chrisw

Joost Buijs
Posts: 1563
Joined: Thu Jul 16, 2009 10:47 am
Location: Almere, The Netherlands

On-line engine blitz tourney November

Post by Joost Buijs »

The monthly on-line engine blitz tourney for November will take place on:

Saturday November 14, 2:00 PM EST (Boston time), 20:00 CET (Amsterdam time)

To connect:

winboard -zp -ics -icshost nightmare-chess.nl -icshelper timeseal -fcp ENGINE.exe -fd ENGINEFOLDER -autoKibitz

(for UCI engines, add -fUCI)

People without a server account can submit a request for a (new) account to admin@nightmare-chess.nl

9 rounds 5/1, our tournament manager will be: Eendje (Lisebeth) from ICC
Joost Buijs
Posts: 1563
Joined: Thu Jul 16, 2009 10:47 am
Location: Almere, The Netherlands

Re: On-line engine blitz tourney November

Post by Joost Buijs »

It was fun as usual! At least I think it was fun.

Schooner clearly ended first place, my guess is that he added NNUE like many others are lately doing.
Belfish ended second, no surprise here. Marvin seems to have made a lot of progress and ended third.

Code: Select all

:Tourney Players: Round 9 of 9 
:
:     Name              Rating Score Perfrm Upset  Results 
:     ----------------- ------ ----- ------ ------ ------- 
:  1 +Schooner          [2912]  7.5  [2995] [  46] +10b +08w =05w +03b =02w +04w +12b +07w =06b 
:  2 +BelFish           [3005]  7.0  [2934] [   0] +15w +07w =04b +05w =01b =06w +09b =03w +12b 
:  3 +Marvin            [2747]  6.5  [2809] [ 202] +23w =12w +09b -01w +11b +13w =04b =02b +14w 
:  4 +ArasanX           [2893]  6.0  [2847] [  56] +11b +13w =02w =06b +14w -01b =03w +12w =05b 
:  5 +WaspX             [2832]  6.0  [2859] [  70] +18w +14w =01b -02b +13w =07b =06b +08w =04w 
:  6 +NightmareX        [2822]  6.0  [2825] [ 176] +16w =09w =12b =04w +08b =02b =05w +14b =01w 
:  7 +Goldbar           [2717]  5.5  [2670] [  57] +17w -02b +11w =12w +18b =05w +13b -01b =09w 
:  8 +TheBaron          [2649]  5.0  [2560] [   0] +19b -01b +18w -10b -06w +11w +15b -05b +13w
:  9 +atomNightmare     [2559]  5.0  [2542] [ 210] +22w =06b -03w -13b +17b +10w -02w +19b =07b 
: 10 +Nemeton           [2343]  5.0  [2515] [ 342] -01w +19w +17b +08w -12b -09b +18w -13w +15b 
: 11 +rpiArminius       [2306]  5.0  [2414] [   0] -04w +20b -07b +23w -03w -08b +24b +22w +19w 
: 12 +atomGoldbar       [2549]  4.5  [2664] [ 383] +24w =03b =06w =07b +10w +14b -01w -04b -02w 
: 13 +WaDuuttie         [2645]  4.0  [2520] [   0] +20w -04b +16w +09w -05b -03b -07w +10b -08b 
: 14 +Arminius          [2613]  4.0  [2467] [   0] +21w -05b +23w +15b -04b -12w +20b -06w -03b 
: 15 +Jumbo             [2379]  4.0  [2328] [   0] -02b =17w +21b -14w =23b +16w -08w +20b -10w 
: 16 +CTdccc            [2217]  4.0  [2212] [  61] -06b +22w -13b -17w +24b -15b -19w +18b +21w
: 17 +Joker             [2144]  4.0  [2207] [ 190] -07b =15b -10w +16b -09w -19b +21w =24w +22b 
: 18 +Barbarossa        [2278]  3.5  [2252] [   0] -05b +21w -08b +20w -07w +23b -10b -16w =24b 
: 19 +Isa               [2126]  3.5  [2170] [ 109] -08w -10b =22b +21w -20b +17w +16b -09w -11b 
: 20 +feligres          [2123]  3.5  [2179] [  44] -13b -11w +24b -18b +19w +22b -14w -15w =23b 
: 21 +KingSlayer        [2055]  3.0  [2069] [ 150] -14b -18b -15w -19b +22w +24w -17b +23w -16b 
: 22 +microMax          [1937]  2.5  [2007] [ 362] -09b -16b =19w +24w -21b -20w +23w -11b -17w 
: 23 +Spartacus         [2205]  2.0  [2040] [  87] -03b +24w -14b -11b =15w -18w -22b -21b =20w 
: 24 +atomSargonI       [1930]  1.0  [1890] [ 281] -12b -23b -20w -22b -16w -21b -11w =17b =18w
D Sceviour
Posts: 570
Joined: Mon Jul 20, 2015 5:06 pm

Re: On-line engine blitz tourney November

Post by D Sceviour »

This is the second time Schooner has won the monthly on-line tournament. The only change made since the last month was to reduce the hash usage size in the settings. Previously, I used 4 Gb / 8 threads for 512 Mb / thread. This was cut in half to 2 GB total usage, not including pawn hash. The smaller hash size seems to produce a faster access on my machine and there was a noticeable improvement in depth.

Bob Hyatt had once produced a formula to maximize hash usage. I remember experimenting with this some time ago but it is not installed in the latest versions of Crafty. It is not really necessary. The only important thing to remember is that too much hash size can be a bad thing.

I have been experimenting with NNUE but not with this tournament. The NNUE results so far have been disappointing. It may be some time before re-considering NNUE use (if ever).

Here are the Schooner games for the November 14, 2020 tournament:
[pgn]
[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "Nemeton"]
[Black "Schooner"]
[Result "0-1"]
[WhiteElo "2346"]
[BlackElo "2912"]
[TimeControl "300+1"]

1. e4 e5 2. Nf3 Nc6 {+0.23/17} 3. Bc4 Be7 {+0.15/18} 4. O-O d6 {+0.29/26}
5. d4 exd4 {+0.38/26} 6. Nxd4 Nf6 {+0.43/26} 7. Nxc6 bxc6 {+0.37/24} 8. Re1
O-O {-0.21/25 7} 9. Nc3 Re8 {+0.22/20} 10. Bf4 Bg4 {+0.53/26} 11. Qd3 Nd7
{-0.20/25 7} 12. Qg3 Bh5 {-0.24/28 9} 13. Rab1 Bf6 {+0.01/25 18} 14. b4 Bg6
{+0.26/24 16} 15. Bd2 a5 {+0.43/25 6} 16. b5 Ne5 {+0.58/28 7} 17. Bf1 cxb5
{+0.51/26 6} 18. Bxb5 c6 {+0.57/27 6} 19. Bf1 Bh4 {+0.57/25 6} 20. Qf4 h6
{+0.61/28 5} 21. Be3 Qe7 {+0.62/27 16} 22. Rb6 Bf6 {+0.71/26 7} 23. Reb1
Nd7 {+0.94/30 7} 24. Rxc6 Be5 {+0.80/31 14} 25. Qf3 Bxh2+ {+0.89/30 7} 26.
Kxh2 Ne5 {+0.90/28 11} 27. Qf4 Nxc6 {+0.74/28 4} 28. Bb5 Rec8 {+0.72/30 7}
29. Nd5 Qxe4 {+0.68/29 7} 30. Bxc6 Rxc6 {+0.79/33 3} 31. Qxe4 Bxe4
{+0.49/34 26} 32. Ne7+ Kf8 {+0.49/31 3} 33. Nxc6 Bxc6 {+0.48/30 6} 34. c4
Ke7 {+0.66/25 4} 35. Bb6 a4 {+0.98/24 3} 36. Bd4 f6 {+0.97/23 5} 37. Kg3 h5
{+1.05/24 6} 38. Re1+ Kf7 {+1.09/26 4} 39. Rb1 Rc8 {+1.20/24 3} 40. Be3 g5
{+1.25/25 2.6} 41. f3 h4+ {+1.28/27 7} 42. Kf2 Be8 {+1.19/28 9} 43. Rc1 Rb8
{+1.25/27 4} 44. Rd1 Ke6 {+1.32/26 4} 45. Bc1 Rc8 {+1.39/26 3} 46. Rd4 Bf7
{+1.35/27 6} 47. Ba3 Rc6 {+1.26/29 1.9} 48. Ke2 Bg6 {+1.38/24 2.9} 49. Bb4
Bb1 {+1.33/30 3} 50. a3 Ba2 {+1.39/30 2.9} 51. Kd3 d5 {+1.35/29 5} 52.
cxd5+ Bxd5 {+1.32/29 2.0} 53. Bc3 Bb3 {+1.32/28 3} 54. Re4+ Kf7
{+1.62/27 3} 55. Rd4 Kg6 {+1.72/31 7} 56. Rb4 f5 {+1.96/28 4} 57. Rd4 Kh5
{+2.09/30 1.5} 58. Kd2 f4 {+2.12/27 1.5} 59. Kd3 Bd1 {+2.55/26 1.8} 60. Re4
h3 {+2.68/27 1.3} 61. Re1 hxg2 {+10.47/27 4} 62. Bd4 Bxf3 {+11.53/27 4} 63.
Bg1 Rd6+ {+11.76/27 1.6} 64. Kc2 g4 {+12.87/26 2.0} 65. Kc3 g3
{+14.45/25 2.0} 66. Re5+ Kh4 {+17.98/28 5} 67. Re1 Bg4 {+21.46/26 0.9} 68.
Re8 f3 {+30.64/28 6} 69. Rh8+ Bh5 {+26.54/24 0.8} 70. Rf8 Kh3
{+31.25/26 2.7} 71. Rf5 Kg4 {+37.35/22 0.8} 72. Rf8 Bg6 {+31.08/21 0.7} 73.
Kb4 Bf5 {+29.81/19 6} 74. Bc5 Rg6 {+327.42/19 0.8} 75. Rd8 f2
{+327.50/19 1.1} 76. Ra8 g1=Q {+327.54/17 0.8} 77. Rxa4 Bc2
{+327.56/20 0.9} 78. Kc3+ Bxa4 {+327.58/33 0.6} 79. Bxf2 gxf2
{+327.60/42 0.6} 80. Kd3 f1=Q+ {+327.62/99 0.4} 81. Kc3 Rc6+
{+327.64/99 0.1} 82. Kb2 Qb1# {+327.66/99 0.1}
{Nemeton checkmated} 0-1

[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "Schooner"]
[Black "TheBaron"]
[Result "1-0"]
[WhiteElo "2912"]
[BlackElo "2650"]
[TimeControl "300+1"]

1. d4 d5 2. c4 c6 3. cxd5 {-0.12/16} cxd5 4. Bf4 {-0.25/20} Qb6 5. Nc3
{+0.08/20} Nf6 6. e3 {+0.06/21} Qxb2 7. Bb5+ {-0.11/21} Nc6 8. Nge2
{+0.00/21} Bf5 9. O-O {+0.00/22} a6 10. e4 {+0.01/25} Bxe4 11. Bxc6+
{-0.52/28} bxc6 12. Qa4 {-0.33/29} Qb7 13. Nxe4 {-0.33/29} Nxe4 14. Rfc1
{-0.59/30} Qb5 15. Qc2 {-0.81/31} g5 16. Rab1 {+1.50/34 9} Qc4 17. Qd1
{+1.72/36 9} Qxa2 18. Ra1 {+1.98/35 23} Qxa1 19. Rxa1 {+1.96/34 6} gxf4 20.
Qa4 {+2.03/34 9} Kd7 21. Nxf4 {+2.40/32 7} e6 22. Rc1 {+2.61/32 10} Rc8 23.
Nd3 {+2.61/32 5} f6 24. Rb1 {+2.83/32 6} Bd6 25. f3 {+3.14/36 15} Ng5 26.
Rb7+ {+3.21/38 17} Rc7 27. Rxc7+ {+3.44/37 8} Kxc7 28. Qxa6 {+3.72/37 35}
Rb8 29. Qa5+ {+3.73/33 5} Kd7 30. Qa7+ {+3.72/34 4} Ke8 31. Qg7
{+3.46/33 6} Rb3 32. Nc5 {+3.96/32 6} Rb1+ 33. Kf2 Nf7 34. Qxf6
{+4.62/30 6} Rb2+ 35. Ke3 {+5.16/31 21} Bxc5 36. dxc5 {+5.37/32 4} Rb7 37.
g4 {+5.45/30 10} Ra7 38. h4 {+6.11/29 7} Kf8 39. g5 {+6.64/28 5} Ra5 40.
Qxe6 {+7.05/25 4} Rxc5 41. f4 {+7.80/29 17} d4+ 42. Kd2 {+8.39/27 6} d3 43.
h5 {+10.71/28 3} Rc2+ 44. Kxd3 {+12.99/30 2.6} Rh2 45. g6 {+17.69/31 1.8}
hxg6 46. hxg6 {+30.39/25 23} Nh8 47. f5 {+327.34/21 1.6} Rh6 48. Qd6+
{+327.46/24 3} Kg8 49. Qd8+ {+327.48/27 1.9} Kg7 50. Qe7+ {+327.50/29 2.1}
Kg8 51. g7 {+327.52/31 1.4} Kh7 52. Qf8 {+327.54/30 1.2} Rh3+ 53. Ke4
{+327.56/68 1.2} Rh4+ 54. Kf3 {+327.58/99 0.2} Rh3+ 55. Kg4
{+327.60/99 0.2} Rg3+ 56. Kxg3 {+327.62/99 0.2} Kh6 57. gxh8=Q+
{+327.64/99 0.2} Kg5 58. Qff6#
{TheBaron checkmated} 1-0

[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "Schooner"]
[Black "WaspX"]
[Result "1/2-1/2"]
[WhiteElo "2914"]
[BlackElo "2834"]
[TimeControl "300+1"]

1. d4 Nf6 2. c4 e6 3. g3 {-0.23/18} d5 4. Bg2 {+0.12/17} Be7 5. Nf3
{+0.01/18} O-O 6. O-O {-0.02/17} dxc4 7. Qc2 {-0.06/18} a6 8. a4 {-0.07/18}
Bd7 9. Qxc4 {-0.07/19} Bc6 10. Bg5 {-0.01/19} Nbd7 11. Rd1 {-0.08/16} Bd5
{+0.00/30 7} 12. Qd3 {+0.00/30 35} Be4 {+0.00/30 9} 13. Qb3 {+0.00/30 8}
Bd5 {+0.00/32 18} 14. Qc2 {+0.00/31 31} Be4 {+0.00/34 16} 15. Qb3
{+0.00/33 1:20} Bd5 {+0.00/35 6} 16. Qc2 {+0.00/32 4} Be4 {+0.00/34 7} 17.
Qb3 {+0.00/31 11}
{Game drawn by repetition} 1/2-1/2

[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "Marvin"]
[Black "Schooner"]
[Result "0-1"]
[WhiteElo "2747"]
[BlackElo "2913"]
[TimeControl "300+1"]

1. e4 e5 2. Nf3 Nc6 {+0.26/16} 3. Bc4 Be7 {+0.35/17} 4. O-O d6 {+0.23/20}
5. d4 exd4 {+0.34/26} 6. Nxd4 Nf6 {+0.35/26} 7. Nc3 O-O {+0.31/26} 8. Bf4
Ne5 {+0.41/23} 9. Bb3 a6 {+0.42/24} 10. a4 Nfd7 {-0.38/25 16} 11. Be3 Nc5
{-0.49/28 13} 12. Ba2 a5 {-0.53/28 9} 13. Qd2 Ng4 {-0.41/24 6} 14. Bf4 Ne5
{-0.41/23 11} 15. Rad1 Bd7 {-0.30/24 7} 16. b3 h6 {-0.06/26 7} 17. f3 Bg5
{+0.00/29 9} 18. Bxg5 hxg5 {+0.01/29 6} 19. h3 Ne6 {+0.00/28 6} 20. Nf5 Nf4
{+0.29/26 6} 21. Ne3 Re8 {+0.33/27 11} 22. Kh2 Be6 {+0.29/25 5} 23. Ng4
Bxg4 {+0.51/27 12} 24. fxg4 c6 {+0.51/28 6} 25. Qxd6 Qb6 {+0.56/31 9} 26.
Qd4 Qxd4 {+0.47/32 8} 27. Rxd4 Ne6 {+0.46/31 11} 28. Rd2 Rad8 {+0.39/31 4}
29. Rfd1 Rxd2 {+0.38/32 17} 30. Rxd2 Kf8 {+0.34/30 7} 31. Kg3 Ke7
{+0.43/28 4} 32. Kf2 g6 {+0.37/28 10} 33. Ne2 Nd7 {+0.29/29 5} 34. Ke3 Kd8
{+0.36/27 8} 35. Nc3 Kc7 {+0.55/24 2.7} 36. Bb1 Nec5 {+0.62/27 5} 37. Rf2
Re7 {+0.65/25 3} 38. Ba2 b6 {+0.63/29 6} 39. Bb1 Nb7 {+0.55/24 2.5} 40. Kd2
Nd6 {+0.68/25 7} 41. Re2 Re8 {+0.65/26 4} 42. Re3 Re5 {+0.65/24 2.6} 43.
Kd3 Kd8 {+0.81/25 2.3} 44. Re2 Nc5+ {+0.77/25 4} 45. Ke3 Ke7 {+0.74/28 5}
46. Kf3 Kf6 {+0.77/29 14} 47. g3 Re8 {+0.86/28 12} 48. h4 Nd7 {+1.13/25 3}
49. Kg2 Ne5 {+1.00/26 1.6} 50. Rf2+ Kg7 {+1.18/27 1.8} 51. Rd2 Nb7
{+1.37/24 2.7} 52. Ba2 gxh4 {+1.40/27 6} 53. gxh4 Nxg4 {+1.24/29 4} 54. Kf3
Ne5+ {+1.70/27 5} 55. Kg3 Nc5 {+1.81/28 7} 56. Kf4 f6 {+1.94/23 1.0} 57.
Kg3 Rh8 {+1.83/26 2.7} 58. Re2 Rh5 {+1.84/24 2.7} 59. Re1 g5 {+2.06/24 1.6}
60. hxg5 fxg5 {+2.19/26 1.7} 61. Nd1 Rh4 {+1.91/23 2.2} 62. Nf2 Kf6
{+1.88/23 0.9} 63. Bb1 Rf4 {+2.05/25 0.9} 64. Rd1 Rf3+ {+2.06/23 0.9} 65.
Kg2 Re3 {+2.27/27 1.0} 66. Rf1 Ke7 {+2.48/26 0.9} 67. Nd1 Re2+
{+2.51/26 1.1} 68. Rf2 Rxe4 {+2.47/26 1.0} 69. Rd2 Rg4+ {+2.67/24 1.0} 70.
Kf1 Rf4+ {+2.72/25 1.4} 71. Rf2 Nf3 {+2.69/27 1.1} 72. Kg2 Nd4 {+2.88/26 4}
73. Rd2 Rg4+ {+3.00/27 1.3} 74. Kf1 Ne4 {+2.95/28 0.9} 75. Rd3 Rf4+
{+2.97/28 6} 76. Kg2 Nf5 {+3.18/26 0.9} 77. Kg1 Nh4 {+3.73/24 2.1} 78. Re3
Kd6 {+3.73/26 4} 79. c3 Nd2 {+3.86/25 0.6} 80. Bd3 Nxb3 {+3.80/22 0.7} 81.
Nb2 Nc5 {+3.81/22 0.6} 82. Nc4+ Kc7 {+3.96/24 1.0} 83. Be2 Nf5
{+4.35/25 2.8} 84. Rh3 Re4 {+4.57/24 0.8} 85. Rh7+ Re7 {+4.64/26 2.4} 86.
Rh2 Nxa4 {+5.50/24 0.6} 87. Bg4 Nh4 {+5.88/26 1.4} 88. Rf2 Nxc3
{+6.04/25 1.8} 89. Rf8 Ne4 {+6.25/26 1.5} 90. Rh8 Nd6 {+6.56/23 0.7} 91.
Nxd6 Kxd6 {+6.93/24 0.8} 92. Ra8 Re4 {+7.01/24 1.5} 93. Bd1 b5
{+7.25/25 1.5} 94. Kh2 a4 {+7.92/23 0.6} 95. Kh3 Nf5 {+8.26/22 0.7} 96. Bg4
Re3+ {+9.48/24 0.8} 97. Kh2 Nd4 {+9.93/25 0.7} 98. Bd1 a3 {+10.35/23 0.4}
99. Kg2 b4 {+10.66/26 0.9} 100. Bh5 c5 {+10.75/26 0.5} 101. Bf7 Ne6
{+11.45/26 1.8} 102. Kf2 Re5 {+12.00/25 0.5} 103. Bg6 Kd5 {+12.97/21 0.7}
104. Ra5 Kd4 {+14.09/23 1.3} 105. Ra8 Kc3 {+14.80/24 1.5} 106. Ra7 Kb2
{+20.63/23 0.7} 107. Bf7 a2 {+32.37/25 1.5} 108. Bh5 a1=Q {+327.24/19 0.5}
109. Rxa1 Kxa1 {+327.44/21 0.7} 110. Bd1 c4 {+327.46/23 1.3} 111. Kf1 b3
{+327.50/22 0.6} 112. Kf2 Nd4 {+327.52/22 0.6} 113. Bh5 c3 {+327.54/22 0.4}
114. Kg2 c2 {+327.56/22 0.5} 115. Bf3 Nxf3 {+327.58/28 0.6} 116. Kh3 c1=Q
{+327.60/78 0.5} 117. Kg4 Qh1 {+327.62/99 0.2} 118. Kg3 Qg1+
{+327.64/99 0.2} 119. Kh3 g4# {+327.66/99 0.1}
{Marvin checkmated} 0-1

[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "Schooner"]
[Black "BelFish"]
[Result "1/2-1/2"]
[WhiteElo "2916"]
[BlackElo "3008"]
[TimeControl "300+1"]

1. d4 Nf6 2. c4 e6 3. g3 {-0.23/19} d5 4. Bg2 {+0.22/17} Bb4+ 5. Nd2
{+0.06/17} O-O 6. Nf3 {+0.10/18} b6 7. O-O {+0.03/17} Bb7 8. b3 {-0.16/16}
c5 9. Bb2 {-0.06/14} Qe7 10. dxc5 {-0.15/16} Bxc5 11. cxd5 {-0.03/22} Bxd5
12. e3 {-0.07/26} Nbd7 13. Qe2 {+0.00/19} a5 14. Rac1 {-0.10/18} a4 15.
Rfd1 {+0.11/24 10} Rfc8 16. Ne5 {+0.09/27 18} a3 17. Ba1 {+0.00/29 7} Nxe5
18. Bxe5 {+0.00/33 15} Bxg2 19. Kxg2 {+0.00/33 6} Qb7+ 20. e4 {+0.00/30 6}
Nd7 21. Bc3 {+0.00/28 6} Bf8 22. Nc4 {+0.26/30 19} Nc5 23. f3 {+0.08/30 5}
Rd8 24. e5 {+0.40/28 5} Qa6 25. Qe3 {+0.32/30 6} Rxd1 26. Rxd1 {+0.24/30 7}
Na4 27. Bd4 {+0.21/32 12} Nb2 28. Bxb2 {+0.21/32 5} axb2 29. a4
{+0.12/33 10} b5 30. Nxb2 {+0.09/31 10} bxa4 31. Nxa4 {+0.09/33 5} Rc8 32.
Kh3 {+0.08/33 6} Rc2 33. Qe4 {+0.00/35 7} Ra2 34. f4 {+0.00/32 19} Qc8 35.
Re1 {+0.00/32 4} g6 36. Re2 {+0.00/34 5} Ra3 37. Rb2 {+0.00/34 3} Be7 38.
Nb6 {+0.00/32 7} Qc3 39. Rc2 {+0.00/33 3} Qxb3 40. Nc8 {+0.00/35 4} Bf8 41.
Nd6 {+0.00/35 6} Qb1 42. Qe2 {+0.00/35 5} Re3 43. Qf2 {+0.00/35 3} Rb3 44.
Rc8 {+0.00/33 4} Qd1 45. Ne4 {+0.00/33 3} Kg7 46. Rc7 {+0.00/35 2.5} Qh5+
47. Kg2 Qd1 48. Nf6 {+0.00/37 4} Rd3 49. Ne8+ {+0.00/36 3} Kg8 50. Nf6+
{+0.00/40 2.6} Kg7 51. Ne8+ {+0.00/43 11} Kh8 52. Rxf7 {+0.00/37 2.5} Bg7
53. Nxg7 {+0.00/40 3} Rd2 54. Qxd2 {+0.00/43 2.4} Qxd2+ 55. Kh3
{+0.00/46 6} Kg8 56. Rc7 {+0.00/46 9} Qd3 57. Ne8 {+0.00/36 2.1} Qf5+ 58.
g4 {+0.00/40 2.2} Qxf4 59. Nf6+ {+0.00/49 1.8} Kf8 60. Rxh7 {+0.00/53 3} g5
61. Kg2 {+0.00/54 7} Qe3 62. Kh1 {+0.00/57 2.1} Qf2 63. h3 {-0.19/39 1.8}
Qe1+ 64. Kg2 {+0.00/68 4} Qa1 65. Kf2 {+0.00/45 1.8} Qc3 66. Kg2
{+0.00/39 1.8} Qd2+ 67. Kh1 {+0.00/61 1.4} Qf2 68. Rc7 {+0.00/59 1.5} Qe1+
69. Kg2 {+0.00/66 2.0} Qd2+ 70. Kg1 {+0.00/55 3} Qe1+ 71. Kg2 {+0.00/96 4}
Qd2+ 72. Kf1 {+0.00/58 4} Qf4+ 73. Kg1 {+0.00/54 1.6} Qg3+ 74. Kh1
{+0.00/60 1.3} Qxh3+ 75. Kg1 Qg3+ 76. Kh1 {+0.00/72 4} Qf2 77. Rh7
{+0.00/64 1.2} Qe1+ 78. Kg2 {+0.00/76 1.2} Qe2+ 79. Kh1 {+0.00/72 3} Qf1+
80. Kh2 Qf3 81. Kg1 {+0.00/66 1.4} Qc6 82. Kh2 {+0.00/66 1.3} Qc1 83. Kg2
{+0.00/68 11} Qe3 84. Rc7 {+0.00/62 0.9} Qe2+ 85. Kg1 {+0.00/69 1.3} Qd1+
86. Kf2 {+0.00/61 4} Qh1 87. Rh7 {+0.00/50 1.3} Qc6 88. Kg3 {+0.00/56 0.8}
Qc2 89. Kh3 {+0.00/50 1.2} Qf2 90. Rc7 {+0.00/57 2.0} Qe1 91. Kg2
{+0.00/55 0.9} Qe2+ 92. Kh3 {+0.00/74 1.1} Qd1 93. Kg2 {+0.00/63 0.9} Qa1
94. Rh7 {+0.00/51 0.8} Qd4 95. Kh3 {+0.00/58 0.8} Qa4 96. Rc7
{+0.00/56 0.8} Qa1 97. Rh7 {+0.00/54 1.4} Qa2 98. Rc7 {+0.00/59 1.0} Qa1
99. Rh7 {+0.00/59 6} Qa2 100. Rc7 {+0.00/64 1.1} Qe2 101. Rh7
{+0.00/63 1.4} Qd1 102. Kh2 {+0.00/64 0.7} Qb3 103. Rc7 {+0.00/54 0.8} Qb1
104. Kh3 {+0.00/53 0.7} Qe1 105. Kg2 {+0.00/60 0.6} Qd1 106. Kf2
{+0.00/59 0.8} Qd2+ 107. Kg1 {+0.00/67 0.8} Qf4 108. Rh7 {+0.00/59 0.7}
Qc1+ 109. Kg2 {+0.00/99 0.4} Qe3 110. Rc7 {+0.00/67 0.8} Qd3 111. Rh7
{+0.00/56 0.8} Qe2+ 112. Kh3 {+0.00/90 0.7} Qd2 113. Kg3 {+0.00/59 0.7}
Qc3+ 114. Kg2 {+0.00/74 5} Qc4 115. Kg3 {+0.00/62 0.6} Qf1 116. Kh2
{+0.00/58 1.1} Qf4+ 117. Kg2 {+0.00/74 0.6} Qc1 118. Kg3 {+0.00/55 0.9}
Qa3+ 119. Kf2 {+0.00/67 0.6} Qa1 120. Kg2 {+0.00/63 1.1} Qb2+ 121. Kh1
{+0.00/79 1.2} Qf2 122. Rc7 {+0.00/73 0.9} Qf3+ 123. Kg1 {+0.00/75 1.3}
Qg3+ 124. Kf1 {+0.00/79 0.7} Qf3+ 125. Kg1 {+0.00/99 0.2}
{Game drawn by the 50 move rule} 1/2-1/2

[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "Schooner"]
[Black "ArasanX"]
[Result "1-0"]
[WhiteElo "2917"]
[BlackElo "2895"]
[TimeControl "300+1"]

1. d4 Nf6 2. c4 e6 3. g3 {-0.24/19} Bb4+ 4. Nd2 {+0.07/18} d5 5. Bg2
{+0.11/18} O-O 6. Nf3 {+0.02/19} b6 7. O-O {+0.02/21} Bb7 8. b3 {-0.14/17}
c5 9. Bb2 {-0.06/16} Qe7 10. dxc5 {-0.21/17} Bxc5 11. cxd5 {-0.07/24} Bxd5
12. e3 {-0.05/27} a5 13. Qe2 {-0.12/26} a4 14. Rfd1 {-0.01/25} Rd8 15. Ne5
{+0.09/23} Bxg2 16. Kxg2 {-0.04/28} a3 17. Bc3 {+0.00/29} Nd5 18. Ne4
{+0.00/30} f6 19. Nf3 {+0.00/30 9} Na6 20. Bd2 {-0.07/26 18} Qb7 21. Nxc5
{-0.11/29 8} Nxc5 22. e4 {+0.00/32 7} Nc7 23. Rac1 {-0.16/32 17} N7a6 24.
Be3 {-0.26/32 16} Rxd1 25. Rxd1 {-0.08/30 8} Nxe4 26. Bxb6 {-0.08/32 7} f5
27. Rd8+ {+0.00/33 7} Rxd8 28. Bxd8 {+0.00/33 6} Nac5 29. Qc4 {+0.00/33 5}
Nd2 30. Qe2 {+0.00/32 8} Nxf3 31. Qxf3 {+0.00/33 5} Qxf3+ 32. Kxf3
{+0.24/32 6} Nxb3 33. Be7 {+0.19/31 6} Nd4+ 34. Ke3 {+0.56/28 5} Kf7 35.
Bc5 {+0.65/31 10} Nc6 36. Bxa3 {+0.77/32 21} e5 37. Kd3 {+0.67/28 4} Ke6
38. Kc4 {+0.66/32 4} g5 39. Bc1 {+0.66/30 9} f4 40. Kc5 {+0.58/30 6} Nb8
41. a4 {+0.60/31 5} Kf5 42. Ba3 {+0.49/29 5} e4 43. gxf4 {+1.74/32 7} gxf4
44. Kd4 {+1.94/33 6} h5 45. Bd6 {+3.20/37 18} Na6 46. h4 {+3.99/39 11} f3
47. a5 {+4.73/40 2.3} Ke6 48. Bf4 {+5.74/39 9} Kf5 49. Be3 {+6.25/37 2.5}
Ke6 50. Kxe4 {+10.91/42 14} Nc7
{ArasanX resigns} 1-0

[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "atomGoldbar"]
[Black "Schooner"]
[Result "0-1"]
[WhiteElo "2562"]
[BlackElo "2921"]
[TimeControl "300+1"]

1. e4 e5 2. Nf3 Nc6 {+0.38/16} 3. Bc4 Be7 {+0.19/17} 4. d4 exd4 {+0.33/19}
5. Nxd4 Nf6 {+0.35/18} 6. Nxc6 dxc6 {+0.32/19} 7. Qxd8+ Bxd8 {+0.58/19} 8.
f3 Nd7 {+0.53/18} 9. Be2 O-O {+0.52/18} 10. Be3 Bh4+ {+0.52/18} 11. g3 Bf6
{-0.48/28 10} 12. c3 a5 {-0.58/27 11} 13. Nd2 a4 {-0.48/27 8} 14. a3 Be7
{-0.53/31 7} 15. Bf4 Bd8 {-0.55/29 15} 16. O-O Nc5 {-0.23/28 10} 17. h4 h5
{+0.00/27 15} 18. Kg2 Re8 {+0.00/31 7} 19. Rae1 g6 {+0.00/31 30} 20. Rd1
Be6 {+0.00/30 6} 21. Bc4 Rc8 {+0.00/30 5} 22. Be3 Be7 {+0.00/32 7} 23. Bxe6
Nxe6 {+0.00/29 4} 24. e5 Red8 {+0.38/30 5} 25. Ne4 Rd5 {+0.41/29 4} 26. f4
Rcd8 {+0.39/32 6} 27. Rxd5 cxd5 {+0.41/33 4} 28. Ng5 Bxg5 {+0.61/28 6} 29.
hxg5 Ng7 {+0.79/28 12} 30. Bf2 Kf8 {+0.94/28 3} 31. Bc5+ Ke8 {+1.02/33 5}
32. Rh1 Nf5 {+1.09/36 4} 33. Kf3 b6 {+1.11/34 4} 34. Bg1 Kd7 {+1.17/33 5}
35. g4 hxg4+ {+1.44/31 3} 36. Kxg4 c5 {+1.27/34 8} 37. Kf3 Ke6
{+1.36/36 18} 38. Rh2 d4 {+1.31/34 3} 39. cxd4 Nxd4+ {+1.42/29 3} 40. Bxd4
Rxd4 {+1.40/32 3} 41. Kg4 b5 {+1.17/30 2.5} 42. Rh8 Rd2 {+1.20/31 2.6} 43.
Re8+ Kd7 {+1.38/30 6} 44. Rf8 Ke7 {+1.54/29 2.4} 45. Rc8 c4 {+1.52/29 2.7}
46. Rc7+ Ke6 {+1.51/31 8} 47. Kf3 Rxb2 {+1.33/28 2.7} 48. Ke4 Rb1
{+1.83/30 10} 49. Kd4 Rf1 {+1.79/31 6} 50. Kc5 Rxf4 {+3.30/34 6} 51. Kxb5
c3 {+3.88/37 2.0} 52. Rxc3 Kf5 {+3.82/38 4} 53. Kc6 Kxg5 {+4.37/33 2.2} 54.
Kd6 Kf5 {+4.67/32 3} 55. Rc7 Kg4 {+4.92/31 1.7} 56. Rc8 g5 {+5.15/31 2.2}
57. Ke7 Kg3 {+5.26/32 1.7} 58. Rc6 Rf5 {+5.32/30 3} 59. Kd6 g4
{+6.07/30 1.7} 60. Kd5 Rf4 {+7.40/28 1.9} 61. e6 fxe6+ {+8.41/26 1.7} 62.
Ke5 Rf3 {+9.95/28 2.2} 63. Kd4 Rxa3 {+16.11/32 6} 64. Rc8 Kh2
{+21.95/28 2.4} 65. Rc7 g3 {+21.75/22 3} 66. Rh7+ Kg1 {+327.16/23 1.2} 67.
Kc4 g2 {+327.18/22 1.4} 68. Kb4 Rf3 {+327.34/24 3} 69. Rd7 a3
{+327.40/21 1.1} 70. Kc5 a2 {+327.46/22 5} 71. Ra7 Kf1 {+327.48/24 1.0} 72.
Ra4 g1=Q+ {+327.52/23 1.4} 73. Kc6 Qg5 {+327.56/27 1.2} 74. Kd6 Rf7
{+327.58/40 1.2} 75. Rf4+ Qxf4+ {+327.60/99 0.7} 76. Kxe6 a1=Q 77. Kd5 Qb1
78. Kc5 Qb5+ 79. Kxb5 Qb4+ 80. Kxb4 Rc7 81. Kb5 Ke2 82. Ka6 Kd2 83. Kb5 Kc3
84. Kb6 Rc4 85. Kb5 Kb3 86. Kb6 Ka4 87. Ka7 Ka5 88. Kb7 Kb5 89. Ka7 Kc6 90.
Kb8 Kb6 91. Ka8 Rc8#
{atomGoldbar checkmated} 0-1

[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "Schooner"]
[Black "Goldbar"]
[Result "1-0"]
[WhiteElo "2922"]
[BlackElo "2718"]
[TimeControl "300+1"]

1. d4 Nf6 2. c4 g6 3. Nc3 {-0.25/17} Bg7 4. e4 {-0.65/18} d6 5. Nf3
{-0.57/19} O-O 6. Be2 {-0.57/19} e5 7. O-O {-0.55/18} Nc6 8. d5 {-0.51/17}
Ne7 9. Ne1 {-0.77/19} a6 10. Nd3 {+0.71/23 17} Nd7 11. f4 {+0.50/25 14}
exf4 12. Bxf4 {+0.46/25 8} h6 13. Be3 {+0.54/26 8} f5 14. Nf4 {+0.55/25 10}
Ne5 15. c5 {+0.48/23 6} g5 16. Ne6 {+1.09/25 6} Bxe6 17. dxe6 {+1.11/28 13}
f4 18. Bf2 {+1.24/24 12} Rf6 19. Qb3 {+1.26/27 5} f3 20. gxf3 {+1.34/28 7}
Nxf3+ 21. Bxf3 {+1.51/30 5} Rxf3 22. Bg3 {+1.39/27 7} Qf8 23. Rad1
{+1.74/28 13} dxc5 24. Rxf3 {+2.16/33 21} Qxf3 25. Nd5 {+2.29/33 6} Qxe4
26. Nxe7+ {+2.35/34 4} Kh8 27. Qd5 {+2.71/32 6} Bd4+ 28. Rxd4 {+2.76/31 4}
Qxd4+ 29. Kg2 {+3.20/32 14} Qxd5+ 30. Nxd5 {+3.18/30 4} Kg7 31. Nxc7
{+3.73/31 24} Rc8 32. Bd6 {+3.86/27 5} Rxc7 33. Bxc7 {+4.55/26 6} Kf6 34.
Bd6 {+4.61/25 4} Kxe6 35. Bxc5 {+4.85/24 4} Kd5 36. Bf8 {+5.70/22 2.8} h5
37. Be7 {+5.98/24 2.2} g4 38. Bd8 {+7.60/24 5} Ke6 39. Kg3 {+8.55/26 2.9}
Kf5 40. Kh4 {+12.02/24 2.5} Kf4 41. Kxh5 {+14.98/25 7} Kf5 42. Bc7
{+15.90/23 2.0} b5 43. b4 {+327.10/28 4} Kf6 44. Kxg4 {+327.34/27 2.3} Ke6
45. h4 {+327.40/28 2.3} Kd5 46. h5 {+327.42/28 2.2} a5 47. Bxa5
{+327.46/27 4} Kd4 48. h6 {+327.52/33 2.0} Kc3 49. h7 {+327.54/37 1.9} Kb2
50. h8=Q+ {+327.56/47 1.8} Kxa2 51. Qb2+ Kxb2 52. Kf4 Kb3 53. Ke5 Kc4 54.
Kd6 Kd4 55. Kc6 Kc4 56. Kb6 Kd4 57. Kxb5 Kd5 58. Ka4 Kc4 59. b5 Kc5 60. b6
Kc6 61. Kb4 Kb7 62. Kb5 Kc8 63. Kc6 Kd8 64. b7+ Ke7 65. b8=Q Kf6 66. Kd5
Kf5 67. Qg8 Kf4 68. Bd2+ Kf3 69. Qg1 Ke2 70. Ke4 Kxd2 71. Qc5 Ke1 72. Ke3
Kf1 73. Kf3 Ke1 74. Qc1#
{Goldbar checkmated} 1-0

[Event "ICS rated blitz match"]
[Site "nightmare-chess.nl"]
[Date "2020.11.14"]
[Round "-"]
[White "NightmareX"]
[Black "Schooner"]
[Result "1/2-1/2"]
[WhiteElo "2823"]
[BlackElo "2924"]
[TimeControl "300+1"]

1. e4 e5 2. Nf3 Nc6 {+0.20/18} 3. Bc4 Be7 {+0.20/19} 4. d4 exd4 {+0.34/20}
5. Nxd4 Nf6 {+0.29/19} 6. Nxc6 dxc6 {+0.32/19} 7. Qxd8+ Bxd8 {+0.53/18} 8.
f3 Nd7 {+0.54/19} 9. Be2 O-O {+0.43/19} 10. Be3 Bh4+ {+0.49/17} 11. Bf2
{+0.34/19 14} Bxf2+ {-0.51/27 10} 12. Kxf2 {+0.28/24 8} a5 {-0.50/27 7} 13.
a3 {+0.30/21 8} h5 {-0.45/28 11} 14. Nd2 {+0.28/20 10} a4 {-0.43/28 7} 15.
Rad1 {+0.40/19 6} Nc5 {-0.39/23 8} 16. Ke3 {+0.25/21 6} g6 {-0.15/25 6} 17.
Rhe1 {+0.17/20 13} Kg7 {-0.08/28 9} 18. h3 {+0.15/21 12} Re8 {+0.00/27 8}
19. h4 {+0.15/21 5} f5 {+0.00/32 7} 20. Kd4 {+0.10/23 7} Ne6+ {+0.00/34 19}
21. Kc4 {+0.02/23 5} fxe4 {+0.00/27 10} 22. Nxe4 {+0.00/26 7} Nf4
{+0.00/31 11} 23. Bf1 {+0.00/27 13} Bf5 {+0.11/32 19} 24. g3 {+0.00/24 21}
Nd5 {+0.36/27 10} 25. Bd3 {+0.00/24 6} Re5 {+0.30/31 33} 26. Rd2
{+0.00/18 4} Rd8 {+0.31/30 11} 27. Rde2 {+0.00/27 6} b6 {+0.35/31 4} 28. b3
{-0.28/24 4} b5+ {+0.20/30 5} 29. Kd4 {-0.17/26 14} Re6 {+0.25/31 5} 30.
Rd2 {-0.25/22 4} Kh6 {+0.35/25 3} 31. Rdd1 {-0.36/24 24} Nf6+ {+0.08/26 5}
32. Ke3 {-0.37/25 5} Ra8 {+0.45/25 2.7} 33. b4 {-0.40/25 6} Rae8
{+0.61/26 4} 34. Kd4 {-0.43/27 4} Rd8+ {+0.62/28 3} 35. Kc3 {-0.51/26 4}
Bxe4 {+0.53/29 4} 36. fxe4 {-0.41/27 2.8} Rf8 {+0.49/27 2.8} 37. Rc1
{-0.20/21 2.2} Kg7 {+0.71/28 8} 38. Kd2 {-0.25/23 5} Rd8 {+0.70/28 5} 39.
Kc3 {-0.34/26 2.9} Re5 {+0.72/29 2.5} 40. Rcd1 {-0.44/23 2.5} Nxe4+
{+0.68/31 4} 41. Bxe4 {-0.34/28 7} Rxd1 {+0.68/32 2.3} 42. Rxd1
{-0.40/31 2.0} Rxe4 {+0.69/32 2.8} 43. Rd7+ {-0.30/28 4} Kf6 {+0.69/31 4}
44. Rxc7 {-0.34/29 2.3} Rc4+ {+0.63/32 4} 45. Kd3 {-0.33/28 2.3} Kf5
{+0.60/33 11} 46. Rg7 {-0.30/31 2.2} Rg4 {+0.50/30 1.8} 47. Rf7+
{-0.27/29 4} Ke6 {+0.50/36 3} 48. Rf3 {-0.25/30 2.0} Kd6 {+0.48/30 2.1} 49.
Kd2 {-0.23/26 1.9} c5 {+0.48/32 6} 50. Rd3+ {-0.21/29 1.8} Kc6
{+0.45/34 2.2} 51. bxc5 {-0.20/29 0.2} Kxc5 {+0.41/32 1.7} 52. Kc3
{-0.20/28} Re4 {+0.40/34 1.7} 53. Rf3 {-0.18/29} Rc4+ {+0.40/34 2.5} 54.
Kb2 {-0.18/30} Kd5 {+0.39/32 2.0} 55. Kc1 {-0.18/29} Rg4 {+0.38/32 3} 56.
Kd2 {-0.17/29} Ke6 {+0.36/31 2.8} 57. Rd3 {-0.17/26} g5 {+0.34/29 1.5} 58.
hxg5 {+0.00/29} Rxg5 {+0.20/32 1.6} 59. Rf3 {+0.00/32} Rf5 {+0.13/28 2.6}
60. Rd3 {+0.00/31} Kf6 {+0.00/31 4} 61. Rd4 {+0.00/32} Rf3 {+0.00/33 3} 62.
Rd5 {+0.00/30} Kg6 {+0.08/31 1.6} 63. Rd3 {+0.00/31} Rf5 {+0.08/33 1.9} 64.
Rd8 {+0.00/33} Rg5 {+0.11/30 3} 65. Rb8 {+0.00/31} Kf6 {+0.08/31 4} 66. Rb7
{+0.00/31} Rxg3 {+1.09/28 1.1} 67. Rb6+ {+0.00/31} Kg7 {+0.66/29 2.4} 68.
Rxb5 {+0.00/32} h4 {+0.42/29 1.5} 69. Rb4 {+0.00/34} h3 {+0.00/34 1.1} 70.
Rxa4 {+0.00/32} Kf6 {+0.00/36 3} 71. Ke2 {+0.00/35} Kg5 {+0.00/41 1.2} 72.
Ra5+ {+0.00/37} Kf6 {+0.00/44 2.0} 73. Ra4 {+0.00/60} Ke5 {+0.00/37 1.2}
74. Kf2 {+0.00/44} h2 {+0.00/36 1.2} 75. Rh4 {+0.00/49} Rc3 {+0.00/40 4}
76. Rh5+ {+0.00/54 4} Kf4 {+0.00/46 1.2} 77. Rh4+ {+0.00/56 1.2} Ke5
{+0.00/47 1.0} 78. Rh5+ {+0.00/60 1.2} Kf4 {+0.00/59 1.0} 79. Rh4+
{+0.00/60 3} Ke5 {+0.00/53 2.0}
{Game drawn by repetition} 1/2-1/2
[/pgn]
Joost Buijs
Posts: 1563
Joined: Thu Jul 16, 2009 10:47 am
Location: Almere, The Netherlands

Re: On-line engine blitz tourney November

Post by Joost Buijs »

In the past Nightmare won the tournament a few times too, in a tournament with such a small number of rounds you also need a bit of luck to get a good ranking.

This time Nightmare had two draws against atomGoldbar and atomNightmare, these engines use the same book and the same evaluation function, this makes a draw very likely even with a 100 fold speed difference. Goldbar is a Nightmare derivative written in Pascal maintained by Bart Goldhoorn and partly by me.

The slowdown with hash >2GB is something that I see over here too, basically due to pressure on the TLB, allocating the hash-table in large page memory helps a lot to reduce the slowdown, this is what I default do (if the OS allows it).

Do you use separate hash-tables for each thread? Probably not, I use separate pawn hash-tables, the main hash-table is shared between all threads otherwise it wouldn't be very useful.

I've not been looking at NNUE at all, it does not seem very difficult to write code for inference, however producing data to train the NN, writing software to train it and getting the NN properly trained is very time consuming. I guess most people grab their data from internet and use py-torch or tensorflow to train the network.

I'm thinking about some small networks for specific things like pawn structure and king-safety, it's my feeling that this would already help quite a lot.
Madeleine Birchfield
Posts: 512
Joined: Tue Sep 29, 2020 4:29 pm
Location: Dublin, Ireland
Full name: Madeleine Birchfield

Re: On-line engine blitz tourney November

Post by Madeleine Birchfield »

Joost Buijs wrote: Sun Nov 15, 2020 8:26 am I've not been looking at NNUE at all, it does not seem very difficult to write code for inference, however producing data to train the NN, writing software to train it and getting the NN properly trained is very time consuming. I guess most people grab their data from internet and use py-torch or tensorflow to train the network.

I'm thinking about some small networks for specific things like pawn structure and king-safety, it's my feeling that this would already help quite a lot.
Coming up with a good evaluation function has always been the hardest and most time consuming part of an engine, whether that be training a neural network, or coming up with new terms for a handcrafted evaluation and tuning it.
chrisw
Posts: 4313
Joined: Tue Apr 03, 2012 4:28 pm

Re: On-line engine blitz tourney November

Post by chrisw »

Joost Buijs wrote: Sat Nov 14, 2020 10:33 pm It was fun as usual! At least I think it was fun.

Schooner clearly ended first place, my guess is that he added NNUE like many others are lately doing.
Belfish ended second, no surprise here. Marvin seems to have made a lot of progress and ended third.

Code: Select all

:Tourney Players: Round 9 of 9 
:
:     Name              Rating Score Perfrm Upset  Results 
:     ----------------- ------ ----- ------ ------ ------- 
:  1 +Schooner          [2912]  7.5  [2995] [  46] +10b +08w =05w +03b =02w +04w +12b +07w =06b 
:  2 +BelFish           [3005]  7.0  [2934] [   0] +15w +07w =04b +05w =01b =06w +09b =03w +12b 
:  3 +Marvin            [2747]  6.5  [2809] [ 202] +23w =12w +09b -01w +11b +13w =04b =02b +14w 
:  4 +ArasanX           [2893]  6.0  [2847] [  56] +11b +13w =02w =06b +14w -01b =03w +12w =05b 
:  5 +WaspX             [2832]  6.0  [2859] [  70] +18w +14w =01b -02b +13w =07b =06b +08w =04w 
:  6 +NightmareX        [2822]  6.0  [2825] [ 176] +16w =09w =12b =04w +08b =02b =05w +14b =01w 
:  7 +Goldbar           [2717]  5.5  [2670] [  57] +17w -02b +11w =12w +18b =05w +13b -01b =09w 
:  8 +TheBaron          [2649]  5.0  [2560] [   0] +19b -01b +18w -10b -06w +11w +15b -05b +13w
:  9 +atomNightmare     [2559]  5.0  [2542] [ 210] +22w =06b -03w -13b +17b +10w -02w +19b =07b 
: 10 +Nemeton           [2343]  5.0  [2515] [ 342] -01w +19w +17b +08w -12b -09b +18w -13w +15b 
: 11 +rpiArminius       [2306]  5.0  [2414] [   0] -04w +20b -07b +23w -03w -08b +24b +22w +19w 
: 12 +atomGoldbar       [2549]  4.5  [2664] [ 383] +24w =03b =06w =07b +10w +14b -01w -04b -02w 
: 13 +WaDuuttie         [2645]  4.0  [2520] [   0] +20w -04b +16w +09w -05b -03b -07w +10b -08b 
: 14 +Arminius          [2613]  4.0  [2467] [   0] +21w -05b +23w +15b -04b -12w +20b -06w -03b 
: 15 +Jumbo             [2379]  4.0  [2328] [   0] -02b =17w +21b -14w =23b +16w -08w +20b -10w 
: 16 +CTdccc            [2217]  4.0  [2212] [  61] -06b +22w -13b -17w +24b -15b -19w +18b +21w
: 17 +Joker             [2144]  4.0  [2207] [ 190] -07b =15b -10w +16b -09w -19b +21w =24w +22b 
: 18 +Barbarossa        [2278]  3.5  [2252] [   0] -05b +21w -08b +20w -07w +23b -10b -16w =24b 
: 19 +Isa               [2126]  3.5  [2170] [ 109] -08w -10b =22b +21w -20b +17w +16b -09w -11b 
: 20 +feligres          [2123]  3.5  [2179] [  44] -13b -11w +24b -18b +19w +22b -14w -15w =23b 
: 21 +KingSlayer        [2055]  3.0  [2069] [ 150] -14b -18b -15w -19b +22w +24w -17b +23w -16b 
: 22 +microMax          [1937]  2.5  [2007] [ 362] -09b -16b =19w +24w -21b -20w +23w -11b -17w 
: 23 +Spartacus         [2205]  2.0  [2040] [  87] -03b +24w -14b -11b =15w -18w -22b -21b =20w 
: 24 +atomSargonI       [1930]  1.0  [1890] [ 281] -12b -23b -20w -22b -16w -21b -11w =17b =18w
I managed to completely miss this, second time it happened, only noticed the thread late Saturday. Shame, would have liked to have been there. Will pay more attention next time.
Joost Buijs
Posts: 1563
Joined: Thu Jul 16, 2009 10:47 am
Location: Almere, The Netherlands

Re: On-line engine blitz tourney November

Post by Joost Buijs »

chrisw wrote: Sun Nov 15, 2020 2:11 pm I managed to completely miss this, second time it happened, only noticed the thread late Saturday. Shame, would have liked to have been there. Will pay more attention next time.
Next time I will notify you personally. In the past there used to be ~40 participants, the last half year it's 24 max.

By the end of the month there is Charles Roberson's ACCA rapid tournament, yet 7 participants, nowadays most people are only interested in shows like TCEC and CCCC.
User avatar
hgm
Posts: 27788
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: On-line engine blitz tourney November

Post by hgm »

All the Raspberry Pi participants have disappeared. There were a lost of those.
Joost Buijs
Posts: 1563
Joined: Thu Jul 16, 2009 10:47 am
Location: Almere, The Netherlands

Re: On-line engine blitz tourney November

Post by Joost Buijs »

hgm wrote: Sun Nov 15, 2020 2:45 pm All the Raspberry Pi participants have disappeared. There were a lost of those.
Indeed, of course the PI engines of Marcel are missing since his passing, the PI engines of Folkert are missing too. At the moment Volker Annuss is the only one with his PI engine permanently online.

This is why I always keep a few engines running on a single core Intel atom which gives about the same performance as the RPI3 (between 0.5 and 1.0 mnps). It is also very handy to have a few engines permanently online for testing purposes.

The server runs under a Hyper-V virtual machine on a 6 core Intel machine, unfortunately the server software is not multi threaded, with a lot of participants (especially when it has to start a new series of games) it gets rather slow.
D Sceviour
Posts: 570
Joined: Mon Jul 20, 2015 5:06 pm

Re: On-line engine blitz tourney November

Post by D Sceviour »

Joost Buijs wrote: Sun Nov 15, 2020 8:26 am The slowdown with hash >2GB is something that I see over here too, basically due to pressure on the TLB, allocating the hash-table in large page memory helps a lot to reduce the slowdown, this is what I default do (if the OS allows it).

Do you use separate hash-tables for each thread? Probably not, I use separate pawn hash-tables, the main hash-table is shared between all threads otherwise it wouldn't be very useful.
Schooner uses a shared hash table. I meant that the total amount for hash size is calculated as a linear increase with the number of threads = 256 Mb x number of threads.

I am using Windows 10 Home edition which apparently disallows large page settings. This has to be set in the Group Policy Editor "gpedit.msc", but the use of gpedit is supposed to be only allowed in the Professional edition of Windows 10. Is the Windows upgrade really worth it? grrr. :evil:

Another alternative is to use Linux for large pages. This is supposed to be easier. However the version of Ubuntu I use is packed with restrictions on viewing folders. It seems like an endless battle with these new operating systems.