Interfacing the Novag Citrine

[sje]

Could you please give details on this cable ? Is it something home made or bought somewhere ?

I believe the cable is a simple six-conductor, straight-through item. A regular four-conductor cable will probably not work.

[sje]

Here's a maximal mating instance of the KQKR endgame (Mate In 35):
[D]K1Q5/8/8/8/8/8/3k4/r7 w - - 0 1
And the best play:

Code: Select all

1. Kb7 {Kb8} Rb1+ 2. Ka6 {Ka7} Ra1+ 3. Kb5 {Kb6} Rb1+ 4. Ka4 {Ka5} Ra1+ 5. Kb3
{Kb4} Rb1+ 6. Ka3 Rf1 7. Kb2 {Qc4 Qc5 Qc6 Qc7 Qd7+ Qe6 Qh3} Rf3 8. Qa8 {Qc4 Qc5
Qc6 Qe6 Qg4} Ke2 {Ke3 Rd3 Rg3} 9. Qe4+ Re3 10. Qc2+ {Qg2+} Ke1 {Kf1} 11. Kc1
Re2 12. Qa4 {Qc5 Qd3} Kf2 13. Qf4+ Ke1 14. Qd4 {Qg3+ Qh4+} Kf1 15. Kd1 {Qc4
Qh4} Re1+ 16. Kc2 {Kd2} Re2+ 17. Kd3 Rh2 18. Qa1+ {Qg4} Kf2 19. Qb2+ {Qf6+} Kg1
{Kg3} 20. Qe5 Kg2 {Ra2 Rh3+} 21. Ke2 Kg1+ 22. Kf3 Rh3+ 23. Kg4 Rh2 24. Kg3 Rg2+
25. Kh3 Kf1 26. Qa1+ Kf2 27. Qb2+ Ke3 28. Kxg2 Kf4 29. Qb5 {Qb8+ Qd4+ Qe2 Qf6+}
Ke4 30. Kf2 {Kg3 Qc5 Qd7} Kd4 31. Kf3 Kc3 32. Ke3 Kc2 33. Qb4 Kc1 34. Kd3 Kd1
35. Qb1# {Qd2#}

I tried the above position with the Citrine in autoplay mode at a game/hour time control. The unit gets a draw by repetition. Few humans could do better.

Code: Select all

[Event "Autoplay"]
[Site "At home"]
[Date "2008.10.21"]
[Round "-"]
[White "Novag Citrine"]
[Black "Novag Citrine"]
[Result "1/2-1/2"]
[Setup "1"]
[FEN "K1Q5/8/8/8/8/8/3k4/r7 w - - 0 1"]

1. Kb7 Rb1+ 2. Ka6 Rc1 3. Qg4 Rb1 4. Ka5 Rb3 5. Qd4+ Kc2 6. Ka4 Rb1 7. Qc4+ Kd2
8. Ka5 Rb2 9. Qd4+ Kc2 10. Ka4 Rb1 11. Qe4+ Kb2 12. Kb4 Rc1 13. Qe2+ Rc2 14.
Qe4 Rc1 15. Ka4 Rb1 16. Qd4+ Kc2 1/2-1/2

[sje]

Here's a maximal mating instance of the KRPKR endgame (Mate In 74):
[D]7k/8/8/8/8/3R4/1P6/r5K1 w - - 0 1
And the best play:

Code: Select all

1. Kf2 Kg7 {Kg8} 2. Ke3 Kf6 {Kf7} 3. Kd4 Ke6 4. Kc5 Ra5+ 5. Kb6 {Kc6} Ra8 6.
Kb7 Ra5 7. Rd2 Ra2 {Ra4 Rc5} 8. Kc6 Ra6+ 9. Kb5 Ra8 10. Rh2 Rf8 {Rg8} 11. Kc6
Rb8 12. Rh6+ Ke7 {Kf7} 13. Rh7+ Ke6 14. Rh2 Ke7 15. Kc7 Rb3 {Rb4 Rb5} 16. Re2+
Kf7 17. Kc6 Rb8 18. Kc5 {Kd5} Rc8+ 19. Kb4 {Kb5 Kd4 Kd5} Rb8+ 20. Kc3 {Kc4}
Rc8+ 21. Kd2 {Kd3} Rb8 {Rd8+} 22. Kc2 Rc8+ 23. Kb1 Rb8 24. Re3 Rd8 {Rg8 Rh8}
25. Kc2 Rc8+ 26. Kd3 Rb8 27. Kc3 Rc8+ 28. Kd4 Rd8+ 29. Kc5 Rc8+ 30. Kd6 Rb8 31.
b3 Rb5 32. Kc6 {Rd3} Rb8 33. Rd3 Kf8 34. Kc5 Ke7 {Rc8+} 35. b4 Rc8+ 36. Kb5
Rb8+ 37. Ka4 {Kc4} Ra8+ 38. Kb3 Rb8 39. Rd4 Ke6 40. Kc4 Ke5 41. Rd5+ Ke6 42.
Rc5 {b5} Kd6 {Kd7 Ke7 Ra8 Rb6 Rb7 Rd8 Re8 Rf8 Rg8 Rh8} 43. b5 Kd7 {Ke6 Ke7 Ra8
Rb6 Re8 Rf8 Rg8 Rh8} 44. Kb4 Kd6 {Ke6 Ke7 Rb7} 45. Rc1 {Rc2 Rc3} Kd7 {Rb7} 46.
Ka5 Ra8+ 47. Kb6 Rb8+ 48. Ka6 Ra8+ 49. Kb7 Ra2 {Ra3 Ra4} 50. Rb1 {Rd1+ b6} Kd6
{Kd8 Ke6 Ke7 Ke8 Ra3 Ra4 Ra5 Rc2 Rd2 Re2 Rf2 Rg2 Rh2} 51. Kb8 {b6} Kc5 {Kd5 Kd7
Ke5 Ke6 Ke7 Ra3 Ra4 Ra5 Rc2 Rd2 Re2 Rf2 Rg2 Rh2} 52. b6 Kc6 {Kd6 Re2 Rf2 Rg2
Rh2} 53. b7 Kd7 {Re2 Rf2 Rg2 Rh2} 54. Rd1+ Ke6 {Ke7} 55. Rd4 Ra1 {Ra3 Rb2} 56.
Kc7 Rc1+ 57. Kb6 Rf1 {Rg1 Rh1} 58. Kc6 {Rb4} Rf8 59. Ra4 {Rb4 Rd5} Rb8 60. Ra5
Rd8 {Rf8 Rg8 Rh8} 61. Ra8 Rd1 62. Re8+ Kf7 63. b8=Q Rc1+ 64. Kd5 Rd1+ 65. Kc4
Rc1+ 66. Kd3 Rd1+ 67. Ke2 Ra1 {Rb1 Rd5 Rh1} 68. Re3 {Re5} Ra2+ {Ra4 Rg1} 69.
Kf3 Ra5 {Ra6 Ra8 Rc2 Rd2} 70. Qe8+ Kg7 71. Re7+ Kf6 {Kh6} 72. Re6+ {Rf7+} Kg5
{Kg7} 73. Qg6+ {Qg8+} Kh4 74. Qg4#

I tried the above position with the Citrine in autoplay mode at a 40/9000 time control. After White drops its pawn, the unit gets a draw by repetition.

Code: Select all

[Event "Autoplay"]
[Site "At home"] 
[Date "2008.10.21"]
[Round "-"]
[White "Novag Citrine"]
[Black "Novag Citrine"] 
[Result "1/2-1/2"]
[Setup "1"]
[FEN "7k/8/8/8/8/3R4/1P6/r5K1 w - - 0 1"]

1. Kf2 Kg7 2. Ke3 Kf6 3. Kd4 Rb1 4. b3 Ke6 5. Rg3 Rb2 6. Kc3 Rb1 7. Rg2 Kd5 8.
b4 Kc6 9. Kc4 Rc1+ 10. Kb3 Kd5 11. Rg3 Rb1+ 12. Kc3 Ke5 13. Kc4 Rc1+ 14. Kd3
Rb1 15. Kc3 Kd5 16. Rg6 Rf1 17. Rg5+ Kc6 18. b5+ Kb6 19. Kb2 Rf3 20. Re5 Rg3
21. Rf5 Re3 22. Rg5 Rf3 23. Rh5 Rg3 24. Re5 Rf3 25. Rg5 Re3 26. Rh5 Rg3 27. Rd5
Ka5 28. Rf5 Ka4 29. Kc2 Rb3 30. Rf4+ Ka3 31. Rf7 Rxb5 32. Kc3 Rb4 33. Ra7+ Ra4
34. Rd7 Rb4 35. Rf7 Rg4 36. Ra7+ Ra4 37. Re7 Rf4 38. Ra7+ Ra4 1/2-1/2

[sje]

This was an eight hour game, a Sicilian won by White with the first 40+ ply coming out of the opening book. Black resigned on its 42nd move, but the game was played to mate for testing purposes.

Code: Select all

[Event "Autoplay"]
[Site "At home"]
[Date "2008.10.22"]
[Round "-"]
[White "Novag Citrine"]
[Black "Novag Citrine"]
[Result "1-0"]
[TimeControl "40/9000"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bg5 e6 7. f4 Be7 8. Qf3
Qc7 9. O-O-O Nbd7 10. g4 b5 11. Bxf6 Nxf6 12. g5 Nd7 13. f5 Nc5 14. f6 gxf6 15.
gxf6 Bf8 16. Rg1 Bd7 17. Rg7 Bxg7 18. fxg7 Rg8 19. e5 O-O-O 20. exd6 Qb6 21.
Ne4 Nxe4 22. Qxe4 Qb7 23. Qxh7 Qd5 24. a4 Qe5 25. Qh3 Qxd6 26. axb5 axb5 27.
Qf3 Be8 28. Rd2 Qb6 29. Qf6 Rd5 30. Bg2 Rd6 31. Rd3 b4 32. Kb1 Qc7 33. Be4 Qd8
34. Qe5 f6 35. Qc5+ Qc7 36. Qxb4 Rb6 37. Qa3 Qxh2 38. Rc3+ Kd8 39. Qa7 Qg1+ 40.
Ka2 Ra6+ 41. Qxa6 Qxd4 42. Rd3 Qxd3 43. Qxd3+ Bd7 44. Qd6 f5 45. Qf8+ Rxf8 46.
gxf8=Q+ Kc7 47. Bg2 e5 48. Qc5+ Kd8 49. Qxe5 Kc8 50. Kb3 Kd8 51. Bh3 Kc8 52.
Qd6 Bb5 53. Bxf5+ Kb7 54. Be4+ Kc8 55. Qc5+ Kd7 56. Qxb5+ Ke7 57. Qe5+ Kd7 58.
Kc4 Kd8 59. Qd6+ Kc8 60. Qe7 Kb8 61. Qb7# 1-0

[bob]

Here's a maximal mating instance of the KQKR endgame (Mate In 35):
[D]K1Q5/8/8/8/8/8/3k4/r7 w - - 0 1
And the best play:

Code: Select all

1. Kb7 {Kb8} Rb1+ 2. Ka6 {Ka7} Ra1+ 3. Kb5 {Kb6} Rb1+ 4. Ka4 {Ka5} Ra1+ 5. Kb3
{Kb4} Rb1+ 6. Ka3 Rf1 7. Kb2 {Qc4 Qc5 Qc6 Qc7 Qd7+ Qe6 Qh3} Rf3 8. Qa8 {Qc4 Qc5
Qc6 Qe6 Qg4} Ke2 {Ke3 Rd3 Rg3} 9. Qe4+ Re3 10. Qc2+ {Qg2+} Ke1 {Kf1} 11. Kc1
Re2 12. Qa4 {Qc5 Qd3} Kf2 13. Qf4+ Ke1 14. Qd4 {Qg3+ Qh4+} Kf1 15. Kd1 {Qc4
Qh4} Re1+ 16. Kc2 {Kd2} Re2+ 17. Kd3 Rh2 18. Qa1+ {Qg4} Kf2 19. Qb2+ {Qf6+} Kg1
{Kg3} 20. Qe5 Kg2 {Ra2 Rh3+} 21. Ke2 Kg1+ 22. Kf3 Rh3+ 23. Kg4 Rh2 24. Kg3 Rg2+
25. Kh3 Kf1 26. Qa1+ Kf2 27. Qb2+ Ke3 28. Kxg2 Kf4 29. Qb5 {Qb8+ Qd4+ Qe2 Qf6+}
Ke4 30. Kf2 {Kg3 Qc5 Qd7} Kd4 31. Kf3 Kc3 32. Ke3 Kc2 33. Qb4 Kc1 34. Kd3 Kd1
35. Qb1# {Qd2#}

I tried the above position with the Citrine in autoplay mode at a game/hour time control. The unit gets a draw by repetition. Few humans could do better.

Code: Select all

[Event "Autoplay"]
[Site "At home"]
[Date "2008.10.21"]
[Round "-"]
[White "Novag Citrine"]
[Black "Novag Citrine"]
[Result "1/2-1/2"]
[Setup "1"]
[FEN "K1Q5/8/8/8/8/8/3k4/r7 w - - 0 1"]

1. Kb7 Rb1+ 2. Ka6 Rc1 3. Qg4 Rb1 4. Ka5 Rb3 5. Qd4+ Kc2 6. Ka4 Rb1 7. Qc4+ Kd2
8. Ka5 Rb2 9. Qd4+ Kc2 10. Ka4 Rb1 11. Qe4+ Kb2 12. Kb4 Rc1 13. Qe2+ Rc2 14.
Qe4 Rc1 15. Ka4 Rb1 16. Qd4+ Kc2 1/2-1/2

however, that is a disappointing result. This is a trivial ending for a computer, if you have two key pieces of evaluation information:

(1) losing king has to be driven to the edge of the board;

(2) winning king has to come to help mate the king by getting close.

Amir Ban and I discussed this opening many years ago in r.g.c.c. He thought it was easy while I thought it would be nearly impossible to win without an endgame table. I tested it with crafty and at 35K node per second speeds, at 1 second per move, it had no problem forcing mate. It didn't play perfectly according to endgame table best moves, but it always won without reaching a repetition or 50 move draw. I played several hundred games with craftyA . 1 sec per move, no EGTB, craftyB using (at the time) your endgame tables. I was really surprised at how easy it was for the computer...

[sje]

however, that is a disappointing result. This is a trivial ending for a computer, if you have two key pieces of evaluation information:

(1) losing king has to be driven to the edge of the board;

(2) winning king has to come to help mate the king by getting close.

Amir Ban and I discussed this opening many years ago in r.g.c.c. He thought it was easy while I thought it would be nearly impossible to win without an endgame table. I tested it with crafty and at 35K node per second speeds, at 1 second per move, it had no problem forcing mate. It didn't play perfectly according to endgame table best moves, but it always won without reaching a repetition or 50 move draw. I played several hundred games with craftyA . 1 sec per move, no EGTB, craftyB using (at the time) your endgame tables. I was really surprised at how easy it was for the computer...

I have some ideas here concerning the Citrine, but they are only speculation; we really need Dave Kittinger here to comment authoritatively.

1) The Citrine node frequency in the endgame is about 8 KHz, and that's a fairly strong limit on effective analysis depth in a full width searcher. Its H8/300 16 bit CPU runs at about 10 MIPS, so there's about 1,250 instruction executions available per endgame node. Are these enough for extensive, class specific, endgame play?

2) I assume that the Citrine, like its Super Constellation grandpappy, uses significant resources (code and time) with root level prepreprocessing. But what works well in the middlegame may not always be as effective in the endgame.

3) The Citrine program ROM space is limited to 60 KB, and 24 KB of that is consumed by the opening book. That's not a very big book by desktop standards, and to make it smaller to include specific endgame class recognizers might have been too risky. On the other hand, I believe that the Citrine has at least a KBNK recognizer as the machine does very well with that class.

My experience so far with the Citrine is that it has been optimized for play against humans at the expense of strength vs other programs. This is not a bad decision given the target audience. I'm guessing that it would score about fifty percent in a long match with a 2200 elo human, but not as well against a typical 2200 elo program. The marketing claim of 2330 elo is an advertising fantasy typical of the commercial chess computer field.

It's still a fun machine, though. I'm glad I bought one and I recommend it to others.

[Steve B]

it might be of interest to look into Kittingers Wchess PC Engine for possible insights into his Citrine program
i would imagine much of the basic programming is the same
in fact i doubt there is much difference in any of his Novag programs since the mid 1990's..only real differences being the hardware and housing of the dedicated units sold
as far as i know his most recent incarnation of Wchess appears in the Majestic Chess software package commonly offered on Ebay

W Regards
Steve

[sje]

it might be of interest to look into Kittingers Wchess PC Engine for possible insights into his Citrine program
i would imagine much of the basic programming is the same

Perhaps there are many similarities with data representation. However, any desktop program with a RAM space ten thousand times greater than the 3 KB Citrine limit will avail itself of all kinds of memory consuming programming techniques. And then there's the matter of using a disk for the opening book and tablebases.

[sje]

Here's anonther example of Citrine autoplay at a 40/9000 time control. This time, the game took about nine hours and it ended with a draw by insufficient material.

Black played 50... axb3, an en passant capture. New to me, the Citrine serial interface outputs this as "a4xb3ep". The needed adjustment to the Lisp reading code has been made.

Code: Select all

[Event "Autoplay"]
[Site "At home"]
[Date "2008.10.24"]
[Round "-"]
[White "Novag Citrine"]
[Black "Novag Citrine"]
[Result "1/2-1/2"]
[TimeControl "40/9000"]

1. e4 e6 2. d4 d5 3. Nd2 c5 4. exd5 exd5 5. Ngf3 Nc6 6. Bb5 Bd6 7. O-O Ne7 8.
dxc5 Bxc5 9. Nb3 Bd6 10. Re1 O-O 11. Bd3 h6 12. h3 Bd7 13. Be3 Re8 14. Nc5 Qc8
15. Nxd7 Qxd7 16. c3 Nf5 17. Bxf5 Qxf5 18. Qb3 Qd7 19. Rad1 Ne7 20. Nd4 Be5 21.
Nc2 a5 22. Nd4 Rec8 23. Qb6 a4 24. Qb5 Qc7 25. Qe2 Re8 26. Qg4 Ng6 27. Qe2 Ne7
28. g3 Bf6 29. Qg4 Ng6 30. f4 Qc4 31. a3 b5 32. Bf2 Ne7 33. Qd7 Rab8 34. Nf5
Rbd8 35. Nxe7+ Bxe7 36. Qb7 Bc5 37. Bxc5 Qxc5+ 38. Kg2 Qc4 39. Re7 Rb8 40. Qd7
Rxe7 41. Qxe7 Qe4+ 42. Qxe4 dxe4 43. Rd5 g6 44. Re5 f5 45. g4 Rd8 46. Rxb5 Kg7
47. Rb7+ Kf6 48. Rb6+ Kf7 49. Rb7+ Kf6 50. b4 axb3 51. Rb6+ Kf7 52. Rxb3 Rd2+
53. Kg1 e3 54. Rb7+ Kf8 55. Rb8+ Kf7 56. Rb7+ Kf8 57. Kf1 Rf2+ 58. Ke1 Rxf4 59.
Ke2 Re4 60. gxf5 gxf5 61. Rb4 Re5 62. a4 Rc5 63. Kd3 e2 64. Kxe2 Rxc3 65. Rb5
Rxh3 66. Rxf5+ Ke7 67. a5 Ra3 68. Kf2 Kd6 69. Rb5 Kc6 70. Rh5 Kb7 71. Ke2 Ka6
72. Rxh6+ Kxa5 73. Rh7 Rb3 74. Rf7 Kb6 75. Rf3 Rxf3 76. Kxf3 1/2-1/2

[sje]

Two more 40/9000 Citrine autoplay games:

Code: Select all

[Event "Autoplay"]
[Site "At home"]
[Date "2008.10.24"]
[Round "-"]
[White "Novag Citrine"]
[Black "Novag Citrine"]
[Result "1-0"]

1. Nf3 Nf6 2. g3 g6 3. Bg2 Bg7 4. O-O O-O 5. d3 d6 6. e4 c5 7. h3 Bd7 8. e5 Ne8
9. exd6 Nxd6 10. Re1 Nc6 11. Nc3 h6 12. Bf4 Qc8 13. Kh2 Rd8 14. a3 Be6 15. Qd2
Nf5 16. Rad1 Ncd4 17. Nxd4 Rxd4 18. Be4 Nd6 19. Bxh6 Bxh3 20. Bxg7 Kxg7 21. Qg5
Bg4 22. Rc1 Qh8+ 23. Kg1 Qh5 24. Qxh5 Bxh5 25. Bg2 Re8 26. f3 f5 27. Re5 b6 28.
Rce1 Nc8 29. Nd5 Kf7 30. Kf2 Ra4 31. Bh3 e6 32. Nf4 Rh8 33. Rxe6 c4 34. d4 c3
35. bxc3 Bxf3 36. Kxf3 g5 37. Rc6 gxf4 38. Bxf5 Ne7 39. Rc7 Kf6 40. Bg4 Ng6 41.
Re6+ Kg5 42. gxf4+ Kh6 43. Re2 Nh4+ 44. Kg3 Rxa3 45. Kxh4 Kg6+ 46. Kg3 Rxc3+
47. Rxc3 Rf8 48. Rc7 Rf7 49. Rxf7 Kxf7 50. d5 a6 51. d6 Kf8 52. d7 Kf7 53. d8=Q
b5 54. Re7+ Kf6 55. Qd6# 1-0

[Event "Autoplay"]
[Site "At home"]
[Date "2008.10.24"]
[Round "-"]
[White "Novag Citrine"]
[Black "Novag Citrine"]
[Result "0-1"]

1. c4 e5 2. Nc3 Nc6 3. Nf3 Nf6 4. g3 Bb4 5. Bg2 O-O 6. O-O e4 7. Ng5 Bxc3 8.
bxc3 Re8 9. Qc2 d5 10. cxd5 Qxd5 11. d3 Bf5 12. Qb2 Rab8 13. Be3 Qd7 14. Bxe4
Bxe4 15. Nxe4 Nxe4 16. dxe4 Rxe4 17. Rfd1 Qe8 18. Rab1 b6 19. Qc2 a6 20. Qd3
Ra8 21. Qd7 Re7 22. Qxe8+ Raxe8 23. Kf1 f6 24. h3 Kf7 25. Kg2 Ke6 26. a4 Rd7
27. Rxd7 Kxd7 28. Rd1+ Kc8 29. Kf1 Re5 30. c4 Ra5 31. Ra1 Ne5 32. Bd2 Rc5 33.
a5 Nxc4 34. Bb4 Rb5 35. Bf8 Rxa5 36. Rxa5 Nxa5 37. Bxg7 f5 38. g4 fxg4 39. hxg4
Nc4 40. Kg2 a5 41. e4 b5 42. Kf3 a4 43. Ke2 a3 44. Kd3 c5 45. f4 Kd7 46. Kc3
Na5 47. Kc2 b4 48. Kb1 b3 49. Bc3 Nc4 50. g5 Nd6 51. e5 Ne4 52. Ba5 Ke6 53. Ka1
Kf5 54. Be1 a2 55. Ba5 c4 56. Bb4 c3 57. Ba3 Nd2 58. e6 Nf3 59. Bb2 Nd4 60.
Bxc3 Nc2+ 61. Kb2 a1=Q+ 62. Kxb3 Qb1+ 63. Bb2 Nd4+ 64. Ka3 Nb5+ 65. Kb3 Kxe6
66. Kc4 Qxb2 67. Kd3 Qc3+ 68. Ke2 Nd4+ 69. Kf2 Qf3+ 70. Kg1 Ne2+ 71. Kh2 Qg3+
72. Kh1 Qh3# 0-1