Author: Omid David Tabibi
Date: 15:31:49 08/02/03
Go up one level in this thread
On August 02, 2003 at 14:51:38, Werner Schuele wrote:
>Hi Ed,
>thanks for the fast answer. Of course in games I do not use such high parameters
>- I only wanted to see, which parameter helps Rebel to see earlier this game ist
>comletely lost. It was not possible to do this with the parameter passed pawn.
>Do you know another way?
Every parameter can backfire. For example, increase the passed pawn bonus, and
the program will start reckless sacrifices to obtain passed pawns; increase the
doubled pawn penalty, and the program will avoid those positions at any cost,
even when they are advantageous; increase the queen value, and the program will
worship it, not trading it for even better pieces; reduce the queen value, and
the program will start unjustified queen exchanges for rook and minor piece, and
so on....
Finding the optimal value for each parameter is impossible, as a good value for
position A could be devastating for position B (you can find many example from
computer tournaments that engines overvalue or undervalue passed pawns, ets).
So, somehow a balancing tradeoff point should be found so that it _usually_
gives a realistic view (again, not always).
>Werner
>
>
>On August 02, 2003 at 14:41:56, Ed Schröder wrote:
>
>>On August 02, 2003 at 14:18:39, Werner Schuele wrote:
>>
>>>Hi Ed,
>>>which parameters help Rebel to see much earlier how dangerous this passed pawn
>>>is. Look at this game - there is a big difference in the evaluation. I made
>>>passed pawn to 500 ! this helps not much and maybe is not the correct way for
>>>other positions:
>>
>>
>>Werner,
>>
>>[Passed Pawns = 500] is complete craziness, values that make sense are between
>>90-115 maybe 120 too, but 500 is asking for trouble :)
>>
>>My best,
>>
>>Ed
>>
>>
>>
>>
>>>Werner
>>>
>>>[Event "Wertungspartie, 60m + 15s [Schüle,Werner]"]
>>>[Site "?"]
>>>[Date "2003.??.??"]
>>>[Round "?"]
>>>[White "cm8000ws,Rebel 12 PRB9"]
>>>[Black "Katzenmaier,Aristarch 4.29"]
>>>[WhiteElo "2450"]
>>>[BlackElo "2625"]
>>>[ECO "B87"]
>>>[Result "0-1"]
>>>
>>>1. e4 {0.01/1 0} c5 {0} 2. Nf3 {0.01/1 3} d6 {0} 3. d4
>>>{0.01/1 4} cxd4 {0} 4. Nxd4 {0.01/1 4} Nf6 {0} 5. Nc3
>>>{0.01/1 3} a6 {0} 6. Bc4 {0.01/1 4} e6 {0} 7. Bb3 {0.01/1
>>>4} b5 {0} 8. O-O {0.01/1 4} Be7 {0} 9. Qf3 {0.01/1 4} Qc7
>>>{0} 10. Qg3 {0.01/1 4} O-O {0} 11. Bh6 {0.01/1 4} Ne8 {0}
>>>12. Rad1 {0.01/1 4} b4 {0.47/14 305} 13. Nce2 {0.01/1 0}
>>>Kh8 {0.47/13 61} 14. Be3 {0.47/12 92} Nf6 {0.27/14 159}
>>>15. f3 {0.29/12 73} e5 {0.19/15 0} 16. Nf5 {0.12/13 121}
>>>Bxf5 {0.04/16 0} 17. exf5 {0.23/9 13} a5 {0.02/14 28}
>>>18. Ba4 {0.08/12 92} d5 {0.00/13 0} 19. f4 {0.19/11 76} e4
>>>{-0.23/13 47} 20. Nd4 {0.12/11 52} Bc5 {-0.25/13 19}
>>>21. Rd2 {0.10/11 71} Nbd7 {-0.31/14 0} 22. Bb5 {-0.01/11
>>>101} Rab8 {-0.53/14 171} 23. Be2 {-0.01/12 102} Nb6
>>>{-0.54/15 0} 24. b3 {-0.08/11 64} Nc8 {-0.84/13 53} 25. Qh3
>>>{-0.12/12 29} Ne7 {-1.34/14 61} 26. g4 {-0.12/12 41} Nc6
>>>{-1.36/14 8} 27. Nxc6 {-0.17/13 99} Qxc6 {-1.50/12 16}
>>>28. Rfd1 {-0.27/12 46} Rbd8 {-1.53/15 188} 29. g5 {-0.33/14
>>>0} Ng8 {-1.65/12 26} 30. Kh1 {-0.25/12 44} Bxe3 {-1.50/14
>>>167} 31. Qxe3 {-0.43/14 0} Ne7 {-1.60/14 56} 32. Bg4
>>>{-0.41/12 6} Kg8 {-1.54/13 29} 33. Qa7 {-0.15/12 82} Rfe8
>>>{-1.55/14 0} 34. Qxa5 {-0.15/12 71} Qd6 {-1.63/14 117}
>>>35. f6 {-0.37/12 64} gxf6 {-1.61/14 0} 36. gxf6 {-0.39/13
>>>130} Ra8 {-1.75/17 0} 37. Qb5 {-0.24/13 106} Reb8 {-1.69/15
>>>15} 38. Qxb8+ {-0.42/13 95} Rxb8 {-1.67/15 37} 39. fxe7
>>>{-0.60/13 1} Qxe7 {-1.67/14 13} 40. Rxd5 {-0.47/14 77} Qc7
>>>{-1.79/14 0} 41. Rh5 {-0.78/13 168} Qxc2 {-2.13/14 146}
>>>42. Rg1 {-0.89/15 0} Kf8 {-2.30/14 51} 43. Rxh7 {-0.67/13
>>>52} Ke7 {-2.32/14 0} 44. f5 {-0.80/12 97} Qc6 {-2.47/13
>>>151} 45. Bh3 {-0.57/13 104} e3+ {-2.94/15 0} 46. Bg2
>>>{-0.80/12 76} Qc3 {-2.97/15 0} 47. Rh4 {-0.94/12 92} Qd2
>>>{-3.32/13 2} 48. Bf1 {-1.03/11 53} Qf2 {-3.55/13 35}
>>>49. Rhg4 {-1.10/12 55} Qf3+ {-3.65/14 0} 50. R4g2 {-1.61/14
>>>57} Rd8 {-3.62/16 0} 51. Be2 {-1.60/14 50} Qxf5 {-3.71/16
>>>0} 52. Rg3 {-2.02/14 71} Qe5 {-4.65/14 32} 53. Re1
>>>{-2.48/13 22} Rd2 {-4.75/14 8} 54. Bc4 {-2.82/12 42} Kf8
>>>{-5.51/15 128} 55. Be2 {cm8000ws,Rebel 12 PRB9 gibt auf
>>>(Lag: Av=0.27s, max=3.1s) -3.72/14 96} 0-1
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.