Arasan 20.2

[jdart]

Arasan 20.2 is available from https://arasanchess.org/.

Changes in Arasan 20.2:

1) Add probcut to search.
2) Detect King and Rook pawn vs wrong-color Bishop configuration
and score near draw, even if opponent has pawns.
3) Change history score calculation (similar to Stockfish).
4) Implement countermove history similar to Stockfish and use in move
ordering. Consolidate history/refutation/countermove history stuff
into SearchContext class and remove separate History and Refutations
classes.
5) Sync with changes to jdart1/Fathom. In particular, pick up a fix for
crashes when running multithreaded.
6) Changes/fixes to strength limitation feature.
7) Fix possible out of range errors in razoring code.
8) Fixes to king cover parameter tuning code. Re-tune parameters.
9) Changes to LMR. Adjust LMR amount based on history values.
10) Remove Util class and replace with std library functions.
11) Allow null move at depth = 1 ply.
12) Changes to futility, reverse futility, and razoring margins.
13) Fix possible bug in output of time control increment to PGN.
14) Separate scoring for outposts in middlegame/endgame.

Plus the usual opening book updates. In my testing this version is about +30 ELO compared to Arasan 20.1.

--Jon

[Lyudmil Tsvetkov]

I used to play a lot with early versions of Arasan years ago, when the engine had very limited king safety, and winning on a slow computer was easy, but now winning against most of the engines of Arasan level is a bit of a challenge.

excuse me for the stupid question, quite probably I have missed some information available somewhere, but where does the name of the engine derive from? (my first thought would be an Indian tribal head...)

[Guenther]

Arasan 20.2 is available from https://arasanchess.org/.

Changes in Arasan 20.2:

1) Add probcut to search.
2) Detect King and Rook pawn vs wrong-color Bishop configuration
and score near draw, even if opponent has pawns.
3) Change history score calculation (similar to Stockfish).
4) Implement countermove history similar to Stockfish and use in move
ordering. Consolidate history/refutation/countermove history stuff
into SearchContext class and remove separate History and Refutations
classes.
5) Sync with changes to jdart1/Fathom. In particular, pick up a fix for
crashes when running multithreaded.
6) Changes/fixes to strength limitation feature.
7) Fix possible out of range errors in razoring code.
8) Fixes to king cover parameter tuning code. Re-tune parameters.
9) Changes to LMR. Adjust LMR amount based on history values.
10) Remove Util class and replace with std library functions.
11) Allow null move at depth = 1 ply.
12) Changes to futility, reverse futility, and razoring margins.
13) Fix possible bug in output of time control increment to PGN.
14) Separate scoring for outposts in middlegame/endgame.

Plus the usual opening book updates. In my testing this version is about +30 ELO compared to Arasan 20.1.

--Jon

Thanks for your long time work Jon!

Guenther

[Damir]

Thanks for the new engine Jon.

[Dokterchen]

Good news, Thanks Jon.

[AdminX]

I used to play a lot with early versions of Arasan years ago, when the engine had very limited king safety, and winning on a slow computer was easy, but now winning against most of the engines of Arasan level is a bit of a challenge.

excuse me for the stupid question, quite probably I have missed some information available somewhere, but where does the name of the engine derive from? (my first thought would be an Indian tribal head...)

What does "Arasan" mean?

Arasan means "king" in Tamil, a language spoken in South India and Sri Lanka. It is related to the Sanskrit word "raja." The Arasan website header has "Arasan" written in Tamil script.

[Dann Corbit]

Any release of Arasan is a happy day.

I have the source, I have the binary.

Many happy hours ahead with these.

[AdminX]

Thanks for the update!

I enjoyed viewing the black side of this game.

[pgn]
[Event "5 Minutes/Game"]
[Site "Engine Match"]
[Date "2017.07.15"]
[Round "4"]
[White "Chiron 4 64bit [3]"]
[Black "Arasan 20.2 [3]"]
[ECO "A06"]
[Result "0-1"]

1. Nf3 {book 0s} Nf6 {book 0s} 2. b3 {book 0s} d5 {book 0s}
3. Bb2 {book 0s} c6 {book 0s} 4. g3 {book 0s} Bf5 {book 0s}
5. Bg2 {book 0s} h6 {book 0s} 6. d3 {book 0s} e6 {book 0s}
7. O-O {book 0s} Nbd7 {book 0s} 8. Nbd2 {book 0s} Be7 {book
0s} 9. e4 {book 0s} Bh7 {book 0s} 10. e5 {book 0s} Ng8
{+0.83/22 8s} 11. Qe2 {book 0s (Ne1)} Bc5 {+0.66/25 8s}
12. d4 {+0.88/26 16s} Bb6 {+0.64/26 0s} 13. c4 {+0.88/27 6s
(Ne1)} Ne7 {+0.64/23 8s} 14. Rac1 {+0.85/25 1s (c5)} O-O
{+0.67/23 8s} 15. Rfd1 {+0.85/24 0s} a5 {+0.63/22 9s (Re8)}
16. a3 {+0.79/21 7s (a4)} Kh8 {+0.60/22 13s (Bc7)} 17. Ne1
{+1.04/21 6s (Nf1)} Qb8 {+0.66/21 7s (Bc7)} 18. Nd3
{+1.24/21 8s (Nf1)} Qa7 {+0.55/22 6s (Qe8)} 19. Nf3
{+1.07/22 8s} Bc7 {+0.52/23 0s (a4)} 20. a4 {+1.08/18 6s}
Rfe8 {+0.44/23 2s (Qb6)} 21. Qe3 {+1.00/19 9s (Ba3)} Rad8
{+0.38/21 6s} 22. Nd2 {+0.88/20 5s (c5)} dxc4 {+0.30/21 6s}
23. Nxc4 {+0.87/21 0s} Nf5 {+0.18/26 7s (Nd5)} 24. Qf4
{+0.88/22 5s} Nf6 {0.00/26 1s (Kg8)} 25. Qf3 {+0.75/21 12s
(Nc5)} Nd5 {+0.18/27 5s} 26. Nc5 {+0.75/20 0s} Nb4
{+0.18/26 7s (Kg8)} 27. Rd2 {+0.66/20 13s (Qh5)} b6
{0.00/23 5s (Kg8)} 28. Ne4 {+0.66/21 5s} Bg6 {+0.18/22 1s
(Re7)} 29. Qe2 {+0.60/20 8s} Bh7 {+0.18/22 0s (Rd7)}
30. Rdd1 {+0.81/19 5s} Na2 {+0.12/22 0s (Re7)} 31. Rc2
{+0.63/22 8s} Nb4 {+0.34/25 0s} 32. Rcd2 {+0.63/19 3s} Rd7
{+0.16/23 3s} 33. f4 {+0.65/20 4s (h4)} Red8 {+0.11/20 5s
(Nd5)} 34. Qf2 {+0.94/22 3s (Kh1)} Rf8 {+0.45/21 6s (Ne7)}
35. g4 {+1.19/18 4s} Ne7 {+0.54/20 1s} 36. Ng3 {+1.15/21 2s
(Rf1)} b5 {0.00/21 4s (Ned5)} 37. axb5 {+0.75/19 4s} cxb5
{+0.23/22 1s} 38. Na3 {+0.35/19 8s} Qb6 {-0.02/22 0s}
39. g5 {+0.23/19 9s (f5)} Ned5 {-0.18/21 4s} 40. Bf1
{+0.10/20 0s (Ne4)} f6 {-0.93/22 4s} 41. gxf6 {0.00/22 4s}
gxf6 {-0.79/23 1s} 42. Nxb5 {0.00/21 1s} Bb8 {-0.84/22 4s
(fxe5)} 43. Kh1 {-0.29/20 16s (Bc1)} Qb7 {-1.24/22 4s}
44. Kg1 {-1.11/20 11s (Nd6)} Rg7 {-1.85/22 4s} 45. Qg2
{-1.08/20 2s (Rc1)} h5 {-1.84/22 13s (Qb6)} 46. Qf3
{-3.35/18 5s (Nd6)} h4 {-3.85/22 3s} 47. Nd6 {-3.45/19 0s
(Rg2)} Qa7 {-4.76/23 10s (Bxd6)} 48. Kh1 {-3.86/17 3s
(Nb5)} hxg3 {-4.49/23 3s} 49. hxg3 {-3.96/20 0s} fxe5
{-4.26/25 3s} 50. dxe5 {-4.32/21 0s} Qe3 {-4.41/26 3s}
51. Qxe3 {-4.48/21 4s} Nxe3 {-4.68/26 0s} 52. Rc1 {-4.59/22
4s (Kh2)} Rxg3 {-4.85/25 2s} 53. Rh2 {-4.70/22 4s} Kg7
{-4.99/25 1s (Kg8)} 54. Be2 {-5.17/19 5s (Bd4)} Nd3
{-5.18/25 2s} 55. Bxd3 {-5.92/20 0s} Bxd3 {-5.87/27 3s}
56. Rh4 {-6.69/24 2s} Be2 {-6.10/27 0s} 57. Rg1 {-6.64/24
0s} Rxg1+ {-6.31/28 2s} 58. Kxg1 {-6.83/24 0s} Ba7
{-6.49/26 3s} 59. Kf2 {-6.89/25 0s} Nf5+ {-6.77/25 2s}
60. Kxe2 {-6.97/25 0s} Nxh4 {-6.96/26 2s} 61. Nb5 {-7.08/24
0s (Bc3)} Bb6 {-7.20/25 2s (Bc5)} 62. Bd4 {-7.05/22 1s
(Nd6)} Ng6 {-8.30/23 2s (Rb8)} 63. Bc3 {-8.06/22 1s (Bxb6)}
Nxf4+ {-9.03/26 2s} 64. Kd2 {-9.34/24 1s (Kd1)} Nd5
{-9.61/26 2s (Rb8)} 65. Bd4 {-11.03/21 2s} Rf5 {-10.67/27
0s} 66. Ba1 {-12.21/25 0s (Kd3)} Rf3 {-11.61/24 1s (Rf2+)}
67. Kc2 {-14.02/25 1s (Na3)} Nb4+ {-16.75/27 1s} 68. Kd2
{-250.00/26 6s} Rf2+ {-17.90/27 0s} 69. Kd1 {-250.00/26 0s}
Rf1+ {-27.51/27 1s} 70. Ke2 {-250.00/28 0s} Rxa1 {-31.45/29
2s} 71. Nd6 {-250.00/26 0s} Rb1 {-31.63/24 2s} 72. Nc8
{-250.00/27 0s (Kf3)} Nd5 {-37.45/25 1s (Rxb3)} 73. b4
{-250.00/24 0s (Nd6)} Rxb4 {-37.45/20 1s} 74. Nd6 {-M11/26
0s (Kd2)} Bd4 {-33.07/19 1s (Rb3)} 75. Ne8+ {-250.00/24 1s
(Kd2)} Kg6 {-38.88/20 1s (Kf8)} 76. Nd6 {-250.00/24 0s}
Bxe5 {-130.62/16 0s} 77. Nc8 {-250.00/27 0s} a4 {-M10/16
1s} 78. Na7 {-250.00/26 0s (Kd3)} Ne7 {-M8/16 1s (a3)}
79. Nb5 {-M8/20 0s (Kd3)} Rxb5 {-M7/13 0s} 80. Kd3 {-M7/20
0s} Rb3+ {-M7/12 0s (Rb4)} 81. Kc4 {-M6/23 0s (Kd2)} Rc3+
{-M6/10 0s (Nd5)} 82. Kb5 {-M5/39 0s} a3 {-M5/19 0s}
83. Kb6 {-M4/87 0s (Ka6)} Bc7+ {-M4/6 0s (a2)} 84. Kb5
{-M3/128 0s (Ka7)} a2 {-M3/4 0s} 85. Kb4 {-M2/128 0s (Ka4)}
a1=Q {-M2/2 0s} 86. Kb5 {-M1/10 0s} Qa5# {-M1/2 0s} 0-1
[/pgn]

[Engin]

first of all many thanks for new Arasan, its become more and more strength to near top engines now.

ok, second what i want to say here, why it is everything allowed take ideas from open sources to implement them in open sources again and why its not allowed take ideas to implement them in closed source project too, everybody is complain and declare that taked ideas from SF or other strong engines as a "clone" or "derived from X

Why is taking ideas WRONG ? i can not found a wheel or circle from new if its allready be found or improved.

it is ok to me study and take only ideas from engine X or sharing ideas between others engines, just only not copy whole source or most of them.

Also it would be fine if Authors are more helping weaker engines to improve them or help to fixing bugs, instead of helping strong engines to improve them more and more, where they want to go, they are allready strong enough.

[Rodolfo Leoni]

first of all many thanks for new Arasan, its become more and more strength to near top engines now.

ok, second what i want to say here, why it is everything allowed take ideas from open sources to implement them in open sources again and why its not allowed take ideas to implement them in closed source project too, everybody is complain and declare that taked ideas from SF or other strong engines as a "clone" or "derived from X

Why is taking ideas WRONG ? i can not found a wheel or circle from new if its allready be found or improved.

it is ok to me study and take only ideas from engine X or sharing ideas between others engines, just only not copy whole source or most of them.

Also it would be fine if Authors are more helping weaker engines to improve them or help to fixing bugs, instead of helping strong engines to improve them more and more, where they want to go, they are allready strong enough.

Hi Engin,

How are you? And how much stronger is Tornado now?

I think it'd be fine if Authors work a bit more on features than on mere strenght. Agreed)

Rodolfo