Author: Reinhard Scharnagl
Date: 02:27:49 01/15/04
Go up one level in this thread
On January 14, 2004 at 23:12:53, Ed Trice wrote: >We have had a few discussions about the piece values where the Chancellor is >always valued more highly than the Archbishop. Here is a demonstration of how >losing a Gothic Exchange of Chancellor for Archbishop on move 23 allows White to >eventually go up a pawn in the endgame on move 48 -- 50 plies later! This is >beyond the scope that any program is capable of searching, and demonstrates how >a strategic mind will be able to outgun "faultless" tactics. > >Reinhardt's opinion on this game would be most interesting :) > > >[Event "8x10 battles!"] >[Site "BrainKing.com (Prague, Czech Republic)"] >[Date "2003.12.03"] > > >[Round "1"] >[White "Ed Trice"] >[WhiteElo "2339"] >[Black "BrainKing Player"] >[BlackElo "1882"] >[Result "1-0"] > >1. Nh3 Nh6 2. c4 g6 3. Nc3 Nc6 4. g3 Bg7 5. d3 d6 6. Bg2 Af6 7. Bg5 Ah5 8. f3 >O-O 9. g4 Ai6 10. Bi3 Rf8 11. Ng5 Ah8 12. Rc1 f6 13. Ni4 Ai6 14. Nxh6+ Bxh6 15. >Ra1 f5 16. gxf5 Bxf5 17. Bh3 Ag7 18. Bxf5 Axf5 19. Cg2 Qd7 20. h3 Ae6 21. Bj4 >Bxj4 22. Axj4 Af4 23. Cxf4 Rxf4 24. Ah2 e5 25. e3 Rf5 26. O-O Cg7 27. Qe2 Raf8 >28. Raf1 Qf7 29. Rhg1 Nb4 30. b3 Ce8 31. Qd2 Cb8 32. e4 Ri5 33. Nb5 a5 34. a3 >Nc6 35. f4 exf4 36. Rxf4 Qe7 37. Rgf1 Rxf4 38. Qxf4 j6 39. Ag4 Rg5 40. Ai3 h6 >41. h4 Re5 42. Ag4 Ch8 43. Af6 Cf7 44. Axe7 Cxf4 45. Rxf4 Rxe7 46. Rf8+ Kj7 47. >Rc8 Ne5 48. Rxc7 Re6 49. d4 1-0 Hello Ed, well, I am not a experienced annotator, but I will try to give my impressions. The estimated exchange values for GC pieces connot be refuted or proven by this game of Gothic Chess. Because both sides obviously do not make the best of their pieces during the phase of opening. Heavy pieces are partially brought into the game much too early and often, whereas some other pieces have been forgotten for a long period. There is (in my view) no direct implication between the addressed exchange in move 23 to the final outcome of the game. Opening here seems to be driven only by the wish to enable castlings. But freeing the squares between king and rook only gave me the impression of having no relevant additional understanding of good positional developing and placing. In games of conventional chess a comparable effect is covered by the implementation of huge opening libraries, which are (luckyly still) missing in GC. And because there are 8 more pieces in GC, this effect has become even bigger in this variant. Positivly spoken GC has the potential to give us a good drosophila (like FRC) to test approaches to create better evaluation functions working during the phase of opening, because it is much more important there in GC (or in FRC) than in traditional chess. For the moment there seems to be only one method to verify or refute estimated exchange values: having there a big number of games, whose endings could be statistically significant analysed. Regards, Reinhard.
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.