WCTCC (World Computer Theban Chess Championship)

Laskos · Post by **Laskos** » Sat May 10, 2014 9:32 pm

Shortly:
I proposed possibly a viable game, which uses mostly the rules and pieces of standard chess, as much as chess engines like Stockfish and Komodo can play it. Basically I change the starting position and the starting set of pieces, the rest being the same. Many (most) engines will refuse to play with pawns on the first rank or with more than 8 pawns, but there are _chess_ engines which do play this game, although they are not built and tuned to play it. Also, the starting position is not legal in regular chess.

The starting postion is:

[d]

Spartans are white, in their standard formation, having the King (always on the right wing in Sparta), a Knight (Hippeis), and 14 pawns, a homogenous phalanx. (Note an inaccuracy in "Spartan Chess", Spartans never had both Kings in a field battle.)

Thebans are black, in their formation of Pelopidas and Epaminondas against the Spartans. The King and the Knight (Sacred Band) are on the left flank, the 12 pawns are forming a deep column opposing the Spartan King, and the phalnax in echelon formation.

These two formations are pretty historically accurate, with Thebans in numerical inferiority (12 pawns against 14).

Rules, pieces and movements are all the same as in the standard chess.

Championship results 2014:
11 out of top 100 _chess_ engines, engines tuned to play regular chess, not Theban variant. 11 engines admitting Theban variant, 40x Round Robin from 40 openings positions.
Ratings using Ordo at 1' + 1''

Code: Select all

   # PLAYER                        : RATING    POINTS  PLAYED    (%)
   1 Stockfish 070514 64 SSE4.2    : 2073.6     367.5     400   91.9%
   2 Komodo TCECr 64-bit           : 2000.0     352.5     400   88.1%
   3 DiscoCheck 5.2                : 1796.0     302.0     400   75.5%
   4 Quazar 0.4 x64                : 1712.9     279.0     400   69.8%
   5 Arasan 17.1                   : 1570.6     238.5     400   59.6%
   6 Delfi 5.4                     : 1517.8     223.5     400   55.9%
   7 LittleThought v1.052          : 1195.7     131.5     400   32.9%
   8 Alfil 13.1 x64 MT             : 1078.4      96.5     400   24.1%
   9 Hannibal 1.4bx64              : 1076.7      96.0     400   24.0%
  10 SOS 5 for Arena               :  966.5      64.5     400   16.1%
  11 Protector 1.6.0 x64           :  904.7      48.5     400   12.1%

Previously, in 2013, the title was held by Komodo 5.1, with Stockfish 3 as third.
http://www.talkchess.com/forum/viewtopic.php?t=40990

The game is balanced white-black, with very few draws

Code: Select all

Games        :   2200 (finished)

White Wins   :   1133 (51.5 %)
Black Wins   :    979 (44.5 %)
Draws        :     88 ( 4.0 %)
Unfinished   :      0

White Perf.  : 53.5 %
Black Perf.  : 46.5 %

bob · Post by **bob** » Sat May 10, 2014 10:42 pm

Laskos wrote:Shortly:
I proposed possibly a viable game, which uses mostly the rules and pieces of standard chess, as much as chess engines like Stockfish and Komodo can play it. Basically I change the starting position and the starting set of pieces, the rest being the same. Many (most) engines will refuse to play with pawns on the first rank or with more than 8 pawns, but there are _chess_ engines which do play this game, although they are not built and tuned to play it. Also, the starting position is not legal in regular chess.

The starting postion is:

[d]

Spartans are white, in their standard formation, having the King (always on the right wing in Sparta), a Knight (Hippeis), and 14 pawns, a homogenous phalanx. (Note an inaccuracy in "Spartan Chess", Spartans never had both Kings in a field battle.)

Thebans are black, in their formation of Pelopidas and Epaminondas against the Spartans. The King and the Knight (Sacred Band) are on the left flank, the 12 pawns are forming a deep column opposing the Spartan King, and the phalnax in echelon formation.

These two formations are pretty historically accurate, with Thebans in numerical inferiority (12 pawns against 14).

Rules, pieces and movements are all the same as in the standard chess.

Championship results 2014:
11 out of top 100 _chess_ engines, engines tuned to play regular chess, not Theban variant. 11 engines admitting Theban variant, 40x Round Robin from 40 openings positions.
Ratings using Ordo at 1' + 1''
Code: Select all
   # PLAYER                        : RATING    POINTS  PLAYED    (%)
   1 Stockfish 070514 64 SSE4.2    : 2073.6     367.5     400   91.9%
   2 Komodo TCECr 64-bit           : 2000.0     352.5     400   88.1%
   3 DiscoCheck 5.2                : 1796.0     302.0     400   75.5%
   4 Quazar 0.4 x64                : 1712.9     279.0     400   69.8%
   5 Arasan 17.1                   : 1570.6     238.5     400   59.6%
   6 Delfi 5.4                     : 1517.8     223.5     400   55.9%
   7 LittleThought v1.052          : 1195.7     131.5     400   32.9%
   8 Alfil 13.1 x64 MT             : 1078.4      96.5     400   24.1%
   9 Hannibal 1.4bx64              : 1076.7      96.0     400   24.0%
  10 SOS 5 for Arena               :  966.5      64.5     400   16.1%
  11 Protector 1.6.0 x64           :  904.7      48.5     400   12.1%
Previously, in 2013, the title was held by Komodo 5.1, with Stockfish 3 as third.
http://www.talkchess.com/forum/viewtopic.php?t=40990

The game is balanced white-black, with very few draws
Code: Select all
Games        :   2200 (finished)

White Wins   :   1133 (51.5 %)
Black Wins   :    979 (44.5 %)
Draws        :     88 ( 4.0 %)
Unfinished   :      0

White Perf.  : 53.5 %
Black Perf.  : 46.5 %

If "the rules are the same as standard chess" exactly how does standard chess mandate that a pawn on the 1st rank moves? Just one square ahead? two squares ahead? What about 2nd rank? Still allowed to move one or two ranks?

Laskos · Post by **Laskos** » Sat May 10, 2014 11:20 pm

bob wrote:
If "the rules are the same as standard chess" exactly how does standard chess mandate that a pawn on the 1st rank moves? Just one square ahead? two squares ahead? What about 2nd rank? Still allowed to move one or two ranks?

1st rank moves just one square ahead. 2nd rank (even the former 1st rank) one or two squares ahead. En passant applies as usual (and only applicable for the second rank).
First rank may be interpreted as assembly area, second rank is the charge

Graham Banks · Post by **Graham Banks** » Sat May 10, 2014 11:34 pm

Laskos wrote:
bob wrote:
If "the rules are the same as standard chess" exactly how does standard chess mandate that a pawn on the 1st rank moves? Just one square ahead? two squares ahead? What about 2nd rank? Still allowed to move one or two ranks?
1st rank moves just one square ahead. 2nd rank (even the former 1st rank) one or two squares ahead. En passant applies as usual (and only applicable for the second rank).

These engines all have a Theban Chess mode?
Otherwise how do they understand the rules regarding pawn moves from the first rank (and that they can only move one square ahead, not two)?

Laskos · Post by **Laskos** » Sat May 10, 2014 11:39 pm

Graham Banks wrote:
Laskos wrote:
bob wrote:
If "the rules are the same as standard chess" exactly how does standard chess mandate that a pawn on the 1st rank moves? Just one square ahead? two squares ahead? What about 2nd rank? Still allowed to move one or two ranks?
1st rank moves just one square ahead. 2nd rank (even the former 1st rank) one or two squares ahead. En passant applies as usual (and only applicable for the second rank).
These engines all have a Theban Chess mode?
Otherwise how do they understand the rules regarding pawn moves from the first rank (and that they can only move one square ahead, not two)?

Well, the starting position and pieces are not legal in standard chess. If _chess_ engines allow for that, and for the moves of first-second rank pawns, it's just by sheer chance.

carldaman · Post by **carldaman** » Sun May 11, 2014 12:57 am

Laskos wrote:
Graham Banks wrote:
Laskos wrote:
bob wrote:
If "the rules are the same as standard chess" exactly how does standard chess mandate that a pawn on the 1st rank moves? Just one square ahead? two squares ahead? What about 2nd rank? Still allowed to move one or two ranks?
1st rank moves just one square ahead. 2nd rank (even the former 1st rank) one or two squares ahead. En passant applies as usual (and only applicable for the second rank).
These engines all have a Theban Chess mode?
Otherwise how do they understand the rules regarding pawn moves from the first rank (and that they can only move one square ahead, not two)?
Well, the starting position and pieces are not legal in standard chess. If _chess_ engines allow for that, and for the moves of first-second rank pawns, it's just by sheer chance.

I believe according to the way the rules of chess are encoded into these particular engines, only pawns from the 2nd rank can move up 2 squares, and any other pawns, including incidental "illegal" 1st rank pawns, can only move 1 square per move. Somehow the engines that can play Theban chess must have a 'loophole' that allows one's pawns to be set up on the 1st rank.

I've always thought that rules of chess stipulated that each pawn may move up 2 squares on their first move from their original home square, which is always on the 2nd rank, whether in normal chess or FRC.

Regards,
CL

Uri Blass · Post by **Uri Blass** » Sun May 11, 2014 1:18 am

Laskos wrote:
bob wrote:
If "the rules are the same as standard chess" exactly how does standard chess mandate that a pawn on the 1st rank moves? Just one square ahead? two squares ahead? What about 2nd rank? Still allowed to move one or two ranks?
1st rank moves just one square ahead. 2nd rank (even the former 1st rank) one or two squares ahead. En passant applies as usual (and only applicable for the second rank).
First rank may be interpreted as assembly area, second rank is the charge

I am surprised that all engines understand the rules and have no problem
with the fact that white has more than 8 pawns.

Laskos · Post by **Laskos** » Sun May 11, 2014 2:29 am

Uri Blass wrote:
Laskos wrote:
bob wrote:
If "the rules are the same as standard chess" exactly how does standard chess mandate that a pawn on the 1st rank moves? Just one square ahead? two squares ahead? What about 2nd rank? Still allowed to move one or two ranks?
1st rank moves just one square ahead. 2nd rank (even the former 1st rank) one or two squares ahead. En passant applies as usual (and only applicable for the second rank).
First rank may be interpreted as assembly area, second rank is the charge
I am surprised that all engines understand the rules and have no problem
with the fact that white has more than 8 pawns.

Not all, some of them, luckily some top engines.

Laskos · Post by **Laskos** » Mon Jul 04, 2016 1:50 pm

That's the starting position of this game, as in the opening post.
[d]

Result of the WCTCC 2016 held on July 4, 2016 in my garage. Blitz games:

Code: Select all

    Program                            Score     %     Elo     Draws

  1 Stockfish 020716 64 BMI2       : 258.0/300  86.0   2262    6.0 %
  2 Komodo 10 64-bit               : 234.5/300  78.2   2181    6.3 %
  3 Andscacs 0.87                  : 143.0/300  47.7   1978    5.3 %
  4 DiscoCheck 5.2.1               : 125.0/300  41.7   1941    8.0 %
  5 Hakkapeliitta TCEC             : 106.5/300  35.5   1903    9.0 %
  6 Cheng 4.39                     : 103.5/300  34.5   1896    7.0 %
  7 Quazar 0.4 x64                 :  79.5/300  26.5   1840    9.0 %

Stockfish is again the champion.

Although engines seem to give about +1.00 advantage to White in the starting position, the game is fairly balanced, with a slight Black advantage:

Code: Select all

Games        :   1050 (finished)

White Wins   :    423 (40.3 %)
Black Wins   :    551 (52.5 %)
Draws        :     76 ( 7.2 %)

Very few draws, as can be seen.

Here is one game as an example, between Stockfish and Komodo:
[pgn][Event "?"]
[Site "?"]
[Date "2016.07.04"]
[Round "3"]
[White "Stockfish"]
[Black "Komodo"]
[Result "0-1"]
[FEN "1p6/2p3kn/3p2pp/4pppp/5ppp/8/PPPPPPPP/PPPPPPKN w - - 0 1"]
[PlyCount "102"]
[SetUp "1"]
[TimeControl "60+0.6"]

1. a4 {book} Nf6 {book} 2. a5 {book} Kf7 {book} 3. b4 {+1.32/19 2.6s}
Ke6 {-0.88/19 1.9s} 4. a2 {+1.37/19 1.1s} h3 {-0.86/20 2.3s}
5. c3 {+1.32/22 3.2s} h4 {-0.90/18 1.2s} 6. d3 {+1.29/21 1.3s}
hxg2 {-0.89/19 2.2s} 7. fxg2 {+1.32/21 0.83s} e4 {-0.96/21 5.9s}
8. d4 {+1.23/21 5.5s} Kd5 {-0.74/20 1.9s} 9. b2 {+1.07/22 5.8s}
c5 {-0.70/21 3.7s} 10. dxc5 {+1.14/20 3.3s} dxc5 {-0.71/20 1.1s}
11. b5 {+1.17/20 1.2s} Kc4 {-0.15/20 0.92s} 12. a4 {-0.96/22 2.8s}
Kb3 {-0.02/21 1.2s} 13. a6 {-1.18/22 3.1s} Nd7 {+0.33/21 3.6s}
14. a5 {-1.19/21 0.36s} Ne5 {+0.55/18 0.89s} 15. b6 {+0.15/21 2.9s}
Nc6 {+0.55/19 0.67s} 16. f3 {-0.34/21 2.5s} e3 {+0.91/18 2.1s}
17. fxg4 {-0.46/20 0.24s} fxg4 {+1.09/19 0.92s} 18. h3 {-0.17/23 0.81s}
g3 {+1.38/25 1.1s} 19. d2 {-2.28/24 2.7s} Kc4 {+1.53/25 0.86s}
20. a7 {-2.24/25 1.5s} bxa7 {+2.11/22 0.88s} 21. c2 {-2.50/28 4.4s}
Kb5 {+2.27/25 1.7s} 22. bxa7 {-3.12/28 9.0s} Nxa7 {+2.76/24 1.7s}
23. Kf1 {-3.16/25 1.3s} Nc6 {+2.88/24 0.68s} 24. Kg1 {-3.32/25 2.4s}
Nxa5 {+3.57/22 1.1s} 25. d3 {-3.79/26 7.0s} c4 {+3.67/20 0.58s}
26. Nf2 {-4.32/21 3.0s} exf2+ {+4.19/21 2.1s} 27. exf2 {-4.30/18 1.1s}
Nb7 {+4.34/23 2.9s} 28. Kf1 {-4.96/19 2.4s} Nc5 {+5.04/23 4.4s}
29. dxc4+ {-5.40/21 3.2s} Kxc4 {+5.68/22 2.2s} 30. e3 {-5.96/21 1.3s}
Ne4 {+6.77/23 4.1s} 31. exf4 {-6.31/23 1.3s} gxf4 {+7.43/24 3.4s}
32. fxg3 {-6.32/24 0.70s} hxg3 {+7.90/24 3.5s} 33. Ke2 {-5.83/17 0.60s}
Nd2 {+8.03/21 1.4s} 34. h4 {-8.45/16 0.54s} h5 {+9.86/20 1.1s}
35. Ke1 {-8.78/19 0.67s} f3 {+11.52/20 0.78s} 36. gxf3 {-22.51/20 0.60s}
Nxf3+ {+17.70/20 3.0s} 37. Kf1 {-50.87/26 0.60s} Nxh4 {+19.08/19 0.97s}
38. Kg1 {-53.12/25 0.60s} Nf5 {+38.40/20 1.3s} 39. Kf1 {-58.72/24 0.60s}
h4 {+M33/21 1.6s} 40. Kg2 {-61.85/23 0.60s} g5 {+M33/18 0.12s}
41. b4 {-63.08/21 0.60s} Ne3+ {+M29/17 0.14s} 42. Kf3 {-63.05/19 0.22s}
g4+ {+M27/14 0.47s} 43. Kf4 {-M18/24 0.81s} g2 {+M19/16 0.20s}
44. b5 {-M16/25 0.71s} g1=Q {+M17/13 0.12s} 45. b6 {-M14/26 0.66s}
Qg3+ {+M15/13 0.11s} 46. Kg5 {-M12/27 0.40s} Qe5+ {+M11/15 0.15s}
47. Kh6 {-M10/36 0.39s} h3 {+M9/18 0.14s} 48. Kg6 {-M8/82 0.40s}
h2 {+M7/28 0.12s} 49. b7 {-M6/109 0.087s} h1=Q {+M5/99 0.055s}
50. Kf7 {-M4/127 0.005s} Qh7+ {+M3/99 0.012s} 51. Kf8 {-M2/1 0s}
Qee7# {+M1/99 0.011s, Black mates} 0-1[/pgn]

Evert · Post by **Evert** » Mon Jul 04, 2016 2:42 pm

Interesting; I'll try this out with SjaakII later. Not that it is any sort of competition for the likes of Stockfish...

WCTCC (World Computer Theban Chess Championship)

WCTCC (World Computer Theban Chess Championship)

Re: WCTCC (World Computer Theban Chess Championship)

Re: WCTCC (World Computer Theban Chess Championship)

Re: WCTCC (World Computer Theban Chess Championship)

Re: WCTCC (World Computer Theban Chess Championship)

Re: WCTCC (World Computer Theban Chess Championship)

Re: WCTCC (World Computer Theban Chess Championship)

Re: WCTCC (World Computer Theban Chess Championship)

WCTCC 2016 - Stockfish is the champion again

Re: WCTCC 2016 - Stockfish is the champion again