Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Should there be a Centaur games thread?
1 2 Previous Next  
- - By Highendman (****) Date 2009-07-24 23:59
So at least I don't end up spamming the forum...

I wanted to share two interesting centaur games from today.
The first one shows just how wrong a chess engine can be in evaluating end-games... even in the hands of a very skilled Centaur.
Ultra-D was white, and played an excellent squeeze. I aimed for what has worked for me few times against the Cluster - giving up some material to stop an attack and get to a material down end-game that I believe can be drawn. It seems Rybka missevaluates rook end-games the most, imho. Ultra-D was quite sure his position was winning. After move 40 he sent me a chat message saying his 40. Kf2 was a winning move - while in fact I believe it's the very move that allowed forcing draw immediately.

position after 40. Kf2
3R4/p3r1k1/5pp1/P1nB3p/4PP1P/6P1/5K2/8 b - - 0 40


And after move 43. Rxa7 he asked again if I wanted to continue or move to the next game to try and take revenge. The reason was clear: the chess engine was jumping with joy scoring this position at >3.00. While it's a known theoretical position which is a dead draw (rook in front of a passed pawn).

position after 43. Rxa7

8/R7/5ppk/P6p/4rP1P/6P1/5K2/8 b - - 0 43


And here is the game:

[Event "Friendly Game, 75m + 15s"]
[Site "Engine Room"]
[Date "2009.07.24"]
[Round "?"]
[White "Ultra-d"]
[Black "Highendman"]
[Result "1/2-1/2"]
[ECO "D11"]
[WhiteElo "1653"]
[BlackElo "2454"]
[PlyCount "110"]
[EventDate "2009.07.24"]
[TimeControl "4500+15"]

1. d4 {5} d5 {2} 2. c4 {9} c6 {42} 3. Nf3 {32} Nf6 {31} 4. g3 {25} dxc4 {339}
5. Bg2 {93} g6 {26} 6. a4 {44} Bg7 {149} 7. Na3 {67} Qa5+ {303} 8. Bd2 {42} Qa6
{6} 9. Rc1 {10} c3 {152} 10. bxc3 {168} Nbd7 {72} 11. c4 {244} O-O {400} 12.
O-O {47} Ne4 {43} 13. Bb4 {69} c5 {195} 14. dxc5 {15} Ndxc5 {24} 15. Nb5 {32}
Be6 {277} 16. Nfd4 {239} Rac8 {20} 17. Qc2 {40} Bxd4 {10} 18. Nxd4 {7} Nd6 {2}
19. Nxe6 {5} Nxe6 {3} 20. Bxd6 {23} exd6 {0} 21. Rfd1 {62} Rc5 {17} 22. Qb3 {
246} Rb8 {14} 23. h4 {23} Re5 {186} 24. e3 {196} Nc5 {8} 25. Qb2 {370} Rd8 {245
} 26. a5 {492} Rd7 {11} 27. Qb4 {75} Kg7 {623} 28. Rd4 {27} f6 {236} 29. Rb1 {
35} h6 {82} 30. Qd2 {211} Rde7 {426} 31. Rxd6 {319} Qxc4 {35} 32. Bd5 {11} Qd3
{8} 33. Qxd3 {7} Nxd3 {0} 34. Rxb7 {6} Rxb7 {4} 35. Bxb7 {5} Nc5 {3} 36. Bd5 {7
} Re7 {3} 37. e4 {30} Rc7 {9} 38. Rd8 {129} h5 {559} 39. f4 {23} Re7 {28} 40.
Kf2 {83} Nxe4+ {168} 41. Bxe4 {76} Rxe4 {3} 42. Rd7+ {12} Kh6 {2} 43. Rxa7 {82}
Ra4 {4} 44. a6 {27} g5 {77} 45. fxg5+ {165} fxg5 {9} 46. Ra8 {103} gxh4 {4} 47.
gxh4 {65} Kg7 {2} 48. a7 {43} Ra6 {42} 49. Ke3 {102} Ra2 {62} 50. Ke4 {29} Ra4+
{10} 51. Kd3 {15} Ra1 {20} 52. Kd4 {27} Ra3 {12} 53. Kc5 {35} Ra1 {10} 54. Kb6
{42} Rb1+ {9} 55. Kc7 {3} Ra1 {
Ultra-d offers a draw  (Lag: Av=0.54s, max=2.6s) 13} 1/2-1/2

The second game Sebi kindly accepted my invite for a game.
You may find it interesting the white Knight moves 6 times in the first 18 moves. The position just allowed to slowly manoeuvre the pieces.
It was a game where black had a significant weakness on his d file. It was kind of a book-like strategy, in a position that allowed ignoring the engine and playing to a simple plan - take control of d5 -> rooks to centre, king out of the way to connect them, aim knight at weak spot etc. It ended abruptly after 25.. g6, that I believe was a blunder that looses the game.

End position:
1rrb2k1/3q1pn1/3p4/1p2p2p/1P2PPP1/pBP1N3/P2Q3R/5KR1 b - - 0 35


And here's the game

[Event "Friendly Game, 60m + 30s"]
[Site "Engine Room"]
[Date "2009.07.24"]
[Round "?"]
[White "Highendman"]
[Black "Sebi-chess"]
[Result "1-0"]
[ECO "B90"]
[WhiteElo "2454"]
[BlackElo "1695"]
[PlyCount "69"]
[EventDate "2009.07.24"]
[TimeControl "3600+30"]

1. e4 {1} c5 {31} 2. Nf3 {2} d6 {68} 3. d4 {6} cxd4 {2} 4. Nxd4 {5} Nf6 {7} 5.
Nc3 {2} a6 {8} 6. f3 {27} e5 {3} 7. Nb3 {13} Be6 {29} 8. Bg5 {43} h6 {11} 9.
Bxf6 {27} Qxf6 {7} 10. Nd5 {28} Qd8 {2} 11. Qd2 {35} Bxd5 {23} 12. Qxd5 {122}
Nc6 {4} 13. Bc4 {37} Qc7 {91} 14. c3 {7} Rb8 {24} 15. Nd2 {78} b5 {180} 16. Bb3
{22} Nd8 {7} 17. Nf1 {57} Be7 {299} 18. Ne3 {11} O-O {260} 19. h4 {155} Ne6 {
220} 20. Qd2 {6} a5 {129} 21. Rd1 {258} Rfd8 {194} 22. Kf2 {199} a4 {167} 23.
Bd5 {47} a3 {329} 24. b4 {21} Rdc8 {301} 25. Rc1 {15} g6 {280} 26. h5 {154}
gxh5 {14} 27. Rxh5 {18} Bg5 {8} 28. g3 {7} Qe7 {308} 29. Rg1 {58} Ng7 {82} 30.
Rh3 {42} h5 {70} 31. f4 {41} Bf6 {9} 32. Bb3 {42} Qa7 {31} 33. Kf1 {24} Qd7 {76
} 34. Rh2 {2} Bd8 {26} 35. g4 {
Sebi-chess gibt auf  (Lag: Av=0.32s, max=0.9s) 35} 1-0
Parent - - By Dadi Jonsson (Silver) [is] Date 2009-07-25 00:14

> And after move 43. Rxa7 he asked again if I wanted to continue or move to the next game


I guess that most experienced tournament players have encountered this type of position (or studied this endgame) and know what the result will be. This game is an example of the advantage that a player with a solid chess knowledge has in centaur games (although many other things can influence the outcome of a game).
Parent - By Miroslav Kvíčala (**) [cz] Date 2009-07-26 04:39
Exactly. If somebody is not able to correctly evaluate so easy endgame...then it is difficult...someone who plays chess at solid level must know very well that it is clear draw :-)
Parent - By TGO (**) [si] Date 2009-07-25 15:20
a thread like this would be intresting, we could all learn something from good centaur observations :)
Parent - - By Lukas Cimiotti (Bronze) [de] Date 2009-07-25 16:24
Thanks for posting these games :)
Do you know if Sebi used HIARCS or Rybka? He's a fanatic HIARCS team member - when Eros lately asked him if he would play the cluster, he answered he wouldn't try it with a dual core - and his quad had more important things to do.

Regards,
Lukas
Parent - By Highendman (****) Date 2009-07-25 17:23
Sorry - don't know. I'll ask him what was he using next time I see him online, and post it here.
I assumed he was using Rybka, but would be very interesting indeed to find out if it was another engine.
Parent - - By David Evans (****) [gb] Date 2009-07-25 16:40
I would like to thank Highend for the series of games i did say during the game kf2 is winning but from a human point of view if it had been otb chess the postion would have been a win for white thas why centaur chess is exciting u always have chances in the face of defeat always battle on.

I hope to play some more games as Highend seems to me a strong centaur players as he has proved playing the cluster.

I look forward to more games and hope to get a little closer to victory we shall see.

Ultra-d
Parent - By Uri Blass (*****) [il] Date 2009-07-25 17:27
I do not understand it.

"from a human point of view if it had been otb chess the postion would have been a win for white"

I think that the position could be also a draw in human-human game OTB so
I do not understand claiming that the position is a win for white from human point of view.

The position may be a win for white only from rybka's point of view(and from the point of view of some other chess engines).

Uri
Parent - - By Uri Blass (*****) [il] Date 2009-07-25 16:41
I do not know Ultra-D but based on the game he does not seem to be a very skilled centaur.

I know for a long time that rybka cannot be trusted about high scores in the endgame and I am surprised that Ultra-D does not have this knowledge.

I have examples from correspondence games

This position happened in analysis of one of my correspondence games(I used rybka2.2n2 at that time because rybka3 was not available but rybka3 is not much better(it also suggest the losing move Kc3 instead of the drawing move Kc4 with a decisive score for white)

Maybe tablebases can help but I think that a good program should see without tablebases.

New game
[D]
8/8/p7/P7/8/1n1K2k1/7p/7B w - - 0 1


Analysis by Rybka 3 1-cpu 32-bit :

1.Kd3-c3
  -+  (-2.90)   Depth: 2   00:00:00
1.Kd3-c3
  -+  (-3.22)   Depth: 3   00:00:00
1.Kd3-c3
  -+  (-3.02)   Depth: 4   00:00:00
1.Kd3-c3
  -+  (-2.82)   Depth: 4   00:00:00
1.Kd3-c3
  -+  (-2.83)   Depth: 4   00:00:00
1.Kd3-c3 Nb3-c5
  -+  (-2.76)   Depth: 5   00:00:00
1.Kd3-c3 Nb3-c5 2.Kc3-d4
  -+  (-2.93)   Depth: 6   00:00:00
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e4 Nd7-f6+
  -+  (-3.05)   Depth: 7   00:00:00  1kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e4 Nd7-f6+ 4.Ke4-e3
  -+  (-3.24)   Depth: 8   00:00:00  1kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e4 Nd7-f6+ 4.Ke4-e3 Nf6-g4+ 5.Ke3-d4 Kg3-f2
  -+  (-3.25)   Depth: 9   00:00:00  2kN
1.Kd3-c4
  -+  (-3.04)   Depth: 9   00:00:00  3kN
1.Kd3-c4 Nb3-c1 2.Kc4-c5
  -+  (-3.15)   Depth: 9   00:00:00  3kN
1.Kd3-c4 Nb3-c1 2.Kc4-d4 Kg3-f2 3.Bh1-a8 Kf2-g1 4.Ba8-f3 Nc1-b3+ 5.Kd4-c3
  -+  (-3.25)   Depth: 10   00:00:01  6kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4
  -+  (-3.13)   Depth: 10   00:00:01  7kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e4
  -+  (-3.06)   Depth: 11   00:00:01  8kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e3 Nd7-e5 4.Bh1-d5 Ne5-c4+ 5.Ke3-d3 Nc4-b2+ 6.Kd3-e3
  -+  (-2.99)   Depth: 12   00:00:01  14kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e3 Nd7-e5 4.Bh1-d5 Ne5-c4+ 5.Ke3-d3 Nc4-b2+ 6.Kd3-e3
  -+  (-2.99)   Depth: 13   00:00:01  15kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-e6+ 3.Kd4-e4 Kg3-f2 4.Ke4-e5 Ne6-c5 5.Ke5-d4
  -+  (-3.38)   Depth: 14   00:00:01  44kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-e6+ 3.Kd4-e4 Kg3-f2 4.Ke4-e5 Ne6-c5 5.Ke5-d4 Nc5-b3+ 6.Kd4-c3 Nb3-c1
  -+  (-3.38)   Depth: 15   00:00:01  49kN
1.Kd3-c3 Nb3-c5
  -+  (-5.12)   Depth: 16   00:00:12  820kN

(so k,  25.07.2009)

No score and no move after 9 minutes at depth 17 and no tablebases is no excuse.

For comparison free programs are able to find the drawing move with draw score easily

New game
8/8/p7/P7/8/1n1K2k1/7p/7B w - - 0 1


Analysis by Stockfish 1.4 JA:

1.Kd3-e3 Nb3xa5
  -+  (-3.73)   Depth: 2   00:00:00
1.Kd3-e3 Nb3xa5 2.Ke3-e2
  -+  (-4.00)   Depth: 3   00:00:00
1.Kd3-e2 Nb3xa5 2.Ke2-f1
  -+  (-3.81)   Depth: 3   00:00:00
1.Kd3-e2 Nb3xa5 2.Ke2-f1 Na5-c4
  -+  (-4.18)   Depth: 4   00:00:00
1.Kd3-c4 Nb3-d2+ 2.Kc4-d3 Nd2-f3 3.Kd3-e4
  -+  (-3.21)   Depth: 4   00:00:00
1.Kd3-c4 Nb3-c1 2.Kc4-d4 Kg3-f4 3.Kd4-d5
  -+  (-3.19)   Depth: 5   00:00:00
1.Kd3-c4 Nb3-c1 2.Kc4-d4 Kg3-f4 3.Kd4-d5 Nc1-d3
  -+  (-3.57)   Depth: 6   00:00:00  3kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e4 Nd7-f6+ 4.Ke4-e3 Nf6-g4+ 5.Ke3-e2
  -+  (-3.03)   Depth: 6   00:00:00  3kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-e6+ 3.Kd4-e3 Ne6-g7 4.Ke3-e4 Ng7-e6
  -+  (-3.11)   Depth: 7   00:00:00  4kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-e6+ 3.Kd4-e3 Ne6-g7 4.Ke3-e4 Kg3-f2 5.Ke4-e5
  -+  (-3.19)   Depth: 8   00:00:00  6kN
1.Kd3-c3 Nb3-c5
  -+  (-3.43)   Depth: 9   00:00:01  8kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e3 Nd7-e5 4.Ke3-d4 Kg3-f4 5.Kd4-d5 Ne5-f3
  -+  (-3.59)   Depth: 10   00:00:01  12kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e3 Nd7-e5 4.Ke3-d4 Kg3-f4 5.Kd4-c5 Kf4-e3 6.Kc5-d5
  -+  (-3.89)   Depth: 11   00:00:01  18kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e4 Kg3-f2 4.Ke4-f4 Kf2-g1 5.Bh1-f3 Nd7-f6 6.Bf3-c6 Nf6-d7
  -+  (-3.93)   Depth: 12   00:00:01  33kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e4 Kg3-f2 4.Ke4-f4 Kf2-g1 5.Bh1-f3 Nd7-f6 6.Bf3-c6 Nf6-h5+ 7.Kf4-g4 Nh5-g7
  -+  (-3.83)   Depth: 13   00:00:01  55kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e3 Nd7-e5 4.Bh1-d5 Ne5-c4+ 5.Ke3-d4 Nc4xa5 6.Kd4-c3 Kg3-f2 7.Kc3-b4 Kf2-g1 8.Kb4xa5
  -+  (-3.97)   Depth: 14   00:00:01  80kN
1.Kd3-c3
  -+  (-3.59)   Depth: 15   00:00:01  136kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e3 Nd7-e5 4.Bh1-d5 Ne5-c4+ 5.Ke3-d4 Nc4-d6 6.Kd4-e5 Nd6-f7+ 7.Ke5-f6 Nf7-h6 8.Kf6-e5 Nh6-g4+ 9.Ke5-d4 Kg3-h3 10.Kd4-d3
  -+  (-3.65)   Depth: 16   00:00:01  172kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e3 Nd7-e5 4.Bh1-d5 Ne5-c4+ 5.Ke3-d4 Nc4-d6 6.Kd4-e5 Nd6-f7+ 7.Ke5-f6 Nf7-h6 8.Kf6-e5 Nh6-g4+ 9.Ke5-d4 Kg3-h3 10.Kd4-d3 Kh3-g3
  -+  (-3.83)   Depth: 17   00:00:01  243kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-d7 3.Kd4-e3 Nd7-e5 4.Bh1-d5 Ne5-c4+ 5.Ke3-d4 Nc4-d6 6.Kd4-e3 Nd6-f5+ 7.Ke3-e2 Kg3-f4 8.Ke2-d3 Nf5-g3 9.Bd5-c6 Ng3-h5 10.Bc6-d5 Nh5-f6
  -+  (-3.79)   Depth: 18   00:00:01  345kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-e6+ 3.Kd4-e3 Ne6-g7 4.Bh1-d5 Ng7-f5+ 5.Ke3-e2 Kg3-f4 6.Ke2-d3 Kf4-g4 7.Bd5-a8 Nf5-g3 8.Kd3-d4 Kg4-f4 9.Ba8-d5 h2-h1Q 10.Bd5xh1 Ng3xh1 11.Kd4-c4 Nh1-g3
  -+  (-3.87)   Depth: 19   00:00:01  683kN
1.Kd3-c3 Nb3-c5
  -+  (-4.08)   Depth: 20   00:00:02  1004kN
1.Kd3-c3 Nb3-c5
  -+  (-4.96)   Depth: 21   00:00:02  1736kN
1.Kd3-c3 Nb3-c5 2.Kc3-d4 Nc5-e6+ 3.Kd4-e4 Kg3-f2 4.Ke4-e5 Kf2-g1 5.Bh1-c6 Ne6-c5 6.Ke5-d4 Nc5-b3+ 7.Kd4-c3 Nb3xa5 8.Bc6-d5 h2-h1Q 9.Bd5xh1 Kg1xh1 10.Kc3-d4 Na5-c6+ 11.Kd4-c3 Kh1-g1 12.Kc3-b3 a6-a5 13.Kb3-a4 Kg1-f2 14.Ka4-b5
  -+  (-5.65)   Depth: 22   00:00:03  3099kN
1.Kd3-c4 Nb3-c1 2.Kc4-c5 Kg3-f2 3.Kc5-b6 Kf2-g1 4.Bh1-a8 Nc1-d3 5.Kb6xa6 Nd3-f4 6.Ka6-b6 Nf4-g2 7.Ba8-c6 h2-h1Q 8.a5-a6 Qh1-h6 9.a6-a7 Qh6-e3+ 10.Kb6-b7 Qe3-b3+ 11.Kb7-c7 Qb3-g3+ 12.Kc7-b7 Qg3-g7+ 13.Kb7-b6 Qg7-b2+ 14.Kb6-c7 Qb2-e5+ 15.Kc7-b6 Qe5-d4+ 16.Kb6-b7
  -+  (-5.53)   Depth: 22   00:00:03  4587kN
1.Kd3-c4 Nb3-c1 2.Kc4-c5 Kg3-f2 3.Kc5-b6 Kf2-g1 4.Bh1-a8 Nc1-d3 5.Kb6xa6 Nd3-f4 6.Ka6-b6 Nf4-g2 7.Ba8-c6 h2-h1Q 8.a5-a6 Qh1-h6 9.a6-a7 Qh6-e3+ 10.Kb6-b7 Qe3-b3+ 11.Kb7-c7 Qb3-g3+ 12.Kc7-b7 Qg3-g7+ 13.Kb7-b6 Qg7-b2+ 14.Kb6-c7 Qb2-e5+ 15.Kc7-b7 Qe5-e7+ 16.Kb7-a6 Qe7-a3+
  -+  (-5.53)   Depth: 23   00:00:04  5827kN
1.Kd3-c4 Nb3-c1 2.Kc4-c5 Kg3-f2 3.Kc5-b6 Kf2-g1 4.Bh1-a8 Nc1-d3 5.Kb6xa6 Nd3-f4 6.Ka6-b6 Nf4-g2 7.Ba8-c6 h2-h1Q 8.a5-a6 Qh1-h6 9.a6-a7 Qh6-e3+ 10.Kb6-b7 Qe3-b3+ 11.Kb7-c7 Qb3-g3+ 12.Kc7-b7 Qg3-g7+ 13.Kb7-b6 Qg7-b2+ 14.Kb6-c7 Qb2-e5+ 15.Kc7-b7 Qe5-e7+ 16.Kb7-a6 Qe7-e2+
  -+  (-5.47)   Depth: 24   00:00:06  8128kN
1.Kd3-c4
  -+  (-5.27)   Depth: 25   00:00:07  10193kN
1.Kd3-c4
  -+  (-4.40)   Depth: 26   00:00:10  13727kN
1.Kd3-c4
  µ  (-0.92)   Depth: 27   00:00:15  20159kN
1.Kd3-c4 Nb3-c1 2.Kc4-c5 Kg3-f2 3.Kc5-b6 Kf2-g1 4.Bh1-a8 Nc1-d3 5.Kb6xa6 Nd3-f4 6.Ka6-b6 Nf4-g2 7.Ba8-c6 h2-h1Q 8.a5-a6 Qh1-h6 9.a6-a7 Qh6-e3+ 10.Kb6-b7 Qe3-b3+ 11.Kb7-c7 Qb3-g3+ 12.Kc7-b7 Qg3-g7+ 13.Kb7-a6 Qg7-a1+ 14.Ka6-b7 Qa1-b2+ 15.Kb7-c8 Qb2-h8+ 16.Kc8-b7 Qh8-b2+
  =  (0.00)   Depth: 28   00:00:21  28708kN
1.Kd3-c4 Nb3-c1 2.Kc4-c5 Kg3-f2 3.Kc5-b6 Kf2-g1 4.Bh1-a8 Nc1-d3 5.Kb6xa6 Nd3-f4 6.Ka6-b6 Nf4-g2 7.Ba8-c6 h2-h1Q 8.a5-a6 Qh1-h6 9.a6-a7 Qh6-e3+ 10.Kb6-b7 Qe3-b3+ 11.Kb7-c7 Qb3-g3+ 12.Kc7-b7 Qg3-g7+ 13.Kb7-a6 Qg7-a1+ 14.Ka6-b7 Qa1-b2+ 15.Kb7-c8 Qb2-h8+ 16.Kc8-b7 Qh8-b2+
  =  (0.00)   Depth: 29   00:00:25  35008kN
1.Kd3-c4 Nb3-c1 2.Kc4-c5 Kg3-f2 3.Kc5-b6 Kf2-g1 4.Bh1-a8 Nc1-d3 5.Kb6xa6 Nd3-f4 6.Ka6-b6 Nf4-g2 7.Ba8-c6 h2-h1Q 8.a5-a6 Qh1-h6 9.a6-a7 Qh6-e3+ 10.Kb6-b7 Qe3-b3+ 11.Kb7-c7 Qb3-g3+ 12.Kc7-b7 Qg3-g7+ 13.Kb7-a6 Qg7-a1+ 14.Ka6-b7 Qa1-b2+ 15.Kb7-c8 Qb2-h8+ 16.Kc8-b7 Qh8-b2+
  =  (0.00)   Depth: 30   00:00:32  44657kN

(so k,  25.07.2009)

Uri
Parent - By TGO (**) [si] Date 2009-07-25 17:04
intresting>>but let us not turn this int another rybka bug thread :) ... although I agree that engines are horrible at endgames. (adding tablebases seems to helps though...at lest to some extent)
Parent - - By David Evans (****) [gb] Date 2009-07-25 17:08
I do not know Ultra-D but based on the game he does not seem to be a very skilled centaur.

I hate sweeping statements as of last year i was ranked 11th in the freestyle lists so i have a little exp nothing wrong with playing on in these types of postions with your opponent in time trouble maybe he makes wrongevaltion of the postion but to be fair highend was clear it was draw if he made correct moves.

Ask Highend if my skill as a centaur player is in question i think he would disagree with your statement.

Maybe Uri u would be willing to a test game with me if iam so unskilled.

Ultra-d
Parent - By Highendman (****) Date 2009-07-25 17:20
Ultra-D is the most difficult human Centaur I had the pleasure of playing against so far.
We played 6 draws so far - I didn't get close to a winning chance against him even once - and was in a very bad position in one of the games.
I look forward to continuing the match with him until one of us manages to get a decisive result more than once ;)
Parent - - By Uri Blass (*****) [il] Date 2009-07-25 17:25
You are probably favourite against me for the simple reason that I have not a good hardware.
The point is that I expect a good centaur not to claim that he is winning only because rybka claims +3.xx

The problem is not that you continued to play in drawn position but that I read that you made arrogant claims(like claiming that Kf2 is a winning move).

Uri
Parent - - By David Evans (****) [gb] Date 2009-07-25 17:42
i said kf2 would prob win in a otb game as the endgame would be very hard for black to hold.

My games with Highend have been played in good and fair spirits and nothing done in Arrogance.

As u stated yourself u dont know me  i never make sweeping statements about anybody as they tend to be unfounded.

Ultra-d
Parent - By Uri Blass (*****) [il] Date 2009-07-25 17:58
I will certainly try to win as white but also try to draw as black so from my human point of view the position after Kf2 is advantage for white(and not clear win or clear draw)

It is probably draw against correct moves of black based on the result of the game but without knowing the result of the game I am going to say unclear if draw or win for white.

I am sure that part of the human-human games from the relevant position can end in a win for white and part of the human-human games can end in a draw.
Getting 50% against highendman is an impressive result and my impression that you are not a good centaur was not because of the mistake (Kf2) in case that highendman is correct but because of the statements during the game.

Uri
Parent - By David Evans (****) [gb] Date 2009-07-25 17:45
Oh by the way i play with a quad i7 only as i sold my skulltrail and iam witing on a good duel board for the i7s.
Parent - - By Roland Rösler (****) [de] Date 2009-07-26 03:50 Edited 2009-07-26 03:54
When I see this position, I know Rybka is working and I could chuck up:

8/R7/5ppk/P6p/4rP1P/6P1/5K2/8 b - - 0 43


We have discussed it in this forum so many times. I can´t believe, a skilled centaur believes here in Rybkas eval. Rybkas eval increases when a-pawn moves forward, but we all know that winning factor decreases when a pawn is pushed forword. That most of the other engines are also blind isn´t an excuse for Rybka and at least for the centaur (Ultra-D).

If a top-15 centaur thinks this is a win, I´m very disaffected about freestyle! :-(

Play this position with R3 (for Alekhine (white)) and you will see why Alkehine would never be WCC with engine support! :-(

8/5pk1/5rpp/8/P2R3P/6P1/5PK1/8 w - - 0 51
Parent - - By Roland Rösler (****) [de] Date 2009-07-26 04:31
Maybe even better is this position (same game; Alekhine vs Capablanca) because of didactical issue:

8/5p2/r5p1/P3k2p/R6P/4K1P1/5P2/8 w - - 0 57
Parent - By M ANSARI (*****) [kw] Date 2009-07-26 10:34
I have to agree that this position is an obvious draw since all black has to do is avoid a rook check which would allow the pawn to queen even if the rook is to be sacrificed, and that seems easy to prevent.  But to be fair it is sometimes very hard to disregard off hand the evaluation of a 3200 ELO engine no matter how good of a chess player you are.  Another situation where Rybka 3 or many other engines seems to totally misevaluate is where the bishop does not control the queening square, in some cases the evaluation can be as high as 6.0+ even though it is an obvious draw that a human can see very quickly.  These situations will slowly disappear as engine hardware and software improve, already there have been huge strides in that field.
Parent - By Banned for Life (Gold) Date 2009-07-26 19:27
I could chuck up

Chuck up? Do you mean upchuck? Projectile vomiting (as in The Exorcist) would probably be more appropriate.

Actually, these are examples of positions that Rybka badly evaluates where people can easily see that the evaluation is far off the mark. There are a huge  number of more complex positions that Rybka badly evaluates where people can't easily see the truth. This is why Vas will be able to continue working on Rybka for a long, long time without even putting curtains in the windows!
Parent - - By Highendman (****) Date 2009-07-29 22:03
Future Breeze and I played a two game centaur match @ 60+0. The first was a French book draw. The second a QID. Here it is:

[Event "Rated game, 60m + 0s"]
[Site "Engine Room"]
[Date "2009.07.29"]
[Round "?"]
[White "Future Breeze"]
[Black "Highendman"]
[Result "0-1"]
[ECO "E16"]
[WhiteElo "2433"]
[BlackElo "2474"]
[PlyCount "90"]
[EventDate "2009.07.29"]
[TimeControl "3600"]

1. d4 {3} Nf6 {3} 2. c4 {2} e6 {20} 3. Nf3 {4} b6 {74} 4. g3 {3} Bb7 {27} 5.
Bg2 {2} c5 {31} 6. d5 {2} exd5 {18} 7. Ng5 {3} h6 {245} 8. Nh3 {5} Qc7 {8} 9.
cxd5 {3} d6 {5} 10. e4 {3} g6 {6} 11. O-O {132} Bg7 {7} 12. f4 {76} Nbd7 {60}
13. Re1 {264} O-O {24} 14. a4 {122} h5 {207} 15. Nf2 {9} Rfe8 {190} 16. Nc3 {4}
a6 {29} 17. Be3 {16} Rab8 {29} 18. h3 {44} Qd8 {41} 19. Bf1 {67} Nh7 {104} 20.
h4 {469} Ndf6 {234} 21. Be2 {36} Bc8 {137} 22. Qd3 {146} Ra8 {67} 23. Kg2 {50}
Qe7 {12} 24. Rad1 {40} Nf8 {134} 25. Qc2 {82} Ra7 {77} 26. Rc1 {39} Qd8 {32}
27. Bd2 {44} Rae7 {41} 28. b3 {67} Qd7 {135} 29. Qd3 {4} N8h7 {37} 30. Rb1 {9}
Ng4 {265} 31. Bxg4 {18} hxg4 {8} 32. h5 {140} gxh5 {32} 33. f5 {40} Bd4 {88}
34. Be3 {68} Bxe3 {3} 35. Rxe3 {0} Qd8 {25} 36. Rh1 {7} Bxf5 {18} 37. Rxh5 {1}
Bg6 {6} 38. Rh4 {7} f5 {26} 39. Re2 {6} Re5 {130} 40. Qxa6 {50} Qf6 {5} 41.
Qxb6 {356} Ng5 {7} 42. Nd3 {9} Rxe4 {36} 43. Nxe4 {7} fxe4 {31} 44. Nf4 {3} e3
{19} 45. Rxg4 {11} Be4+ {Future Breeze gibt auf  (Lag: Av=0.44s, max=1.2s) 22}
0-1
Parent - By The Wizard (***) Date 2010-01-09 10:03
Hi,
    In other games Future Breeze had chosen 10.a4

Future Breeze,Rybka 3 DynamicBree (2787) - Sonic der Igel,Rybka 3 (2691) [E16]
5m + 0s, rated, 12.08.2009

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Bb7 5.Bg2 c5 6.d5 exd5 7.Ng5 h6 8.Nh3 Qc7 9.cxd5 d6 10.a4 Be7 11.0-0 0-0 12.e4 Re8 13.f4 Bf8 14.Nf2 Na6 15.Bd2 Nb4 16.Na3 a5 17.Nb5 Qd8 18.Bc3 Rc8 19.Qd2 h5 20.Bf3 h4 21.gxh4 Kh7 22.h5 Ba6 23.Rae1 Bxb5 24.axb5 c4 25.e5 dxe5 26.d6 Bxd6 27.fxe5 Bc5 1-0

Future Breeze,Rybka 3 DynamicBree (2726) - Jefk,Rybka 3aqua (2574) [E16]
60m + 15s, rated, 22.08.2009

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Bb7 5.Bg2 c5 6.d5 exd5 7.Ng5 h6 8.Nh3 Qc7 9.cxd5 d6 10.a4 Be7 11.0-0 0-0 12.e4 Nbd7 13.Na3 Rae8 14.f4 c4 15.Nb5 Qb8 16.Nf2 Bd8 17.Qc2 a6 18.Nd4 Nc5 19.b4 cxb3 20.Nxb3 a5 21.Bb2 Nxb3 22.Qxb3 Nd7 23.Bd4 Qc7 24.Qa2 Ba6 25.Rfe1 f6 26.Bh3 Nc5 27.Rec1 Kh8 28.Be6 g6 29.f5 g5 30.Qd2 Qg7 31.Qd1 Re7 32.Rcb1 Qh7 33.Qh5 Rg7 34.Ra3 Rc7 35.Rc3 Ra7 36.Re1 Ra8 37.Ng4 Ra7 38.Nxf6 Bxf6 39.Bxf6+ 1-0

Future Breeze,Rybka 3 DynamicBree (2696) - Jefk,Rybka 3egtb (2683) [E16]
16m + 0s, rated, 15.09.2009

1.Nf3 Nf6 2.d4 e6 3.c4 b6 4.g3 Bb7 5.Bg2 c5 6.d5 exd5 7.Ng5 h6 8.Nh3 Qc7 9.cxd5 d6 10.a4 Be7 11.0-0 Nbd7 12.e4 a6 13.Na3 0-0 14.f4 Rfe8 15.Nf2 h5 16.h3 Rac8 17.b3 Nf8 18.Bb2 h4 19.g4 N6d7 20.g5 b5 21.axb5 axb5 22.Ng4 Qb6 23.Rf2 Ra8 24.Bf1 Ba6 25.Qf3 Rec8 26.e5 b4 27.Nc4 Bxc4 28.Rxa8 Rxa8 29.Bxc4 Ra2 30.e6 Rxb2 31.Rxb2 Qa7 32.exf7+ Kxf7 33.Re2 Nb6 34.Qe3 Nxc4 35.bxc4 Qd7 36.f5 Bd8 37.Qe6+ Qxe6 38.dxe6+ Ke8 39.g6 Nxg6 40.fxg6 b3 41.Rb2 Ke7 42.Rxb3 Kxe6 43.Rf3 Bg5 44.Rf7 1-0

Future Breeze,Rybka 3 DynamicBree (2666) - Jefk,Rybka 3egtb (2623) [E16]
16m + 0s, rated, 27.09.2009

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Bb7 5.Bg2 c5 6.d5 exd5 7.Ng5 h6 8.Nh3 Qc7 9.cxd5 d6 10.a4 Be7 11.0-0 Nbd7 12.e4 0-0 13.Na3 a6 14.f4 Rab8 15.Nf2 Rfe8 16.Re1 Bf8 17.Bd2 Bc8 18.Bf1 Bb7 19.Bc3 h5 20.Bg2 g6 21.Qd2 Nh7 22.Re2 b5 23.axb5 axb5 24.Nxb5 Qb6 25.Na3 Ba6 26.Ree1 Bg7 27.Bxg7 Kxg7 28.Qc3+ Nhf6 29.Nc2 c4 30.Ne3 Bb5 31.Ra5 Kh7 32.Rea1 Qc5 33.Bf3 Kg8 34.Kg2 Qc8 35.Ra7 Kf8 36.R1a5 Kg8 37.g4 Qd8 38.gxh5 Rf8 39.h6 Nc5 40.e5 dxe5 1-0

Regards
Parent - - By Highendman (****) Date 2009-08-01 09:57
A short wild game in the Sicilian against Eros. White sacs early for initiative. Most engines think it's 0.00 until quite late in the game when they realize things are not what they seem.

[Event "Rated game, 60m + 0s"]
[Site "Engine Room"]
[Date "2009.07.31"]
[Round "?"]
[White "Highendman"]
[Black "Auryn"]
[Result "1-0"]
[ECO "B90"]
[WhiteElo "2514"]
[BlackElo "1843"]
[PlyCount "63"]
[EventDate "2009.07.31"]
[TimeControl "3600"]

1. e4 {2} c5 {1} 2. Nf3 {2} d6 {1} 3. d4 {1} cxd4 {0} 4. Nxd4 {1} Nf6 {0} 5.
Nc3 {12} a6 {2} 6. f3 {8} e5 {13} 7. Nb3 {30} Be6 {8} 8. Be3 {62} Be7 {25} 9.
h4 {89} O-O {14} 10. h5 {46} Nc6 {256} 11. Nd5 {76} Bxd5 {568} 12. exd5 {5} Nb4
{1} 13. c4 {6} a5 {1} 14. Qd2 {11} b6 {2} 15. g4 {9} Nd7 {6} 16. g5 {188} f5 {
119} 17. g6 {5} h6 {33} 18. Bxh6 {114} f4 {5} 19. Bxg7 {62} Bh4+ {4} 20. Kd1 {
15} Kxg7 {3} 21. a3 {66} Na6 {10} 22. Qh2 {8} Bg3 {16} 23. h6+ {4} Kxg6 {12}
24. Bd3+ {14} Kf6 {10} 25. Qh3 {62} Qe8 {587} 26. h7 {16} Qf7 {208} 27. Qg4 {5}
Ke7 {48} 28. Rh6 {7} Nf6 {90} 29. Qg5 {5} Rh8 {35} 30. Bf5 {24} a4 {152} 31.
Nd2 {4} Nc5 {3} 32. Kc2 {Auryn abbandona  (Lag: Av=0.30s, max=1.3s) 5} 1-0
Parent - - By Nelson Hernandez (Gold) [us] Date 2009-08-01 11:15
Do you understand the significance of beating the reigning freestyle champion...with black, in his favorite opening?
Parent - By TGO (**) [si] Date 2009-08-01 11:31
It means we should all be looking forward to seeing the next freestyle- which is when btw? :)
Parent - - By David Evans (****) [gb] Date 2009-08-01 11:55
Thats twice he has beaten him now Highend is i belive a strong centaur player we have played a series of games in which no one has won both sides have had chances i llok forward to more games to get rid of any rust before freestyle proper starts.

Ultra-d
Parent - - By TGO (**) [si] Date 2009-08-01 12:01
Good he is, but he should start hiding his best lines!...lol, or noone will play them againts him again. (for example I saw a stuning demolishion in the french the other day- what line won't happen again-ever.
Parent - - By Highendman (****) Date 2009-08-01 17:58
Thanks TGO.
Here is that nice French win you mentioned. I hope someone allows that line against me again hehe. I spent few hours analysing it from all angles now ;)
I believe my 16. b4 is a novelty that wins by force in this line. But Nelson would surely by able to let us know if it has been played before or not. He knows everything, I've learned by now.
I'm actually just playing for fun so I don't at all mind people avoiding some lines. I'm also not trying to 'save' any novelties I find when I analyse to some important future date as this is more a fun hobby for me than something I need to make a living out of lol. So I rather test something I find in the earliest opportunity. Nelson suggested that if ever there is a new freestyle event I should pick another alias to play with. Which is a good advice. It seems though that there's no sponsor for these sort of events as no one seems to know if/when another championship will happen at all??

[Event "Friendly Game, 60m + 30s"]
[Site "Engine Room"]
[Date "2009.07.26"]
[Round "?"]
[White "Highendman"]
[Black "Big Rabbit"]
[Result "1-0"]
[ECO "C14"]
[WhiteElo "2476"]
[BlackElo "1646"]
[PlyCount "42"]
[EventDate "2009.07.26"]
[TimeControl "3600+30"]

1. e4 {2} e6 {63} 2. d4 {3} d5 {7} 3. Nc3 {103} Nf6 {7} 4. Bg5 {27} Be7 {14} 5.
e5 {36} Nfd7 {5} 6. h4 {14} Bxg5 {20} 7. hxg5 {3} Qxg5 {2} 8. Nb5 {56} Qd8 {8}
9. Qg4 {35} g6 {37} 10. Qf4 {120} a6 {81} 11. Nc3 {7} c5 {14} 12. dxc5 {27} Nc6
{13} 13. Nf3 {Big Rabbit reconnected. 25} Qa5 {137} 14. Ng5 {14} Ncxe5 {4} 15.
Nxh7 {129} Kd8 {10} 16. b4 {17} Qa3 {206} 17. Kd2 {11} Nc6 {19} 18. Qg5+ {17}
Ne7 {11} 19. Rb1 {16} d4 {69} 20. Ne4 {6} Qxa2 {164} 21. Rb3 {14} Qa1 {
Big Rabbit gibt auf  (Lag: Av=2.00s, max=38.0s) 18} 1-0
Parent - - By Nelson Hernandez (Gold) [us] Date 2009-08-01 18:35
16.b4 definitely a novelty on my end.  The last game I had in book up to that point was New Man-Zunkor, 2008.07.26.  New Man continued with 16.O-O-O.
Parent - - By Highendman (****) Date 2009-08-01 18:47
Super - good to know, thanks for confirming Nelson. From my end - that b4 wins by force on all variations.
Parent - - By Nelson Hernandez (Gold) [us] Date 2009-08-01 19:42 Edited 2009-08-01 19:49
I think 2.d4 wins by force in all variations.  I really hate the French; historically it has been by far our worst-performing line with black.  Anson's observation is that this opening is only suitable for grandmasters.  Engines tend to misunderstand the position on the board until it is far too late to salvage the situation.  Even an experienced centaur can't see as far ahead as they need to; you simply need significant chess knowledge to navigate it.  I would think you would be devastating against any opponent who dared play it against you.

One of our most memorable games was a near-defeat which required a 31-minute think after 20.Nf3 to figure out a do-or-die defense, which subsequent analysis showed may still not be enough, but was sufficient on this day.  (The engine wants to play 20...g6, which is losing.)  But the grimmest part was to go backwards in our book and try to figure where we went wrong.  You have to go back something like a dozen moves, and bear in mind we only had theory up through 2007 at the time.  It is quite difficult to figure this stuff out under tournament pressure unless you have a first-rate book and steely nerves with respect to using up the clock when you need to.  If you're like us, you don't want to be in situations like this.

[Event "5th Freestyle Main Event"]
[Site "playchess.com #091606"]
[Date "2007.03.04"]
[Round "7"]
[White "Etaoin Shrdlu"]
[Black "Cato the Younger"]
[Result "1/2-1/2"]
[ECO "C11"]
[WhiteElo "2308"]
[BlackElo "2487"]
[PlyCount "63"]
[EventDate "2007.03.04"]

1. e4 e6 2. d4 d5 3. Nc3 Nf6 4. e5 Nfd7 5. f4 c5 6. Nf3 Nc6 7. Be3 Qb6 8. Na4
Qa5+ 9. c3 cxd4 10. b4 Nxb4 11. cxb4 Bxb4+ 12. Bd2 Bxd2+ 13. Nxd2 O-O 14. Bd3
b5 15. Nb2 Nb6 16. O-O Nc4 17. Nbxc4 bxc4 18. Bxh7+ Kxh7 19. Qh5+ Kg8 20. Nf3
f6 21. exf6 Rxf6 22. Ng5 Qd8 23. Qh7+ Kf8 24. Qh8+ Ke7 25. Qxg7+ Kd6 26. Nh7
Rf5 27. g4 Qd7 28. Qxd7+ Bxd7 29. gxf5 exf5 30. Ng5 d3 31. Rac1 Rc8 32. Nf3
1/2-1/2
Parent - - By Highendman (****) Date 2009-08-01 20:03
Great game - thank you for posting it. It seemed black had a death wish up to move 20 ;)
But then seeing that drawn end-game attempt was a very nice save. Yes, seems White could still have won with a different plan, but probably not under the time control.
Really an exciting game to have played I'm sure. I had fun quickly analysing it now.
Parent - By Roland Rösler (****) [de] Date 2009-08-02 00:24
Okay, enough time for analysing! :-)
You have to find 28. Qh6! (and maybe 30. f5!). Nick Carlin showed it about 4 months later in his Freestyle game against New Man.
Parent - - By Highendman (****) Date 2009-08-01 17:28
Your style makes it extremely difficult to beat you, for me, Ultra-d.
I noticed your playchess profile signature mentions you had a series of draws in the last freestyle event you participated. After getting 8-9 draws against you I know why ;)
I analysed all our games, and found one forced win I missed in one of them, and no real chance for either side in any of the others.
Yes, I look forward to playing some more games with you until one of us wins 2 games and settles the score hehe. It may take many more games though until one of us concedes a full point to the other lol.
Parent - By stephanie (**) Date 2009-08-02 06:38

> I analysed all our games, and found one forced win I missed in one of them


Please do post this game with your analysis on it, thanks.
Parent - - By Highendman (****) Date 2009-08-01 17:24
Pardon my ignorance, but who is the reigning freestyle champion you refer to? Eros? I didn't know that. When did that most recent championship take place, btw?
I didn't win against him in black if he was the one you were referring to. I won two white games, and got a draw in 4 black/white games, in a couple of which I had to work very very hard for the draw. I did miss one win that I found 10 minutes after one of the games was over, in a very tricky ending.
Parent - By Nelson Hernandez (Gold) [us] Date 2009-08-01 19:04
The last Chessbase-run Freestyle was April 2008.  (So long ago!)  Since then there was an InfinityChess tournament in December 2008 that ended in chaos as the server simply was not ready for prime time.  The TD was overmatched by the repeated crashes (probably fifteen or so) and isolated cases of players having nervous breakdowns as a result.  Eros won both tournaments, so nobody can really dispute that he's performed best in tournaments lately.  In general, tournaments with significant prize funds bring out tremendous competitors at the Ultra-d level and even better.  Certainly Vas is extremely formidable and a two-time tournament winner.  Steve Cramton's team is extremely strong as well.  There are a number of 2000+ players in Europe that turn out as well with quite strong machines and are always a threat; some of them specialize in oddball openings.  Basically what it comes down to is that the winner must be extremely skilled, have the right setup, and as in human tournaments they need a little luck.  Unexpected things just happen: fingerslips, server crashes or bugs, ISP crashes and hardware failures by individual players, players dropping out or missing rounds, disqualifications, games that appear to be fixed, players gaining or losing hardware, erroneous pairings.  Anything can happen and something always does.

I think you'd really get a charge out of the whole thing, though you need a lot of stamina.  Imagine as many as eight very tough matches in a Friday-Sunday span.  I think the biggest tournament had something like 130 contestants but usually it has been in the 100 range for Chessbase.  (Much lower for Infinity.)

We await news from the Infinity website.  According to the patron saint of Freestyle chess, Arno Nickel, a tournament schedule will be forthcoming in October.  However this is an update from the previous statement, that we would have a schedule in June.  As other sponsors are not available we have no choice but to be patient.
Parent - - By Vasik Rajlich (Silver) [pl] Date 2009-08-03 07:26
Highendman is putting together a nice scalp collection. Hopefully there will be some official tournaments and we can all re-sort this out. If not, we'll probably do some formal matches.

Vas
Parent - - By Highendman (****) Date 2009-08-03 09:35
Two more 'scalps' from Sunday night:

I think Future-Breeze gave up too quickly on the first game, but I really was 'on top' in it. I was analysing way ahead when he resigned and while I had no forced winning lines, I did have +- position.

[Event "Rated game, 60m + 0s"]
[Site "Engine Room"]
[Date "2009.08.02"]
[Round "?"]
[White "Highendman"]
[Black "Future Breeze"]
[Result "1-0"]
[ECO "B90"]
[BlackElo "2418"]
[PlyCount "35"]
[EventDate "2009.08.02"]
[TimeControl "3600"]

1. e4 {3} c5 {3} 2. Nf3 {2} d6 {2} 3. d4 {1} cxd4 {1} 4. Nxd4 {1} Nf6 {1} 5.
Nc3 {2} a6 {1} 6. f3 {1} e6 {2} 7. a3 {33} b5 {5} 8. h4 {58} e5 {10} 9. Nb3 {34
} Be7 {15} 10. g4 {93} Be6 {102} 11. g5 {94} Nfd7 {4} 12. Nd5 {24} Nb6 {9} 13.
f4 {83} Bxd5 {2} 14. exd5 {2} Nc4 {2} 15. f5 {67} Nd7 {65} 16. Bd3 {6} Qb6 {74}
17. a4 {239} Qb7 {80} 18. Qf3 {
Future Breeze gibt auf  (Lag: Av=0.48s, max=1.2s) Future Breeze gibt auf 6} 1-0

On the second game 41. Bxa6 by white loses by force. But other lines should lose as well. Unfortunately Future-Breeze got disconnected and had technical problem resuming the game after my 41. .. Ra1, and conceded the game. So we didn't get a chance to play the truth of the position, but with some analysis it should be pretty clear.

[Event "Rated game, 60m + 0s"]
[Site "Engine Room"]
[Date "2009.08.02"]
[Round "?"]
[White "Future Breeze"]
[Black "Highendman"]
[Result "0-1"]
[ECO "D30"]
[WhiteElo "2407"]
[PlyCount "82"]
[EventDate "2009.08.02"]

1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. Bg5 dxc4 5. e4 b5 6. Nc3 a6 7. a4 b4 8. Na2 Bb7
9. e5 h6 10. Be3 Ng4 11. Bxc4 c5 12. Bf4 g5 13. Bg3 h5 14. h4 Nc6 15. dxc5 Bxc5
16. Qe2 Nd4 17. Nxd4 Bxd4 18. O-O-O Qb6 19. Rhe1 Bxf2 20. Bxf2 Qxf2 21. Nxb4
Qxe2 22. Rxe2 gxh4 23. a5 Ke7 24. b3 Rhg8 25. Rh1 Rg5 26. Nd3 Nh6 27. Rxh4 Bxg2
28. Kb2 Nf5 29. Rh2 Bc6 30. Ka3 Nd4 31. Re3 Bg2 32. Ne1 Bb7 33. Nd3 h4 34. Reh3
Bg2 35. Rxh4 Nf3 36. Rh5 Rg3 37. Rxg2 Rxg2 38. Nc5 Rc2 39. Na4 Rc1 40. Nb6 Rb8
41. Bxa6 Ra1+ 0-1

I'm now at +3 against Future-Breeze, +2 against Eros, +2 against the Cluster, +1 against Sebi and = against Ultra-D.
I hope to get a chance to play few more games before I start my vacation - this sort of chess is a lot of fun.
Parent - - By jaboo2 (**) Date 2009-08-08 03:18
Highendman plays like Kasparov played during his best years as a world champion. Strong tactical and positionally impressive play, just like Kasparov in 1996-7.

If we could turn back time back to 1996 and Highendman could challenge Kasparov instead of Deep Blue, do you think Highendman would win against Kasparov during his best years?
Parent - - By Uly (Gold) [mx] Date 2009-08-08 07:55
Highendman is seemingly playing at a rating much higher than Kasparov at his peak. It's difficult to comprehend, but Cluster Rybka performed relatively impressive against centaurs, and I think Highendman is playing stronger than her.
Parent - - By Gaмßito (****) [cr] Date 2009-08-08 20:13

> and I think Highendman is playing stronger than her.


I think he seems to be stronger but also because the mighty Cluster used a limited book and in fact it was also not a very strong opening book. Of course that was also part of Luka's experiment, but it seems that against very strong centaurs these things are simply not good and will be paid very expensive in most cases.

And I do not try nor want to remove his merits - Highendman's merits. The Cluster was outplayed by him in some games in great style; that show us clearly that human+computer can be a very fearful combination, even against a superior hardware and program but without human assistance.

Regards,
Gaмßito.
Parent - - By Uly (Gold) [mx] Date 2009-08-09 02:14
Highendman has also played and defeated the Freestyle Champion, twice, and remains undefeated if we ignore the lose on time against him (i.e. the game was drawn at some point, Eros metabeat him.)
Parent - - By Auryn (***) [it] Date 2009-08-09 03:16 Edited 2009-08-09 03:20
He has Defeated me, and as I said he's for sure stronger than me with the big hardware advantage he has. (I used only Quad 6600 overclocked at 3000Mhz, with lag when using all 4 processors).
Maybe When he is back we can arrange a similar hardware match... and then I am sure it will be much harder for him to beat me when he has white using the "trick" of taking me out of book as soon as possible, and will also be much harder for him to defend almost desperate positions when he has Black ;-)

Eros.
Parent - - By Roland Rösler (****) [de] Date 2009-08-09 04:09
To play freestyle with Q6600 3 GHz against two Nehalem EP 3.2 GHz is suicide! :-)
It would be fun to see the real strength in a freestyle tournament with ~ $10,000 price money. Then all the centaurs are best armed! :-)
Parent - - By Auryn (***) [it] Date 2009-08-09 11:44
" It would be fun to see the real strength in a freestyle tournament with ~ $10,000 price money. Then all the centaurs are best armed! "

Exactly, the games we played were only for fun, we both (at least I) didn't put maximum efforts in them, he everytime played a different variation, and also I would have played better variations if it was an official game. I also liked to "experiment" some sublines with him, just to see what would happen. Also in my loss with the Cluster, I tried to avoid a draw by repetition for the curiosity to see "what would have happened". If that was an official game, and I was with my Q6600 against the Cluster... I would have probably taken the draw.
But again, these games don't say the truth about the real strenght of a player, they are helpful anyway for practice in centaur play.
Same thing can be said for the Cluster, where the "not so good" book it used caused the machine some defeats that wouldn't have happened in a real freestyle tournament where for sure a much better book would be used.

Eros
Parent - - By Vasik Rajlich (Silver) [pl] Date 2009-08-11 05:35
Things look pretty good right now I think for some serious events in the fall.

Vas
Parent - By Banned for Life (Gold) Date 2009-08-11 07:28
They can't be that serious if they let Eros play...
Parent - By Master Om (Bronze) [in] Date 2010-01-10 15:27

>with the big hardware advantage he has.


Thats the advantage Shahar has.

>Maybe When he is back we can arrange a similar hardware match... and then I am sure it will be much harder for him to beat me when he has white using the "trick" of taking me out of book as soon as possible, and will also be much harder for him to defend almost desperate positions when he has Black ;-)


Totally Agree. +1
Parent - - By Master Om (Bronze) [in] Date 2010-01-10 15:22
He has a Good Hardware to back up his analysis. can he perform like this with only a quad ?
Up Topic Rybka Support & Discussion / Rybka Discussion / Should there be a Centaur games thread?
1 2 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill