Author: blass uri
Date: 02:13:59 03/26/99
Go up one level in this thread
On March 26, 1999 at 01:38:37, Harald Faber wrote: >On March 25, 1999 at 16:15:12, Howard Exner wrote: > >>>>Instead of Kd6, h4 looks like it should win. Black will have to eventually >>>>play h6 or h5 but then white plays f6. > >>The march of the white king to the kingside, then the march back to >>pick up the queenside pawn takes a few plys, and just seems to make it by one >>tempo. >> >>44. h4 h5 45. f6 gxf6+ 46. Kxf6 Kd5 47. Kg5 Ke5 48. Kxh5 Kf5 >>49. Kh6 Kf6 50. h5 Kf7 51. Kg5 Kg7 52. Kf5 Kh6 53. Ke6 Kxh5 54. Kd6 Kg6 >>55. Kc7 Kf6 56. Kb7 Ke7 57. Kxa7 Kd7 (just misses the the c8 square) 1-0. >> >> Is this just sheer calculating power or can knowledge intervene >>to assist? This one for humans is friendly as it is just a straight forward >>series of moves - a sort of "toes and fingers" counting exercise. > >I tought knowledge should do it too. And with a depth of at least 14 MCP should >find it. I think that it is possible that the only difference between fast searchers and slow searchers is that slow searchers see more plies by evaluation function so they can see more tactics in the same depth. Example:If I give Junior or Fritz or crafty to play at level of Depth 1 they can do stalemate with positive evaluation. They cannot see evaluation of 0 for stalemates in the leaves. Slow searchers may see that the position is a stalemate in the leaf so they see one more ply by evaluation function. I also guess that slow searchers may see by evaluation that they can do something tactical in the next move when fast searchers need to search the next move to see it. 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.