Rebel TNG (version 1.0) approximately 2950....

[carldaman]

Pretty sure that ProDeo 1.88 was okay too?
Graham.

Hi Graham, I am not sure - how to stop this engine using all the databases?

PS: here Rebel TNG uses 58% of a quad - more than 2 CPUs

regards
Werner

You can not turn off the database search, even if you would manage it will play exactly as ProDeo 1.88. If you don't like the program in its current form siimply don't test it

Hi Ed,

I am curious if there was something special and indispensable about Prodeo v1.88 (as opposed to the earlier versions 1.86 and 1.87) so that perhaps Rebel TNG could be based solely on it... (?)

Uh, no.

The reason I'm asking is that v1.88 seems to be markedly weaker than the previous ones (about 100 points weaker on the CCRL 40/40 list). Could Rebel TNG have been even stronger if based on either prior version instead of 1.88?

Yes, the 1.88 results are odd. I think I messed up, don't know how. And since TNG is based on 1.88 (no other change than adding the pattern knowledge) you could be right that I took the wrong version as a base for TNG.

Thanks,CL

You are welcome Carl.

Ed, I hope it was useful to point this out to you A new paradigm is always refreshing...

In any case, it's great to see the Rebel name back in business!

I've been running tests continuously since yesterday. One time forfeit so far, straight out of book, otherwise all looks fine. I'll let you know if I see more.

Cheers,
CL

[mclane]

Somebody said today is 1st April.

[mclane]

The strongest version so far was 1.84.

[carldaman]

The strongest version so far was 1.84.

On CCRL's 40/40 list v1.86 is the strongest. However, I don't see v1.84 listed.
CEGT 40/20 has v1.85 as the strongest.

[carldaman]

In games with repeating time controls, such as 40 moves in 20 minutes, Rebel TNG will start moving almost instantly (spending exactly 1 sec per move) after move 40, move 80, etc. for an indefinite number of moves.

This seriously cripples its performance, as one would expect.

One example below (but it behaves like this in every game):

[pgn]

[Event "T 40-20"]
[Date "2014.03.31"]
[Round "3"]
[White "Rebel TNG"]
[Black "Rodent 1.3 x64"]
[Result "0-1"]
[BlackElo "2750"]
[ECO "C08"]
[Opening "French"]
[Time "05:53:52"]
[Variation "Tarrasch, Open, Main Line, 4.exd5 exd5 5.Ngf3 Nf6, 10.Nb3"]
[WhiteElo "2800"]
[TimeControl "40/1200:40/600:40/300"]
[Termination "adjudication"]
[PlyCount "133"]
[WhiteType "program"]
[BlackType "program"]

1. e4 e6 2. d4 d5 3. Nd2 c5 4. exd5 {+0.76/16 40} exd5 {-0.08/18 28} 5.
Ngf3 {+0.80/15 22} Nf6 {0.00/18 29} 6. Bb5+ {+0.58/15 32} Bd7 {0.00/20 30}
7. Bxd7+ {+0.64/18 34} Nbxd7 {0.00/22 29} 8. O-O {+0.17/17 23} Be7 {0.00/22
49} 9. dxc5 {+0.50/17 36} Nxc5 {+0.04/21 28} 10. Nb3 {+0.50/16 25} Ne6
{-0.04/21 32} 11. Be3 {+0.44/16 32} O-O {+0.04/19 27} 12. Re1 {+0.63/16 35}
Re8 {0.00/18 28} 13. c3 {+0.58/15 54} a6 {0.00/18 26} 14. Nfd4 {+0.69/15
27} Qc7 {0.00/18 24} 15. Nf5 {+0.66/15 23} Bf8 {0.00/20 29} 16. f3
{+0.74/15 26} g6 {-0.08/19 33} 17. Nh6+ {+0.76/16 28} Kh8 {-0.08/21 36} 18.
Qd2 {+0.80/16 30} Bg7 {-0.11/20 33} 19. Bf2 {+0.77/16 35} Nh5 {-0.11/19 26}
20. Ng4 {+0.71/17 35} Nhf4 {-0.11/20 33} 21. Rad1 {+0.32/15 49} Rad8
{-0.11/19 32} 22. Bg3 {+0.98/16 29} Qc4 {-0.12/21 35} 23. Ne5 {+0.96/16 24}
Bxe5 {-0.32/22 26} 24. Rxe5 {+0.87/12 2} Kg8 {-0.24/21 26} 25. Kh1
{+0.94/16 26} f6 {-0.44/22 34} 26. Ree1 {+0.97/18 34} Nh5 {-0.36/21 25} 27.
Bf2 {+0.91/18 43} Nhg7 {-0.44/22 32} 28. Kg1 {+0.77/16 71} Nf5 {-0.40/21
35} 29. Bb6 {+0.72/16 41} Rd6 {-0.10/22 29} 30. Bc5 {+0.35/16 21} Rdd8
{-0.10/25 34} 31. Bf2 {+0.37/16 32} h5 {-0.24/21 33} 32. Bb6 {+0.92/16 23}
Rd6 {-0.10/24 28} 33. Bd4 {+0.33/15 24} Kf7 {-0.08/21 29} 34. Bf2 {+0.85/16
27} b6 {-0.10/21 31} 35. Qc2 {+0.81/16 28} h4 {-0.09/22 35} 36. Kh1
{+0.09/16 53} Qb5 {-0.04/21 40} 37. Qe2 {+0.81/14 17} Qc4 {-0.10/21 33} 38.
Kg1 {+0.64/16 25} Re7 {-0.08/22 42} 39. Qxc4 {+0.19/16 24} dxc4 {-0.08/25
23} 40. Rxd6 {+0.52/20 23} Nxd6 {-0.16/24 36} 41. Nd2 {+0.37/13 1} h3
{-0.12/22 15} 42. Bg3 {+0.17/12 1} Rd7 {-0.12/21 15} 43. Bxd6 {+0.73/13 1}
Rxd6 {-0.03/25 16} 44. Nxc4 {+0.66/14 1} Rd3 {-0.03/23 23} 45. Nxb6
{-0.21/13 1} hxg2 {-0.03/24 15} 46. Nc4 {0.00/14 1} Nf4 {+0.24/22 13} 47.
a3 {0.00/15 1} Rd5 {+0.40/22 19} 48. a4 {+0.30/13 1} Kg7 {+0.40/24 15} 49.
h4 {+0.32/13 1} g5 {+1.40/23 14} 50. h5 {0.00/14 1} Kh6 {+1.60/23 14} 51.
Kh2 {-0.01/15 1} Kxh5 {+1.56/23 15} 52. Kg1 {-0.42/13 1} Rd3 {+3.04/23 11}
53. Ne3 {-1.34/14 1} Rd2 {+3.48/24 11} 54. Kh2 {-2.25/14 1} Rf2 {+3.60/23
12} 55. Nf5 {-2.40/14 1} Rxf3 {+5.16/23 15} 56. Kg1 {-3.03/15 1} Rh3
{+6.04/22 21} 57. Kf2 {-2.72/15 1} g4 {+6.12/21 23} 58. Rg1 {-3.23/14 1}
Rf3+ {+7.52/22 12} 59. Ke1 {-2.86/7 0} g3 {+8.08/21 13} 60. Kd2 {-6.66/17
10} Kg5 {+8.92/21 23} 61. Nxg3 {-7.39/18 23} Rd3+ {+10.16/22 14} 62. Kc2
{-6.95/20 36} Rxg3 {+11.08/22 15} 63. Rb1 {-7.96/20 30} Nh3 {+14.48/20 16}
64. c4 {-6.89/19 29} Rf3 {+18.32/20 14} 65. Rd1 {-14.51/18 18} Rf1
{+19.72/19 16} 66. a5 {-12.93/18 27} Rxd1 {+M9/24 14} 67. Kxd1 {-19.48/15 4
Arena Adjudication} 0-1

[/pgn]

CL

[Rebel]

In games with repeating time controls, such as 40 moves in 20 minutes, Rebel TNG will start moving almost instantly (spending exactly 1 sec per move) after move 40, move 80, etc. for an indefinite number of moves.

This seriously cripples its performance, as one would expect.

Never happened to me, odd.

[snino64]

It's very interesting, thank you very much!
I try to use it in scidvspc and I find the error message:The analysis engine terminated without exit code: ""
Can be used with scid?

[Rebel]

It's very interesting, thank you very much!
I try to use it in scidvspc and I find the error message:The analysis engine terminated without exit code: ""
Can be used with scid?

I have no idea, the engine runs fine under Arena and ChessPartner and ChessBase although the latter I haven't used for a while.

[mike angel]

Maestro Ed ... I love you !!!

... today, yesterday and ever -- also I like your webside

[carldaman]

In games with repeating time controls, such as 40 moves in 20 minutes, Rebel TNG will start moving almost instantly (spending exactly 1 sec per move) after move 40, move 80, etc. for an indefinite number of moves.

This seriously cripples its performance, as one would expect.

Never happened to me, odd.

Very baffling, indeed.
I'm not surprised though, since if I don't find bugs, they find me, LOL

I'm curious to see if any of the regular rating list testers can reproduce the bug.

Increment time controls fortunately are not affected. Rebel TNG really shows its true strength there It's a magnificent beast!