Strange. Looks like Skipper plays better with null move disabled. For instance this game has been played without doing any pruning or reductions.
[pgn]
[Event "Computer Chess Game"]
[Site "HP"]
[Date "2017.03.27"]
[Round "-"]
[White "Fairy-Max 4.8S"]
[Black "SkipperWinb"]
[Result "0-1"]
[TimeControl "120"]
[Annotator "1. +0.11 1... +0.11"]
1. c4 {+0.11/8} Nc6 {+0.11/70 2.0} 2. f4 {+0.03/8 3} b6 {+0.21/70 2.0} 3.
Nf3 {+0.14/8 1.6} g6 {+0.14/70 2.0} 4. g3 {+0.00/8 1.7} d5 {+0.11/60 1.9}
5. Qa4 {+0.20/9 2.4} Qd7 {-0.05/70 1.9} 6. Nc3 {+0.23/9 4} dxc4
{-0.15/70 1.9} 7. Bh3 {-0.09/8 1.6} e6 {-0.14/70 1.8} 8. Qxc4 {-0.10/8 2.2}
Bb7 {-0.10/60 1.8} 9. O-O {-0.12/7 2.0} Na5 {+0.01/60 1.8} 10. Qb5
{-0.28/9 1.5} Nf6 {-0.07/60 1.8} 11. Qxd7+ {+0.15/10 3} Kxd7 {+0.01/70 1.7}
12. Ne5+ {+0.24/9 1.9} Ke8 {-0.07/70 1.7} 13. d4 {-0.04/9 2.3} Nd5
{-0.01/60 1.7} 14. Bg4 {-0.08/7 1.2} Rc8 {+0.00/60 1.7} 15. Nxd5
{-0.18/8 3} Bxd5 {-0.02/70 1.6} 16. Bf3 {-0.13/8 1.1} c5 {+0.01/70 1.6} 17.
e4 {+0.02/9 1.4} Ba8 {-0.01/70 1.6} 18. d5 {-0.21/9 1.2} Bg7 {-0.18/70 1.6}
19. Bg4 {-0.24/8 2.0} Ke7 {-0.35/70 1.5} 20. d6+ {+0.47/7 1.1} Kf8
{-0.08/80 1.6} 21. Bf3 {+0.16/8 1.3} f6 {-0.03/60 1.5} 22. Ng4
{+0.34/9 1.3} f5 {-0.12/70 1.5} 23. exf5 {+0.28/10 1.7} Bd4+ {-0.12/80 1.5}
24. Be3 {+0.54/10 1.5} Bxb2 {+0.49/70 1.4} 25. d7 {+0.27/9 1.2} Rd8
{+0.50/80 1.4} 26. Bxa8 {-0.04/10 1.3} Bxa1 {+1.57/80 1.4} 27. fxe6
{+0.07/10 1.0} Rxa8 {+0.93/80 1.4} 28. Rxa1 {+0.09/10 1.1} Ke7
{+0.94/80 1.4} 29. Ne5 {-0.05/10 1.7} Kxe6 {+1.57/70 1.3} 30. Re1
{-0.23/9 1.2} Rhf8 {+1.73/60 1.3} 31. h4 {-0.17/8 1.1} Rad8 {+1.63/60 1.3}
32. Bc1 {-0.02/9 1.2} Kd6 {+1.70/70 1.3} 33. Rd1+ {-0.21/9 2.4} Kc7
{+1.74/80 1.3} 34. f5 {-0.92/10 1.2} Nc6 {+1.81/80 1.3} 35. Bf4
{-0.69/10 1.0} Kb7 {+1.79/80 1.3} 36. Nxc6 {+0.08/10 1.7} Kxc6
{+1.69/80 1.2} 37. Rd6+ {-1.15/11 1.0} Kb7 {+1.69/90 1.2} 38. Bg5
{-1.47/11 1.6} Kc7 {+2.09/80 1.2} 39. Bf4 {-1.43/12 1.4} Rxf5
{+2.16/80 1.2} 40. Rd1+ {-1.71/13 0.9} Rxf4 {+2.25/90 1.2} 41. gxf4
{-1.44/13 0.9} Rxd7 {+2.78/80 1.1} 42. Re1 {-1.57/12 16} c4 {+3.01/80 1.1}
43. Re3 {-1.52/9 0.5} b5 {+2.82/70 1.1} 44. a3 {-1.48/9 0.9} a5
{+2.85/70 1.1} 45. Kf2 {-1.46/9 0.9} Rf7 {+2.83/70 1.1} 46. Kg3
{-1.54/9 0.5} Kd7 {+2.72/70 1.1} 47. Rc3 {-1.52/10 0.9} Kd6 {+2.75/80 1.0}
48. a4 {-1.78/10 1.0} Kc5 {+2.76/80 1.0} 49. axb5 {-1.95/10 0.6} Kxb5
{+2.76/80 1.0} 50. Rc2 {-2.13/10 1.5} Rc7 {+2.87/80 1.0} 51. h5
{-2.11/10 1.4} gxh5 {+3.06/80 1.0} 52. f5 {-2.27/10 0.8} Rc5 {+3.02/80 1.0}
53. Kf4 {-2.17/10 0.8} a4 {+2.75/80 1.0} 54. f6 {-1.68/10 0.5} Kb4
{+2.45/80 0.9} 55. f7 {-1.06/10 0.5} Rc8 {+2.01/90 0.9} 56. Rf2
{-1.15/10 0.5} Rf8 {+2.25/90 0.9} 57. Ke5 {-0.56/10 0.4} a3 {+1.57/90 1.0}
58. Ke6 {-0.44/11 0.5} Kb3 {+1.29/90 0.9} 59. Ke7 {-0.49/11 0.5} a2
{+0.00/90 0.9} 60. Rf1 {-1.94/12 0.4} Rxf7+ {+2.66/90 0.8} 61. Kxf7
{-6.78/12 0.5} c3 {+2.78/90 0.8} 62. Ke6 {-6.83/12 0.5} c2 {+8.38/80 0.8}
63. Rg1 {-7.61/11 0.3} Kb2 {+9.45/80 0.8} 64. Kf6 {-9.35/12 0.3} c1=Q
{+12.08/80 0.8} 65. Rxc1 {-17.01/14 4} Kxc1 {+12.13/100 0.8} 66. Kg7
{-17.22/14 0.2} a1=Q+ {+22.21/90 0.8} 67. Kf7 {-18.01/13 0.8} h4
{+21.53/80 0.8} 68. Ke7 {-18.63/12 0.4} h3 {+23.25/70 0.8} 69. Ke6
{-79.95/12 0.8} Qa8 {+23.37/70 0.8} 70. Ke5 {-18.64/11 0.3} Kd1
{+23.39/70 0.8} 71. Kd4 {-18.66/10 0.2} h2 {+23.50/70 0.7} 72. Ke5
{-79.96/10 0.2} Qf8 {+23.44/70 0.7} 73. Kd5 {-79.96/10 0.2} Qh6
{+23.35/70 0.7} 74. Kd4 {-79.96/11 0.2} h1=Q {+23.51/70 0.7} 75. Kc5
{-79.97/13 0.2} Q1h3 {+26214.39/60 0.7} 76. Kd4 {-79.98/25 0.2} Qd6+
{+26214.39/70 0.8} 77. Kc4 {-79.99/28 0.2} Qhd3# {+26214.39/80 0.7}
{Xboard adjudication: Checkmate} 0-1
[/pgn]
Clean
Moderators: hgm, Rebel, chrisw
-
- Posts: 1357
- Joined: Wed Mar 08, 2006 10:15 pm
- Location: San Francisco, California
Re: Clean
There certainly have been times when I thought Myrddin would play better with 90% of the evaluation disabled.Sven Schüle wrote:Next step: disable move generator. Number of bad moves played by Skipper will decrease.
-
- Posts: 759
- Joined: Fri Jan 04, 2013 4:55 pm
- Location: Nice
Re: Clean
Sven wrote this joke because of your useless topicHenk wrote:Illustrates the quality of the recommendations of these experts.Daniel Anulliero wrote:Sven Schüle wrote:Next step: disable move generator. Number of bad moves played by Skipper will decrease.
Everybody know an engine play more deeper and yes much better with a null move correctly implemented.
Isa is far from a top engine , I never said I'm an expert but I
understand some things , and sure , you have the same bugs in Skipper since 3-4 years , so , sometimes you have some Joking answers , think why
Isa download :
-
- Posts: 7220
- Joined: Mon May 27, 2013 10:31 am
Re: Clean
If you haven't seen the source code you don't understand what's going on. Strangely null move does not give any improvement anymore and I did not change the code. So I don't understand why. Might be it never gave any improvement.Daniel Anulliero wrote:Sven wrote this joke because of your useless topicHenk wrote:Illustrates the quality of the recommendations of these experts.Daniel Anulliero wrote:Sven Schüle wrote:Next step: disable move generator. Number of bad moves played by Skipper will decrease.
Everybody know an engine play more deeper and yes much better with a null move correctly implemented.
Isa is far from a top engine , I never said I'm an expert but I
understand some things , and sure , you have the same bugs in Skipper since 3-4 years , so , sometimes you have some Joking answers , think why
These tests are so tremendously boring and wasting so much computer time. Also what holds on one minute games may not hold on five minutes games.
-
- Posts: 2559
- Joined: Fri Nov 26, 2010 2:00 pm
- Location: Czech Republic
- Full name: Martin Sedlak
Re: Clean
Well, it's simple, nullmove gives ~70+ elo, YMMV. If you get regression then it means that you don't test your code at all or or's pure crap (both is possible)Henk wrote:If you haven't seen the source code you don't understand what's going on. Strangely null move does not give any improvement anymore and I did not change the code. So I don't understand why. Might be it never gave any improvement.
These tests are so tremendously boring and wasting so much computer time. Also what holds on one minute games may not hold on five minutes games.
- note that disabling nullmove will lose elo BUT will make your engine solve some testpostions where others need randomly lobotomized piece of shite like "matesolvers" (troll trademark) and so on.
So yeah, cry ad lib. Basing conclusions on one game, c'mon. You can do better Anyway - you got what you asked for.
-
- Posts: 7220
- Joined: Mon May 27, 2013 10:31 am
Re: Clean
If time control is less than about three minutes null move reductions gives no improvements in my code.mar wrote:Well, it's simple, nullmove gives ~70+ elo, YMMV. If you get regression then it means that you don't test your code at all or or's pure crap (both is possible)Henk wrote:If you haven't seen the source code you don't understand what's going on. Strangely null move does not give any improvement anymore and I did not change the code. So I don't understand why. Might be it never gave any improvement.
These tests are so tremendously boring and wasting so much computer time. Also what holds on one minute games may not hold on five minutes games.
- note that disabling nullmove will lose elo BUT will make your engine solve some testpostions where others need randomly lobotomized piece of shite like "matesolvers" (troll trademark) and so on.
So yeah, cry ad lib. Basing conclusions on one game, c'mon. You can do better Anyway - you got what you asked for.