Author: Slater Wold
Date: 10:34:59 08/02/02
Go up one level in this thread
On August 02, 2002 at 13:32:42, Uri Blass wrote: >On August 02, 2002 at 13:19:20, Slater Wold wrote: > >>On August 02, 2002 at 12:54:00, Uri Blass wrote: >> >>>I stop the analysis. >>>Even at depth 14 Deep Fritz cannot see even one >>>single negative number. >>> >>>[D]r7/3nbpk1/4qnp1/4p1Bp/2P1P3/1pP2Q2/4N1PP/3R1R1K w - - 0 1 >>> >>>Analysis by Deep Fritz: >>> >>>1. ² (0.63): 3.Ng3 Ra6 4.Rd5 Nb6 5.Qe2 Ra4 6.Rb5 Nxc4 7.Qd1 >>>2. ² (0.59): 3.Rb1 Qxc4 4.Ng3 Rh8 5.Bh4 Rb8 6.Rb2 Rb6 7.Bg5 >>>3. ² (0.41): 3.Nc1 Qxc4 4.Ne2 b2 5.Rb1 Qxe4 6.Rxb2 Qxf3 7.gxf3 >>>4. ² (0.41): 3.Rd5 Nb6 4.Rb1 Nxc4 5.Bxf6+ Qxf6 6.Qd3 Ra2 7.Qxc4 Qf2 8.Rg1 >>>5. ² (0.41): 3.Bh4 Qg4 4.Bxf6+ Nxf6 5.Rb1 Qxe4 6.Rxb3 Ra2 7.Rb7 Qxf3 >>>6. ² (0.41): 3.Rd2 Qg4 4.Qxg4 hxg4 5.Rb2 Nc5 6.Nc1 Ra3 7.Kg1 g3 >>>7. ² (0.28): 3.Rde1 Rb8 4.Ng3 Nh7 5.Bxe7 Qxe7 6.Rf2 Nc5 7.Qd1 Nf6 >>>8. ² (0.28): 3.Rfe1 >>>9. = (0.25): 3.h3 Ra4 4.Ng3 Nc5 5.Kg1 Rxc4 6.Rde1 Nfd7 7.Bxe7 >>> >>>(Blass, Tel-aviv 02.08.2002) >>> >>>Uri >> >>Here's an example of a program seeing "right" for "wrong" reasons. > >It still does not see right and has a positive score >for h3 that means that white should not lose. > >> >>h3 was the move played by Brutus - Brutus is a pretty strong program, and is >>very unlikely to play "unsound" moves. >> >>However, in the original game it played h3 in this position. DF after 14 ply >>can already tell you, h3 is _no_ good. And it doesn't even have the right line. >> >>Make DF play h3 and you play b2. Let it go 12 ply. See what the score is. I >>bet it is a lot worse than 0.25. > > >No > >Even at depth 14 the score is above 0.25. > > >I use the Deep Fritz of the ssdf list that >came before Fritz7. > >New position >[D]r7/3nbpk1/4qnp1/4p1Bp/2P1P3/2P2Q1P/1p2N1P1/3R1R1K w - - 0 1 > >Analysis by Deep Fritz: > >2.Bxf6+-- > ± (1.06) Depth: 1/3 00:00:00 >2.Bxf6+-- > ± (1.06) Depth: 1/3 00:00:00 >2.Rb1! > ± (1.22) Depth: 1/7 00:00:00 >2.Rb1! > ± (1.34) Depth: 1/8 00:00:00 >2.Rb1 Rb8 3.Qd3 Bc5 > ± (1.25) Depth: 2/10 00:00:00 >2.Rb1 Rb8 3.Bxf6+ Nxf6 4.Qd3 > ± (1.03) Depth: 3/11 00:00:00 1kN >2.Rd2! > ± (1.06) Depth: 3/13 00:00:00 2kN >2.Rd2! Rb8 > ± (1.16) Depth: 3/13 00:00:00 2kN >2.Rd2 Rb8 3.Rb1 Ba3 > ± (1.31) Depth: 4/17 00:00:00 6kN >2.Rd2 Rb8 3.Rb1 Ba3 4.Qd3 > ± (1.13) Depth: 5/26 00:00:00 15kN >2.Rd2 Ba3 3.Bxf6+ Nxf6 4.c5 Qc6 5.Ng3 > ± (1.16) Depth: 6/25 00:00:00 57kN >2.Rd2 Rb8 3.Rb1 Ba3 4.Bxf6+ Nxf6 5.c5 h4 > ± (1.13) Depth: 7/25 00:00:00 153kN >2.Rd2 Rb8 3.Rb1 Ba3 4.Kg1 Qxc4 5.Be3 h4 > ± (0.88) Depth: 8/25 00:00:00 354kN >2.Rd2 Rb8 3.Rb1 Ba3 4.Ng3 Qxc4 5.Rf2 Rb6 6.Be3 > ± (0.88) Depth: 9/28 00:00:01 1234kN >2.Rd2 Rb8 3.Rb1 Ba3 4.Ng3 Qxc4 5.Rf2 Rb6 6.Be3 > ± (0.84) Depth: 10/29 00:00:05 3858kN >2.Rd2 Rb8 3.Rb1 Ba3 4.Ng3 Qxc4 5.Rf2 Nh7 6.Bh4 f5 7.Kg1 > ± (0.75) Depth: 11/31 00:00:12 8851kN >2.Rd2 Rb8 3.Rb1 Ba3 4.Ng3 Qxc4 5.Rf2 Nh7 6.Be3 Nhf6 7.Rd2 > ± (0.75) Depth: 12/35 00:00:28 20610kN >2.Rd2 Rb8 3.Rb1 Ba3 4.c5 Qa2 5.Rdd1 Bxc5 6.Bh4 Qe6 7.Bxf6+ > ² (0.59) Depth: 13/34 00:01:11 53770kN >2.Rb1! > ² (0.63) Depth: 13/34 00:01:28 65346kN >2.Rb1! Rb8 3.Qd3 Ba3 4.Ng3 Ne8 5.c5 f6 6.c6 Qxc6 7.Be3 Nc5 > ² (0.66) Depth: 13/34 00:01:38 72177kN >2.Rb1 Rb8 3.Rfd1 Ba3 4.Ng3 Qxc4 5.Qe3 Rb7 6.Rf1 Be7 > ² (0.59) Depth: 14/39 00:04:48 201777kN > >(Blass, Tel-aviv 02.08.2002) > >Uri That's interesting. DJ7 wouldn't play h3 because of b2. If I forced h3, and then played b2, it went into the negatives pretty quick. I will have to look at this when I get home. :) Working from memory is no good for me.
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.