TSCP zh

MikeB · Post by **MikeB** » Thu Dec 15, 2016 6:41 am

Recently came across TSCP zh (http://www.tckerrigan.com/Chess/TSCP/Community/) , and thought I would play a game against it, fully expecting to win. My gosh, it was much tougher than I expected. So then I matched sjeng 11.3 ( yes, if you download the source from Gian-Carlo Pascutto's GitHub page https://github.com/gcp/sjeng), it must even been updated to 11.3 about 5 years ago, thinking sjeng would wipe TSCP out ( at one time, sjeng was top rated in zh #1 over both humans and programs on both FICS and ICC.

Did not happen - stunning upset:

Code: Select all

[Event "Computer Chess Game"]
[Site "Mac-Pro.local"]
[Date "2016.12.15"]
[Round "-"]
[White "Sjeng 11.3"]
[Black "TSCP zh 1.1"]
[Result "0-1"]
[TimeControl "15+0"]
[Variant "crazyhouse"]
[Annotator "1. +1.09   1... +0.05"]

1. e3 {+1.09/8} Nc6 {+0.05/5 0.5} 2. Nf3 {+0.86/6 0.7} e5 {+0.03/5 0.5} 3.
Bc4 {+0.01/6 0.7} e4 {+0.23/4 0.5} 4. Nd4 {+0.38/7 1.4} Nxd4 {+0.45/5 0.5}
5. exd4 {+0.29/5 0.3} d5 {+0.37/4 0.5} 6. Bb5+ {+0.35/6 0.4} c6
{+0.85/5 0.5} 7. Bxc6+ {-0.46/5 0.4} bxc6 {+3.82/4 0.5} 8. N@e5
{-2.40/4 0.4} N@e7 {+4.23/3 0.4} 9. Na3 {-1.28/5 0.9} Ng6 {+4.00/4 0.4} 10.
Nxf7 {-0.39/6 0.7} Kxf7 {+6.52/4 0.4} 11. Rb1 {-3.96/6 0.7} N@f4
{+8.41/3 0.4} 12. P@f6 Nxg2+ {+11.19/3 0.4} 13. Kf1 {-9.30/4 0.3} P@f3
{+13.32/3 0.4} 14. Kg1 {-10.00/6 0.6} Bxa3 {+14.25/4 0.4} 15. bxa3
{-11.36/6 0.7} N@h3+ {+22.79/4 0.4} 16. Kf1 B@e2+ {+99.93/4 0.2} 17. Qxe2
fxe2+ {+99.95/3 0.1} 18. Kxe2 {-9999.95/4 0.3} Q@f3+ {+99.97/2 0.1} 19. Kf1
Qxf2# {+99.99/1 0.1}
{Black mates} 0-1

For those with a Mac, source and exe's for both engines.

https://www.dropbox.com/s/2xiaydxcjfxh8 ... g.zip?dl=1

edit: It might have been time control - 15sec/0.25

"Longer" 1/1 game below:

Code: Select all

[Event "Computer Chess Game"]
[Site "Mac-Pro.local"]
[Date "2016.12.15"]
[Round "-"]
[White "Sjeng 11.3"]
[Black "TSCP zh 1.1"]
[Result "1-0"]
[TimeControl "60+1"]
[Variant "crazyhouse"]
[Annotator "1. +1.11   1... -0.27"]

1. e3 {+1.11/8} e5 {-0.27/6 2.8} 2. Nc3 {+1.10/8 2.9} Nc6 {-0.18/6 2.7} 3.
Nf3 {+0.81/8 2.8} d5 {+0.06/5 2.7} 4. Bb5 {+0.70/8 2.7} e4 {-0.27/6 2.6} 5.
Ne5 {+1.65/8 2.6} Ne7 {-0.28/6 2.6} 6. Qh5 {+5.00/7 2.3} Be6 {-2.95/6 2.5}
7. Bxc6+ {+4.87/7 2.5} Nxc6 {-0.15/6 2.5} 8. Nxc6 {+0.26/6 2.4} bxc6
{+0.03/5 2.4} 9. N@f4 {+0.40/6 2.2} N@b7 {+0.12/4 2.4} 10. Nxe6
{+8.82/7 2.3} Qd6 {-5.98/5 2.3} 11. Qxf7+ {+10000.00/4 0.1} Kxf7
{+6.32/5 2.3} 12. N@g5+ {+3.63/6 2.3} Ke7 {+7.92/4 2.2} 13. B@a3
{+2.49/5 1.8} B@c5 {+5.17/5 2.2} 14. P@a6 {+2.60/7 2.2} Bxa3 {-2.88/5 2.2}
15. axb7 {+5.75/7 2.1} Rb8 {-4.73/5 2.1} 16. N@f5+ {+11.70/7 2.1} Kd7
{-4.93/5 2.1} 17. Nxd6 {+12.07/7 2.0} Kxd6 {-10.45/4 2.0} 18. bxa3
{+12.99/6 1.9} N@a6 {-11.53/3 2.0} 19. B@f4+ {+9999.90/6 1.9} Kd7
{-99.92/5 1.0} 20. Nxf8+ {+9999.92/6 1.8} Rbxf8 {-99.94/4 0.2} 21. Q@e6+
{+10000.00/3 0.2} Kd8 {-99.94/4 0.1} 22. B@e7+ {+10000.00/3 0.1} Ke8
{-99.96/3} 23. Bed6+ {+9999.96/5 0.1} N@e7 {-99.98/2 0.1} 24. Qxe7#
{+9999.98/3}
{White Mates} 1-0

MikeB · Post by **MikeB** » Thu Dec 15, 2016 7:00 am

ok , there was a little note in the read me file for sjeng to undefine this line

Code: Select all

#undef HAVE_FTIME

in config.h for greater time precision, mystery solved

Code: Select all

[Event "Computer Chess Game"]
[Site "Mac-Pro.local"]
[Date "2016.12.15"]
[Round "-"]
[White "Sjeng 11.3"]
[Black "TSCP zh 1.1"]
[Result "1-0"]
[TimeControl "15+0"]
[Variant "crazyhouse"]
[Annotator "1. +0.99   1... -0.30"]

1. Nf3 {+0.99/8} d5 {-0.30/6 0.5} 2. Nc3 {+0.74/7 0.7} d4 {+0.20/5 0.5} 3.
Nb5 {+1.17/7 1.4} Nc6 {+0.15/5 0.5} 4. Ng5 {+0.90/6 0.5} a6 {+1.35/5 0.5}
5. Nxc7+ {+0.07/6 0.3} Qxc7 {+4.26/5 0.5} 6. Nxf7 {+0.07/5 0.4} Kxf7
{+6.90/4 0.5} 7. P@d5 {-1.69/5 0.5} Nb4 {+6.83/4 0.5} 8. P@c3 {-2.33/6 0.5}
Nxd5 {+7.47/4 0.4} 9. cxd4 {-2.20/6 0.7} Bf5 {+7.90/3 0.4} 10. d3
{-2.87/6 0.4} P@e4 {+7.22/3 0.4} 11. dxe4 {+0.16/6 0.4} Bxe4 {+6.30/4 0.4}
12. P@e6+ {-1.21/5 0.3} Kxe6 {+7.42/4 0.4} 13. P@d3 {-1.50/5 0.3} P@c3
{+9.45/3 0.4} 14. dxe4 {+1.33/6 0.4} P@d2+ {+3.32/4 0.4} 15. Bxd2
{+4.05/5 0.4} cxd2+ {+3.37/4 0.4} 16. Qxd2 {+4.80/5 0.3} Ndf6 {+3.37/3 0.4}
17. B@f5+ {+3.99/5 0.4} Kf7 {+1.41/4 0.4} 18. P@e6+ {+10000.00/3 0.2} Ke8
{+1.08/5 0.4} 19. P@f7+ {+4.00/6 0.4} Kd8 {+1.08/4 0.4} 20. fxg8=R
{+3.44/6 0.4} Rxg8 {+1.32/3 0.4} 21. e3 {+1.50/6 0.4} P@g6 {+3.10/3 0.4}
22. Bxg6 {+1.62/6 0.4} hxg6 {+3.53/3 0.4} 23. N@f7+ {+2.03/5 0.4} Ke8
{+2.33/4 0.4} 24. Ne5 {+3.48/5 0.3} Kd8 {+0.00/3 0.4} 25. P@d7
{+2.53/5 0.4} N@h8 {+2.41/3 0.4} 26. Nf7+ {+2.35/6 0.3} Nxf7 {+0.14/5 0.4}
27. exf7 {+2.36/5 0.3} Qxd7 {+0.14/3 0.4} 28. fxg8=R {+2.92/5 0.4} Nxg8
{-3.61/3 0.4} 29. R@c8+ {+10000.00/4 0.2} Rxc8 {+6.49/4 0.3} 30. N@f7+
{-3.02/5 0.4} Ke8 {+8.92/5 0.3} 31. Bxa6 {-2.01/5 0.4} Kxf7 {+11.07/3 0.3}
32. P@h7 {-3.64/5 0.4} bxa6 {+9.34/3 0.3} 33. hxg8=N {-3.60/5 0.4} Kxg8
{+14.55/3 0.3} 34. N@b6 {-7.34/5 0.3} P@c3 {+12.53/3 0.3} 35. Qxc3
{-5.83/4 0.4} Rxc3 {+13.10/3 0.3} 36. Nxd7 {-5.83/3 0.3} P@d2+
{+11.50/2 0.3} 37. Kxd2 {-5.86/3 0.3} Rxc2+ {+10.30/2 0.3} 38. Kxc2
{-6.28/3 0.3} P@b3+ {+9.70/2 0.3} 39. axb3 {-4.06/3 0.3} P@d3+
{+3.62/2 0.3} 40. Kxd3 {-1.86/4 0.3} N@h7 {+3.62/2 0.3} 41. P@f7+
{+9995.00/5 0.4} Kxf7 {-99.94/4 0.2} 42. P@e6+ {+10000.00/3 0.2} Kg8
{-99.96/3 0.1} 43. Q@f7+ {+10000.00/2 0.1} Kh8 {-99.98/2} 44. R@g8#
{+9999.98/3 0.1}
{White Mates} 1-0

source and mac exe updated - same link as above

Ferdy · Post by **Ferdy** » Thu Dec 15, 2016 10:58 am

Thanks for the post, I am testing on blitz my compilation now vs the old 11.2.

Code: Select all

Score of Sjeng 11.3 32bit vs Sjeng 11.2 32bit: 6 - 7 - 0 [0.462]
Elo difference: -26.78 +/- 212.76

13 of 1040 games finished.

Ferdy · Post by **Ferdy** » Thu Dec 15, 2016 1:14 pm

2 Connection stalls from 11.2.

Code: Select all

[Event "My Tournament"]
[Site "i7-2600K"]
[Date "2016.12.15"]
[Round "1"]
[White "Sjeng 11.2 32bit"]
[Black "Sjeng 11.3 32bit"]
[Result "0-1"]
[PlyCount "122"]
[Termination "stalled connection"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. e4 {book} e5 {book} 2. Be2 {book} Be7 {book} 3. Nh3 {book} Nh6 {book}
4. f4 {book} f5 {book} 5. d4 {book} d5 {book} 6. Be3 {book} Be6 {book}
7. Kd2 {book} Kd7 {book} 8. Re1 {book} Re8 {book} 9. Bg1 {book} Bg8 {book}
10. Nf2 {book} Nf7 {book} 11. Nh1 {book} Nh8 {book} 12. Na3 {book} Na6 {book}
13. Kc1 {book} Kc8 {book} 14. Kb1 {book} Kb8 {book} 15. Qd3 {book} Qd6 {book}
16. Bd1 {book} Bd8 {book} 17. Nb5 {book} Nb4 {book} 18. Nc3 {book} Nc6 {book}
19. Ne2 {book} Ne7 {book} 20. Nc1 {book} Nc8 {book} 21. fxe5 {book} fxe4 {book}
22. Qxe4 {book} Qxe5 {book} 23. Qxd5 {book} Qxd4 {book} 24. Qf3 {book}
Qf6 {book} 25. Qf1 {book} Qf8 {book} 26. P@f2 {book} P@f7 {book} 27. P@e2 {book}
P@e7 {book} 28. P@d2 {book} P@d7 {book} 29. e4 {+0.93/10 9.0s} e6 {-0.24/9 7.8s}
30. Nb3 {+0.44/9 17s} Nb6 {-0.86/9 8.7s} 31. f3 {+0.70/8 7.9s}
Bh4 {-0.85/9 8.4s} 32. Bc5 {+0.57/9 6.2s} Be7 {-0.43/9 8.1s}
33. Bd4 {+0.73/8 7.4s} Rd8 {-0.44/8 7.8s} 34. Na5 {+0.75/8 7.1s}
Bb4 {-0.67/9 7.5s} 35. Bc3 {+0.37/8 6.9s} Bxa5 {+0.73/9 7.0s}
36. Bxa5 {-0.21/9 6.2s} N@c4 {-0.51/8 6.5s} 37. Bc3 {+0.47/8 6.4s}
Qd6 {-1.01/9 13s} 38. d3 {+1.75/9 6.2s} Nxb2 {-2.03/9 12s}
39. Bxb2 {+2.66/8 6.0s} Qf8 {-2.27/8 5.2s} 40. a4 {+2.64/8 5.8s}
P@d2 {-3.11/8 11s} 41. Re2 {+3.24/8 5.6s} f6 {-3.95/8 4.1s}
42. Rxd2 {+4.84/8 5.4s} Qd6 {-5.20/8 4.0s} 43. N@h5 {+4.05/7 3.8s}
Qxh2 {-3.05/8 3.9s} 44. P@e7 {+3.29/8 5.2s} Rc8 {-3.43/8 4.2s}
45. e8=Q {+4.19/9 5.0s} Rxe8 {-4.38/9 9.3s} 46. Nxg7 {+3.90/9 4.9s}
Re7 {-4.85/8 3.4s} 47. Bxf6 {+5.45/8 6.3s} P@e3 {-5.99/8 6.8s}
48. B@e5 {+7.86/8 4.2s} Qxe5 {-8.52/8 5.1s} 49. Bxe5 {+7.52/9 12s}
B@d8 {-10.15/7 3.8s} 50. Q@f8 {+10.33/8 3.9s} Nf7 {-12.52/7 3.5s}
51. Re2 {+12.38/8 3.8s} d6 {-13.01/7 2.8s} 52. Bb2 {+13.52/8 3.7s}
P@c4 {-14.22/9 3.6s} 53. Qxg8 {+15.05/9 3.6s} cxd3 {-15.48/9 7.1s}
54. cxd3 {+15.49/9 3.6s} P@c4 {-16.07/8 2.6s} 55. P@f6 {+14.70/8 3.5s}
Rd7 {-15.50/8 3.3s} 56. P@e7 {+15.53/8 3.4s} cxd3 {-15.68/8 3.2s}
57. Qxf7 {+14.93/7 3.2s} dxe2 {-15.47/7 3.1s} 58. exd8=Q+ {+15.03/6 2.6s}
R@c8 {-17.20/6 6.1s} 59. Qdxd7 {+16.94/6 8.6s} Nxd7 {-20.42/7 2.9s}
60. Qxe2 {+18.72/6 2.9s} P@d3 {-22.61/8 5.6s} 61. R@e8 {+21.49/6 2.9s}
P@c2+ {-28.77/6 4.3s, White's connection stalls} 0-1

Code: Select all

[Event "My Tournament"]
[Site "i7-2600K"]
[Date "2016.12.15"]
[Round "1"]
[White "Sjeng 11.2 32bit"]
[Black "Sjeng 11.3 32bit"]
[Result "0-1"]
[PlyCount "56"]
[Termination "stalled connection"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. d4 {book} d5 {book} 2. Nf3 {book} Bf5 {book} 3. Bg5 {book} Nc6 {-0.03/9 8.7s}
4. Nc3 {+0.23/9 8.1s} h6 {+0.22/9 6.3s} 5. Bh4 {+0.02/9 8.7s} Nb4 {-0.35/9 7.3s}
6. Rc1 {+1.00/9 8.4s} Nf6 {-0.67/8 8.2s} 7. Ne5 {+1.57/8 8.1s} e6 {-0.14/8 7.9s}
8. e3 {+1.70/8 7.8s} Bd6 {-0.42/7 7.6s} 9. Nxf7 {+1.29/8 7.5s}
Kxf7 {-0.44/8 7.3s} 10. P@e5 {+1.76/8 5.6s} Nxc2+ {-0.25/8 7.0s}
11. Rxc2 {+2.75/7 7.0s} Bxc2 {-3.24/7 5.1s} 12. Qxc2 {+3.20/7 6.8s}
N@b4 {-2.55/7 6.6s} 13. B@g6+ {+3.35/7 6.5s} Kg8 {-3.77/7 6.4s}
14. Qc1 {+3.36/7 4.7s} Nxa2 {-4.01/7 5.2s} 15. Nxa2 {+5.90/8 6.2s}
R@c6 {-5.73/7 5.5s} 16. Qd2 {+6.34/8 5.9s} P@f7 {-6.16/8 5.8s}
17. exd6 {+5.58/8 5.8s} fxg6 {-5.95/8 5.7s} 18. N@e7+ {+7.26/8 5.6s}
Qxe7 {-7.43/7 5.5s} 19. dxe7 {+7.47/7 5.4s} N@c2+ {-7.99/7 5.3s}
20. Kd1 {+7.75/6 5.2s} B@h5+ {-7.45/6 5.2s} 21. f3 {+7.33/6 5.1s}
g5 {-6.76/7 5.0s} 22. Q@d8+ {+7.43/7 4.9s} Rxd8 {-8.36/6 4.8s}
23. exd8=Q+ {+7.80/6 4.8s} Kh7 {-8.52/5 4.7s} 24. Bd3+ {+8.82/6 4.6s}
P@e4 {-9.37/6 9.1s} 25. N@f8+ {+8.61/6 3.3s} Rxf8 {-8.79/6 3.4s}
26. Qxf8 {+7.14/7 4.4s} N@g6 {-7.79/6 4.1s} 27. Qxf6 {+7.47/6 4.3s}
gxf6 {-7.27/6 4.0s} 28. R@d7+ {+10.93/6 4.2s}
P@e7 {-7.25/6 7.8s, White's connection stalls} 0-1

Test stopped.

Code: Select all

Score of Sjeng 11.3 32bit vs Sjeng 11.2 32bit: 54 - 54 - 2 [0.500]
Elo difference: 0.00 +/- 64.99

111 of 1040 games finished.

I am going to use 11.3 in the champ's tour.

Guenther · Post by **Guenther** » Thu Dec 15, 2016 2:53 pm

Ferdy wrote: I am going to use 11.3 in the champ's tour.

Will you make your compilation of Sjeng 11.3 available?
(it seems it could be more stable?)

Carlos777 · Post by **Carlos777** » Thu Dec 15, 2016 2:54 pm

Ferdy wrote: I am going to use 11.3 in the champ's tour.

Hi Ferdy,

Please, share the windows binary.

Thanks in advance,
Carlos

Ferdy · Post by **Ferdy** » Thu Dec 15, 2016 7:07 pm

Guenther wrote:
Ferdy wrote: I am going to use 11.3 in the champ's tour.
Will you make your compilation of Sjeng 11.3 available?
(it seems it could be more stable?)

Try here.
https://drive.google.com/file/d/0BwAOsu ... sp=sharing

Guenther · Post by **Guenther** » Thu Dec 15, 2016 9:04 pm

Ferdy wrote:
Guenther wrote: Will you make your compilation of Sjeng 11.3 available?
(it seems it could be more stable?)
Try here.
https://drive.google.com/file/d/0BwAOsu ... sp=sharing

Thanks Ferdinand!

TSCP zh

TSCP zh

Re: TSCP zh

Re: TSCP zh

Re: TSCP zh

Re: TSCP zh

Re: TSCP zh

Re: TSCP zh

Re: TSCP zh