Author: Bruce Moreland
Date: 12:19:25 03/19/99
Go up one level in this thread
On March 19, 1999 at 14:36:48, blass uri wrote: > >On March 19, 1999 at 13:49:07, Mark Young wrote: > >>On March 19, 1999 at 12:48:04, Bruce Moreland wrote: >> >>>5rk1/5ppp/p1Q1p3/1R6/q7/4b1P1/P2RPP1P/6K1 w - - 0 1 >>> >>>The key is 1. Rd8, which Kasparov didn't find. >>> >>>It isn't hard to find for a computer, but I think they will typically think it >>>is a draw. Can anyone find a score that indicates *significant* advantage to >>>white? >> >>Hiarcs7 has a nice + score for this position for the start. Here is Hiarcs7 line >>after about 30 secs on a P II 400. >> >>10->30 27 1487kN d2-d8 f8xd8 b5-d5 a4-d4 d5xd4 e3xd4 c6xa6 g8-f8 = 309 >> >>Hiarcs 7 score only increased as it searched deeper, and the line was almost the >>same. >> >>> >>>What this means in practice is, can anyone resolve (hopefully from the root) the >>>near-perpetual after: >>> >>>1. Rd8 Qxb5 2. Qd6 Bxf2+ 3. Kxf2 Qf5+ >> >>Hiarcs 7 scored the position after this line at +593 > >The question is if it scored it at +593 for the correct reason because I saw a >case in the ssdf games when Hiarcs7 scored a repetition line as positive. > >Fritz5.32 did a draw against Hiarcs7 because of this reason. This is possible, but it's hard to check. It avoided this line entirely in the first position, going instead for a very lost endgame, with the appropriate bad score. Mine does the same thing but it takes a lot longer, it takes it 13 minutes on a P2/300 to see that white is winning (although it does play Rd8 with a draw in under two seconds). I can imagine a program seeing the win a lot faster if it was more aggressively extending, as well as if it were being speculative, as you suggest. bruce
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.