Author: Uri Blass
Date: 21:57:35 01/15/06
Go up one level in this thread
On January 15, 2006 at 23:33:53, Keith Ian Price wrote: >On January 15, 2006 at 22:28:12, Christopher Conkie wrote: > >>On January 15, 2006 at 20:55:17, Keith Ian Price wrote: >> >>>[D]b1r1qn1k/1rb3p1/2p1p2p/2N1P3/1P1P4/R2B1N2/3Q1PPP/2R3K1 b - - 0 43 >>> >>>In this position both Rybka Beta 1.0 and Beta 10d think they have a draw, when >>>they are losing. If you make the move that they think gets them a draw, they >>>both show that they are losing. Is this a null move problem, or why can't they >>>see they are losing until the move is made? >>> >>>kp >> >>Are you sure you have given the correct position? >> >>Christopher >> >>Rybka 1.0 Beta 32-bit: >> 3 00:00.011 228 233.472 +1.97 Rb6 >> 4 00:00.030 556 51.758 +2.03 Rb6 >> 5 00:00.060 1.722 33.910 +2.51 Rb6 Rca1 >> 6 00:00.141 4.375 40.000 +2.57 Rb6 Rca1 Kg8 >> 7 00:00.251 10.361 43.841 +2.59 Rb6 Rca1 Rbb8 Na6 >> 8 00:00.662 34.390 54.852 +2.55 Rb6 Ra7 Nd7 Ba6 Rcb8 >> 9 00:00.903 53.220 61.718 +2.61 Rb6 Ra7 Nd7 Ba6 Rcb8 Ra1 >> 10 00:01.934 137.416 73.518 +2.89 Rb6 Rca1 Rbb8 Na6 Rb7 Nxc7 Rbxc7 Ra7 >> 10 00:04.437 293.753 68.085 +2.58 Rbb8 Na6 Rb6 Nxc7 Rxc7 Ra5 Kg8 h3 >> 11 00:08.574 524.467 62.710 +2.52 Rbb8 Na6 Rb6 Nxc7 Rxc7 Ra5 Kg8 h3 Bb7 >> 12 00:17.386 1.057.238 66.077 +2.59 Rbb8 b5 Bd8 Ne4 cxb5 Rxc8 Rxc8 Nd6 Qe7 >>Bxb5 >> 13 00:37.055 2.267.229 62.978 +2.81 Rbb8 Na6 Rb6 Nxc7 Rxc7 b5 Bb7 d5 exd5 >>Qe3 >> 14 00:56.443 3.528.392 64.152 +2.66 Rbb8 Na6 Rb7 b5 Bb6 Rxc6 Rxc6 bxc6 Qxc6 >>Nb4 >> 15 01:14.068 4.806.607 65.843 +2.72 Rbb8 Na6 Rb7 b5 Bb6 Rxc6 Rxc6 bxc6 Qxc6 >>Nb4 >> 16 03:32.096 13.332.175 64.406 +2.87 Rbb8 Qa2 Rd8 Na6 Rbc8 Nxc7 Rxc7 b5 c5 >>b6 >> >>Rybka 1.01 Beta 10d 32-bit: >> 3 00:00.361 304 0 +1.97 Rb6 >> 4 00:00.361 1.352 0 +2.03 Rb6 >> 5 00:00.361 3.240 0 +2.51 Rb6 Rca1 >> 6 00:00.412 9.880 0 +2.57 Rb6 Rca1 Kg8 >> 7 00:00.552 10.683 0 +2.59 Rb6 Rca1 Rbb8 Na6 >> 8 00:00.953 35.012 0 +2.55 Rb6 Ra7 Nd7 Ba6 Nxc5 >> 9 00:01.303 57.196 0 +2.61 Rb6 Ra7 Nd7 Ba6 Nxc5 Bxc8 >> 10 00:02.355 128.790 0 +2.87 Rb6 Rca1 Rbb8 Na6 Bb7 b5 c5 >> 10 00:04.938 293.069 0 +2.58 Rbb8 Na6 Rb6 Nxc7 Rxc7 Rca1 Bb7 Ba6 >> 11 00:09.374 557.426 0 +2.52 Rbb8 Na6 Rb6 Nxc7 Rxc7 Ra5 Kg8 h3 Bb7 >> 12 00:21.292 1.250.215 0 +2.78 Rbb8 Ne4 Qe7 Ra7 Kg8 Nc5 Nd7 Nxe6 Bb6 >> 13 00:26.739 1.605.807 0 +2.72 Rbb8 Ne4 Qe7 Ra7 Kg8 Nc5 Qe8 Bc4 Qf7 Rxc7 >> 14 00:37.274 2.346.126 0 +2.80 Rbb8 Ne4 Qe7 Ra7 Kg8 Nc5 Qe8 Bc4 Kh7 Rxc7 >>Rxc7 Na6 >> 15 01:48.828 6.526.991 0 +2.75 Rbb8 Na6 Rb7 b5 Bb6 Rxc6 Rxc6 bxc6 Qxc6 Nc5 >>Bxc5 Rxa8 Rc7 Ra2 >> 16 02:37.628 9.313.279 0 +2.75 Rbb8 Na6 Rb7 b5 Bb6 Rxc6 Rxc6 bxc6 Qxc6 Nc5 >>Bxc5 Rxa8 Be7 Qf4 >> 17 06:28.349 23.124.849 0 +2.87 Rbb8 Qa2 Bd8 Bc4 Be7 Bxe6 Nxe6 Qxe6 Rxb4 Ra7 >>Bxc5 Qxe8+ Rxe8 dxc5 > > >Well, I am running the 64-bit editions, and both of them show 0.00 for Rbb8. >Here is the game from which I took the position. Rybka knew it was losing up to >that point, but then thought it had found a draw for 2 moves, then went back to >knowing it was losing. I tried the same position with Beta 10d to see if it >still did the same, and it does, even though there was a clear hash. See move >43: The reason is simple. It is known that many programs simply evaluate repetition as a draw and rybka is one of them(many programs are going to see a draw in the same conditions). It is the simple solution to the problem of avoiding repetitions. programs of course can evaluate only 3 time repetition as a draw but it may cause them to search bigger trees for no good reason so it is not clear if it is better for playing strength. programs that evaluate first repetition not as a draw may waste time by allowing repetitions and in some cases it even may cause draw by the 50 move rule because in superior position they may not be able to be lucky to find the right plan on time when avoiding repetition may force them to find the right plan if they evaluate first repetition as a draw. I choose to evaluate first repetition as a draw because when I have 2 options when it is not clear which one is better I prefer the simpler option. 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.