Author: Jan K.
Date: 07:09:36 07/23/04
Go up one level in this thread
Thanks for reply...I turned everything that could spoil the search off except repetition detection,check extension,null move and hashtables. now with null off repetition off: 4 3.21 c1a3 e6e1 a1e1 f2e1 h1g2 c5b4 5 2.98 c1a3 c5b4 a3b4 e6e2 a4a5 6 2.63 c1f4 e6e1 a1e1 f2e1 h1g2 e1b4 d3b3 b4b3 c2b3 7 1.91 c1f4 e6e2 d3c4 f7g6 c4d3 f6f5 d3a6 g6h5 g3g4 f5g4 a6d3 8 0.73 d3c4 f2f3 h1h2 f3e2 c4e2 e6e2 h2g1 c5b4 c1f4 e2c2 a4a5 ... although pv nor the score in this variation at depth 8 is right(and cannot be right with this setting), is sees d3c4 as the best move since depth 8 with null move off repetition on: 4 2.44 c1a3 e6e1 a1e1 f2e1 h1g2 c5b4 5 0.11 d3c4 f7g6 c1e3 e6e3 c4c5 6 0.00 d3c4 f2f3 h1h2 f3f2 h2h1 ... that's alright with null move on repetition on: 4 2.43 c1a3 e6e1 a1e1 f2e1 h1g2 c5b4 5 2.56 c1f4 e6e1 a1e1 f2e1 h1g2 e1b4 d3b3 b4b3 6 2.56 c1f4 e6e1 a1e1 f2e1 h1g2 e1b4 d3b3 b4b3 c2b3 7 2.60 c1b2 e6e1 a1e1 f2e1 h1g2 e1b4 b2a3 b4a4 a3c5 8 0.00 c1b2 e6e2 d3c4 f7g6 c4d3 g6f7 ... wrong, but it sees d3c4 as repetition too at root, it's just that c1b2 comes first in move ordering :)) with repetition off null move on: 4 2.45 c1a3 e6e1 a1e1 f2e1 h1g2 c5b4 5 2.95 c1a3 c5b4 a3b4 e6e2 a4a5 6 2.54 c1f4 e6e1 a1e1 f2e1 h1g2 e1b4 d3b3 b4b3 c2b3 7 2.61 c1b2 e6e1 a1e1 f2e1 h1g2 e1b4 b2a3 b4a4 a3c5 8 2.61 c1b2 e6e1 a1e1 f2e1 h1h2 e1f2 h2h1 f2e1 9 1.95 c1f4 e6e2 d3c4 f7g6 c4d3 g6f7 d3c4 f7g6 c4d3 f6f5 d3a6 g6h5 g3g4 f5g4 a6d3 10 0.68 d3c4 f2f3 h1h2 f3f2 h2h1 f2f3 h1h2 f3e2 c4e2 e6e2 h2g1 c5b4 c1f4 e2c2 a4a5 ... strange.... So it seems it's really a 1)null move problem or 2)null move+repetition problem. Despite the fact that 2) is VERY unlikely, could you please try what happens when you turn off repetition detection? Depth 4-10 should be enough to see if you get more logical pv than me. I hope it doesn't bother you...for i want to make everything as clear as possible before spending hours exploring the whole search tree......
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.