Not logged inRybka Chess Community Forum
Up Topic The Rybka Lounge / Computer Chess / TCEC 19 Premier and CCC CPU Blitz
- By MrKris (***) Date 2020-09-09 06:19
As expected in the TCEC 19 Premier Stockfish is leading.
Lc0's J92-100 is doing well, close second so far.
https://tcec-chess.com/#

At CCC's CPU Blitz Madness Stockfish is of course first, Sf-Classic of course is 2nd but maybe surprisingly close so far
https://www.chess.com/computer-chess-championship

Game at: https://tcec-chess.com/#div=p&game=38&season=19 [Event "TCEC 19 Premier"]
[Site "tcec-chess.com"]
[Date "2020.09.08"]
[Round "10.2"]
[White "Stockfish 208026"]
[Black "AllieStein0.8-LS15.0"]
[Result "1-0"]
[BlackElo "2000"]
[Time "21:40:24"]
[WhiteElo "2400"]
[TimeControl "1800+5"]
[Termination "normal"]
[PlyCount "0"]
[WhiteType "human"]
[BlackType "program"]

1. Nf3 f5 2. g3 Nf6 3. Bg2 g6 4. d4 Bg7 5. O-O O-O 6. b3 d6 7. Bb2 Qe8
8. c4 Na6 9. Re1 b6 10. Nbd2 Bb7 11. a3 e6 12. b4 Qe7 13. Qb3 c5
14. b5 Nc7 15. a4 Rab8 16. Qa3 Ba8 17. e3 a6 18. Rac1 Rfe8 19. Bc3 axb5
20. cxb5 Ncd5 21. dxc5 bxc5 22. Bb2 e5 23. Red1 e4 24. Ne1 Nb4 25. Bf1 g5
26. a5 f4 27. Nc4 d5 28. Nb6 Ng4 29. Bxg7 Qxg7 30. Rxc5 Qh6 31. h3 Nxf2
32. Rdxd5 Nxd5 33. Nxd5 fxg3 34. Qb3 {TCEC 0.22}

[FEN "br2r1k1/7p/7q/PPRN2p1/4p3/1Q2P1pP/5n2/4NBK1 b - - 1 34"]

34... Rf8 {??} ({TCEC Stockfish's PV for its 0.22:}34...Kh8 35. Qb2+ Qg7 36. Qxg7+ Kxg7 37. Nc7 Rec8 38. b6 h5 39. Rxg5+ Kh6 40. Rc5 Rd8
41. Be2 Nxh3+ 42. Kg2 Ng5 43. Kxg3 Bb7 44. Rxg5 Kxg5 45. Ne6+ Kf5 46. Nxd8 Rxd8
47. a6 Rb8 48. axb7 Rxb7 49. Bxh5 Rxb6 )
35. Nb6+ {TCEC 2.23}  Kg7 36. Qb2+ {TCEC 3.74} Qf6
37. Qxf6+ {TCEC 4.44} Rxf6 38. Nd7 Rd8 39. Nxf6 Kxf6 40. Nc2 Rd2 41. a6 h5 42. Rc8 Bd5
43. Nb4 Bc4 44. Rc6+ Kf7 45. Bxc4+ Ke7 46. Bf1 g4 47. hxg4 h4 48. Rh6 1-0

Current TCEC GPU server - for AllieStein G/90'+10"
GPUs: 4x V100 {Amazon $8,250 each, x4 = $33,000}

Here, Toga found the correct black 34th move in 1 second:
TogaII 4.0.1 64 Bit Core-i7: move/30", Ryzen 7 2700X {Amazon $223}
...
  8/24-  00:01           91k  935k  -0.08  34. ... Rf8 35.Nf4+ Kh8 36.Qc3+ Kg8
  8/24  00:01           91k  943k  +0.20  34. ... Kf8 35.Nc7 Nxh3+ 36.Bxh3 Qxh3...
...
20/56  00:24       19,625k  11,808k   -0.00  34. ... Kh8 35.Qc3+ Qg7 36.Qxg7+ Kxg7...
- By MrKris (***) Date 2020-09-13 19:14
https://tcec-chess.com/
Season 19 - Premier Division 90|10 (8 engines)

Stockfish is leading but by only a 1/2 point above Lc0: 18.0/29 vs. 17.5/29.


Stockfish-aug26 1-0 Komodo-2576, Rd. 12.3
After 66.Qe3+ {TECE 16.64} Kf7 {TCEC 4.79 (?)}
8/5k2/3q4/1p6/1Pb5/4QBP1/5PK1/8 w - - 2 67

Then 67.Qf4+ {TCEC 21.60} Qxf4 {TCEC 11.05 (ok)}
(-Maybe human-like by Komodo: low disadvantage with a queen and a remote possibility of its oponent allowing a perpetual.
-Several years ago at TCEC it got out of a lost game vs. Stockfish that way - there was an Sf patch in a few days.)


So far 6th-8th is Komodo/Ethereal/Fire all with 13.0/30.

- - -

https://www.chess.com/computer-chess-championship
CPU Blitz Madness 3|2 (14 engines)

Quite a decent showing by Classic so far I thought:

1 Stockfish 102.5/132 (77.7%!!)
2 Sf-Classic 99.0/132 (75.0%!)
3 Komodo 81.5/130 (62.7%)
4...

Considering this, although 11 is old now:
http://www.cegt.net/40_40%20Rating%20List/40_40%20All%20Versions/rangliste.html
1 Stockfish 12 82215 1 CPU 3546
2 (Sfnnue)
3 (Sfnnue)
4 Lc0 Ls15 3514
5 Stockfish 11 8 CPU 35414
6...
- By MrKris (***) Date 2020-09-14 09:36
https://tcec-chess.com/
TCEC 19 Premier, after 31 games each:
1-2 Stockfish 19.0
1-2 Lc0 19.0
3 AllieStein 16.5
4 Stoofvlees 15.5
5-8 Komodo 13.5
5-8 Ethereal 13.5
5-8 ScorpioNN 13.5
5-8 Fire 13.5
Upcoming: Stockfish-Stoofvless, Lc0-AllieStein.

https://tcec-chess.com/#div=p&game=121&season=19
Lc0_J92-100 1-0 ScorpioNN

After 18...d4 {TCEC 0.12} 19.c5 {TCEC 0.41. --Is there really a 0.41 += or +/ defence for black, white looks winning, +-, to me holing in the R(a8) and B(b8) with just its N(a4).}
rb1q1rk1/1p1nnpp1/1Pp1b2p/p1P1p3/N2pP3/P2P1NPP/1BQ2PB1/1R3RK1 b - - 0 19


After 38...Bb3. Then 39.Ra1 {TCEC 3.75. --As if knowing black has nothing better to do than chase it?!, as if thinking black's N could defend more than its rook could attack?!} Nc2 {TCEC 3.46} 39.g4 {TCEC 4.08}
rb3r1k/1p3qp1/1Pp4p/p1P1p2P/Pn1pN3/1b1P1PP1/3BQNK1/4RR2 w - - 1 39


After 53...Qf5, black had given up material to escape the bind, but now is B(b5) is trapped and useless.
White's king's motto is: 'the best defense is a strong offense'.
Then 54.Qxg8+ Mate in 13:
6rk/1pP5/2p4B/1bP1pq1P/3pN2K/p2P1P2/Q7/3R4 w - - 3 54


- - -

https://www.chess.com/computer-chess-championship
CPU Blitz Madness 3|2
1 Stockfish 109/142 (76.8%)
2 Stockfish Classic 107/142 (75.4%)
3 Komodo 88.5/142 (62.3%)
4...
                         go Classic!
- By MrKris (***) Date 2020-09-15 04:57
https://tcec-chess.com/#div=p&game=131&season=19
Komodo 0-1 Stockfish, TCEC Premier, Rd. 33.3, G/90'+15"

After 42...Rf7 {TCEC 0.00} 43.Rb2-a2? h5 {TCEC -1.44}
Stockfish's TCEC PV said 43.Rb2-b1 was 0.00.

After Stockfish's recommendation 43.Rb2-a1 (1.Rb2-b1 here) Classical was able to draw as white vs Pure NNUE: [Event "G/5'+1''"]
[Site "AMD R7 2700X"]
[Date "2020.09.14"]
[Round ""]
[White "Cfish-200911Classical"]
[Black "Cfish-200911Pure"]
[Result "1/2-1/2"]
[FEN "r1bb1qk1/3n1r2/2Rp3p/1QpPp1p1/p1P1PpP1/P1BN1P2/6KP/1R1B4 b - - 13 1"]
[PlyCount "70"]
[SetUp "1"]
[Termination "adjudication"]
[TimeControl "300+2"]

1... Nf6 {+0.34/30 15s} 2. Bxe5 {0.00/34 6.3s} dxe5 {+0.67/26 2.2s}
3. Nxe5 {0.00/38 4.0s} Rfa7 {+0.65/28 4.1s} 4. Qxc5 {-0.12/36 6.3s}
Qxc5 {+0.33/33 21s} 5. Rxc5 {-0.16/38 7.3s} Bd7 {0.00/36 9.4s}
6. Nc6 {-0.25/32 6.0s} Bxc6 {0.00/36 3.7s} 7. Rxc6 {-0.16/32 3.7s}
Nd7 {0.00/36 3.7s} 8. Rg6+ {0.00/35 3.5s} Kf7 {0.00/33 4.0s}
9. Rxh6 {0.00/39 5.1s} Kg7 {0.00/34 3.9s} 10. Re6 {0.00/37 4.5s}
Nc5 {0.00/37 8.7s} 11. Re5 {0.00/37 5.4s} Rb7 {+0.33/30 4.2s}
12. Rb5 {0.00/40 4.1s} Rxb5 {+0.20/32 4.6s} 13. cxb5 {0.00/39 4.9s}
Rc8 {0.00/37 13s} 14. Re8 {0.00/41 6.8s} Nd7 {0.00/38 5.0s}
15. h4 {0.00/36 4.7s} Kf7 {0.00/39 4.3s} 16. Rh8 {0.00/40 5.0s}
Nf8 {0.00/37 5.0s} 17. Rh6 {0.00/35 5.0s} gxh4 {0.00/39 4.0s}
18. Rd6 {0.00/37 6.2s} Ng6 {0.00/39 4.9s} 19. Bxa4 {0.00/34 4.5s}
Ne5 {0.00/40 5.4s} 20. b6 {0.00/39 8.7s} Rb8 {0.00/41 7.5s}
21. Re6 {0.00/40 5.0s} h3+ {0.00/42 4.3s} 22. Kxh3 {0.00/47 4.6s}
Nxf3 {0.00/47 6.0s} 23. Re8 {0.00/43 4.5s} Ng5+ {0.00/45 5.1s}
24. Kg2 {0.00/47 16s} Rxb6 {0.00/48 6.3s} 25. Rxd8 {0.00/49 9.8s}
Rb2+ {0.00/45 7.1s} 26. Kf1 {0.00/47 6.7s} f3 {0.00/49 15s}
27. Rd7+ {0.00/46 5.1s} Kg8 {0.00/50 6.4s} 28. Rd8+ {0.00/42 7.0s}
Kg7 {0.00/47 4.1s} 29. Rd7+ {0.00/46 8.7s} Kf8 {0.00/48 5.0s}
30. Rd8+ {0.00/54 24s} Kf7 {0.00/49 12s} 31. Rd7+ {0.00/49 6.2s}
Kf8 {0.00/47 5.3s} 32. Rd8+ {0.00/57 4.1s} Kg7 {0.00/55 7.3s}
33. Rd7+ {0.00/51 4.1s} Kg6 {0.00/47 4.4s} 34. Rc7 {0.00/48 6.1s}
Nh3 {0.00/50 4.3s} 35. Bc2 {0.00/47 7.0s} Ra2 {0.00/48 4.7s}
36. Ke1 {0.00/50 8.1s, Draw by adjudication} 1/2-1/2

With Komodo's move 43.Rb2-a2?, 1.Rb2-a2 here, Classical lost quickly: [Event "G/6'+1''"]
[Site "AMD R7 2700X"]
[Date "2020.09.14"]
[Round ""]
[White "Cfish-200911Classical"]
[Black "Cfish-200911Pure"]
[Result "0-1"]
[FEN "r1bb1qk1/3n1r2/2Rp3p/1QpPp1p1/p1P1PpP1/P1BN1P2/R5KP/3B4 b - - 13 1"]
[PlyCount "60"]
[SetUp "1"]
[Termination "adjudication"]
[TimeControl "360+1"]

1... h5 {+1.28/28 26s} 2. gxh5 {-0.55/27 7.1s} Nf6 {+1.41/26 2.5s}
3. Rb2 {-0.62/31 29s} g4 {+1.27/27 4.0s} 4. Kh1 {-1.24/28 11s}
g3 {+1.69/26 3.2s} 5. Rxd6 {-1.31/30 12s} Qxd6 {+3.11/25 3.7s}
6. Bxe5 {-1.36/27 2.7s} Qa6 {+3.39/26 6.8s} 7. Bxf4 {-1.53/33 26s}
Qxb5 {+3.98/26 3.5s} 8. Rxb5 {-1.57/29 2.3s} Nxh5 {+4.47/26 4.7s}
9. Be5 {-1.88/29 7.7s} Rh7 {+4.91/26 3.8s} 10. Rb2 {-1.90/31 7.0s}
Bf6 {+5.28/27 3.8s} 11. Bxf6 {-2.09/29 11s} Nxf6 {+5.25/26 7.0s}
12. Rg2 {-2.14/29 13s} Nd7 {+6.19/26 4.2s} 13. Rxg3+ {-2.99/28 26s}
Rg7 {+6.69/28 5.4s} 14. Kg2 {-3.37/27 17s} Ba6 {+8.07/27 6.3s}
15. Kf2 {-3.48/28 9.8s} Bxc4 {+8.42/27 4.4s} 16. Nb2 {-4.04/27 23s}
Bb3 {+10.92/27 5.2s} 17. Be2 {-4.09/25 2.8s} Rxg3 {+11.43/29 4.7s}
18. hxg3 {-4.33/26 11s} Rb8 {+11.77/27 4.5s} 19. g4 {-4.28/25 13s}
Kf8 {+12.49/31 7.7s} 20. g5 {-4.51/26 19s} Bxd5 {+14.18/26 4.2s}
21. Nxa4 {-5.30/25 5.3s} Bc6 {+14.75/27 6.2s} 22. Nc3 {-6.02/27 15s}
Rb3 {+15.03/28 5.8s} 23. Nd1 {-6.12/24 4.5s} Ne5 {+15.20/27 5.4s}
24. a4 {-6.69/27 20s} Bxa4 {+16.07/26 4.9s} 25. Ne3 {-6.69/25 1.6s}
Rb2 {+16.21/25 5.2s} 26. Ke1 {-7.70/27 9.3s} Bb5 {+19.97/28 11s}
27. Bd1 {-152.50/42 4.0s} Nd3+ {+152.50/31 2.9s} 28. Kf1 {-3.40/9 0.001s}
Nf4+ {+152.55/46 5.0s} 29. Ke1 {-152.49/245 28s} Bd3 {+152.58/41 4.2s}
30. Ba4 {-152.50/27 2.5s} c4 {+152.59/51 4.4s}
31. Nd5 {-152.58/31 2.3s, Black wins by adjudication} 0-1

The point of the difference: [Event "Round 33.3"]
[Site "TCEC Premier"]
[Date "2020.09.14"]
[Round "?"]
[White "Komodo"]
[Black "Stockfish"]
[Result "0-1"]
[BlackElo "2400"]
[Time "20:42:55"]
[WhiteElo "2400"]
[TimeControl "0+60"]
[SetUp "1"]
[FEN "r1bb2k1/3n3r/8/2RPN1pP/p1P1Pp2/P4P2/R5KP/3B4 w - - 2 49"]
[Termination "normal"]
[PlyCount "4"]
[WhiteType "program"]
[BlackType "program"]

{The TCEC game:} 49. Rxc8 ({If} 49. Nxd7 Bxd7 {Now black is threatening 50...Bb6 trapping white's
R(c5). Had Komodo played 43.Rb2-b1, instead of the pointless, off an
important file, 43.Rb2-a2?, 49.Nxd7 would have been possible because 43.Rb2-b1 would already have defended against 50...Bb6.} 50.
Rb2 Bf6 51. Rb7 (51. Rd2 {?} Rxh5 52. d6 Kf7 53. Rc7 Ke6 54. c5 Rah8 {is
worse}) 51. .. Bh3+ 52. Kxh3 Rxb7 {Similar to the TCEC game, white pawns
vs. R.}) 49. .. Rxc8 50. Nxd7 Rxd7 {(TCEC -2.50, black won on move 83.)}
0-1
Up Topic The Rybka Lounge / Computer Chess / TCEC 19 Premier and CCC CPU Blitz

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill