Author: Uri Blass
Date: 06:27:04 05/02/03
Go up one level in this thread
On May 02, 2003 at 09:16:13, Uri Blass wrote:
>On May 02, 2003 at 09:08:08, Uri Blass wrote:
>
>>On May 02, 2003 at 09:01:29, Mogens Larsen wrote:
>>
>>>On May 02, 2003 at 08:36:30, Djordje Vidanovic wrote:
>>>
>>>>Hello Tony,
>>>>
>>>>which book is Ruffian using? Sorry if I am pestering you with a question
>>>>already asked, but I am really interested. Fritz 8 must be using its own Fritz
>>>>8 book (by A. Kure), how about Ruffian?
>>>>
>>>>Thanks in advance.
>>>
>>>Ruffian is using its own book available at the website. And since it's installed
>>>as an UCI engine there's no book learning. Essentials from answers to similar
>>>questions like the one above :-).
>>>
>>>Regards,
>>>Mogens
>>
>>Does Ruffian has some other kind of learning that is not book learning to
>>prevent losing another game in the same way?
>>
>>I can think about positional learning that is based only on the scores.
>>
>>Uri
>
>For example using more time in move 51 when you remember that you had a fail low
>at move 52
>
>I am interested to know what are the practical results of the following way of
>learning:
>
>1)playing many the moves that you played in a previous games in 0 seconds
>2)starting to calculate and using more time one move before the move that you
>remember a negative evaluation of more than 0.5 pawn against yourself.
>
>What is the probability that a program that lose is going to lose again in these
>conditions?
>
>Note that one of the possibilities may be that the opponent may have no time to
>use on your time so the opponent may play a different move earlier.
>
>Uri
I can add that maybe this kind of learning could help here
[Event "SSDF"]
[Site "Tony Hedlund"]
[Date "2003.05.02"]
[Round "5"]
[White "Fritz 8 A1200"]
[Black "Ruffian 1.0.1 A1200"]
[BlackElo "2671"]
[ECO "C11"]
[Result "1-0"]
1. e4 {0s} e6 {book 0s} 2. d4 {0s} d5 {book 0s} 3. Nc3 {0s}
Nf6 {book 0s} 4. Bg5 {0s} dxe4 {book 0s} 5. Nxe4 {0s} Be7
{book 0s} 6. Bxf6 {0s} Bxf6 {book 0s} 7. Nf3 {0s} O-O {book
0s} 8. Qd2 {0s} Nd7 {book 0s} 9. O-O-O {0s} b6 {book 0s}
10. Bc4 {0s} c6 {-0.13/13 4:23m} 11. Qf4 {9:34m (Nxf6+)}
Be7 {-0.05/14 2:44m} 12. Ne5 {4:48m} Qc7 {-0.06/14 0s}
13. Qg3 {4:02m} Bd8 {-0.09/13 28s} 14. Kb1 {9:39m (Ng5)} a5
{-0.14/13 7:50m} 15. Rhe1 {3:21m} a4 {-0.14/13 3s} 16. a3
{9:18m (Re3)} Nf6 {-0.17/13 5:07m} 17. Bd3 {11:13m} Nd7
{-0.25/14 0s} 18. f4 {4:12m} Ba6 {-0.40/13 4:09m} 19. Bxa6
{6:00m (Qf3)} Rxa6 {-0.48/16 22:50m} 20. Qf3 {0s (Qc3)}
Nxe5 {-0.44/15 17:21m} 21. dxe5 {0s} Be7 {-0.45/14 2:11m}
22. Qg4 {4:00m} Ra5 {-0.30/13 1:51m} 23. h4 {1:01m (c4)}
Rd8 {-0.16/13 2:02m} 24. h5 {2:52m} Rad5 {0.00/14 0s}
25. Nc3 {5:21m (Nf6+)} Rd4 {+0.04/13 1:07m} 26. Rxd4
{3:21m} Rxd4 {-0.02/17 0s} 27. h6 {3:46m (Re4)} g6
{+0.05/15 1:26m} 28. Ne4 {4:10m (Qf3)} Qd8 {+0.08/14 56s}
29. Qe2 {6:26m} b5 {0.00/16 0s} 30. c3 {3:52m} Rd5 {0.00/16
0s} 31. Rc1 {3:28m (g4)} Kf8 {+0.01/13 1:11m} 32. Rf1
{5:16m (g4)} Qd7 {+0.01/13 53s} 33. Ka1 {1:43m (Kc1)} c5
{+0.07/14 1:00m} 34. g4 {6:28m} c4 {0.00/16 42s} 35. Ka2
{1:04m (g5)} Ke8 {0.00/13 49s} 36. Qg2 {1:08m (Ng5)} Qc7
{0.00/14 54s} 37. Qf2 {50s (Ng5)} Kf8 {-0.04/13 49s} 38. f5
{1:41m} gxf5 {-1.78/16 7:40m} 39. gxf5 {0s} Qxe5 {-1.74/15
45s} 40. f6 {0s} Qxe4 {-1.95/15 1:13m} 41. fxe7+ {0s} Kxe7
{-4.84/18 21:26m} 42. Qxf7+ {0s} Kd6 {-4.34/16 4:57m}
43. Qb7 {0s} Qc2 {-6.28/16 6:40m} 44. Qb6+ {1:00m} Ke7
{-9.01/18 2:22m} 45. Qc7+ {0s} Rd7 {-9.01/16 2:09m}
46. Qc5+ {0s} 1-0
Ruffian lost after Kf8
With more time Ruffian says
depth=13 -0.04 Kf8 f5 gxf5 gxf5 Qxe5 f6 Bd6 Qg2 Ke8 Qg8+ Bf8 Qxh7 Qh2 Re1 Bxh6
Nodes: 22661583 NPS: 499153
Time: 00:00:45.40
depth=14 +0.00 Qd7 g5 Kf8 Re1 Kg8 Nf6+ Bxf6 exf6 Rd2 Re2 Rd1 Qc5 Ra1+ Kxa1 Qd1+
Ka2 Qb3+ Ka1 Qd1+
Nodes: 72064174 NPS: 498196
Time: 00:02:24.65
depth=15 -0.01 Qd7 Re1 Qc7 g5 Kf8 Nf6 Bxf6 gxf6 Qd7 Rg1 Rd2 Qc5+ Ke8 Qb6 Kf8
Qb8+ Qe8 Qc7
Nodes: 147861573 NPS: 508954
Time: 00:04:50.52
Note also that I do not understand the reason that Ruffian used so little time
because based on the reported time it had time on the clock.
I wonder if Ruffian's time management is better in Leo's tournament(40/40 time
control).
Uri
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.