Author: Robert Hyatt
Date: 10:31:14 02/28/01
Go up one level in this thread
On February 28, 2001 at 11:32:34, Joshua Lee wrote: >I don't know much about this subject but i've noticed something interesting in >reguards to it. Namely when looking at endgame positions one move will have a +- >winning score and even extenions 22/22Ply but if you back up one move all you >get is a falling score with odd extentions. this is with Fritz.... > >Spassky,B - Byrne,R >3b4/p7/Pp2k3/1P2Bp2/2K2P1p/7P/8/8 w - - 0 1 > >Analysis by Fritz 6: > >58.Kd4! > ² (0.69) Depth: 1/2 00:00:00 >58.Kd4 Be7 > ² (0.59) Depth: 2/5 00:00:00 >58.Bb8! > ² (0.63) Depth: 2/5 00:00:00 >58.Bb8! Bf6 59.Bxa7 > +- (2.25) Depth: 2/5 00:00:00 >58.Bb8 Kd7 59.Bxa7 Bc7 > +- (2.22) Depth: 3/5 00:00:00 >58.Bb8 Be7 59.Bxa7 Bc5 > +- (2.31) Depth: 4/8 00:00:00 >58.Bb8 Be7 59.Bxa7 Bc5 60.Bb8 > +- (2.25) Depth: 5/8 00:00:00 >58.Bb8 Be7 59.Bxa7 Bc5 60.Bb8 Kd7 61.a7 > +- (2.22) Depth: 6/8 00:00:00 >58.Bb8 Be7 59.Bxa7 Bc5 60.Bb8 Be3 61.a7 > +- (2.34) Depth: 7/11 00:00:00 1kN >58.Bb8 Kd7 59.Kd5 Kc8 60.Bxa7 Bc7 61.Kc6 Bxf4 62.Bxb6 > +- (2.53) Depth: 8/11 00:00:00 4kN >58.Bb8 Kd7 59.Kd5 Kc8 60.Bxa7 Bc7 61.Kc6 Bxf4 62.Bxb6 Kb8 > +- (2.31) Depth: 9/12 00:00:00 5kN >58.Bb8! > +- (2.63) Depth: 10/13 00:00:00 9kN >58.Bb8-- > +- (2.31) Depth: 11/11 00:00:00 11kN >58.Bb8-- > +- (2.31) Depth: 11/14 00:00:00 14kN >58.Bb8! > +- (2.63) Depth: 12/15 00:00:00 23kN >58.Bb8-- > +- (2.31) Depth: 13/13 00:00:00 25kN >58.Bb8-- > +- (2.31) Depth: 13/16 00:00:00 33kN >58.Bb8! > +- (2.63) Depth: 14/18 00:00:00 54kN >58.Bb8-- > +- (2.31) Depth: 15/15 00:00:00 60kN >58.Bb8-- > +- (2.31) Depth: 15/19 00:00:00 83kN >58.Bb8-- > +- (2.00) Depth: 16/17 00:00:00 90kN >58.Bb8-- > +- (2.00) Depth: 16/17 00:00:00 97kN >58.Bb8-- > +- (1.69) Depth: 17/18 00:00:00 113kN >58.Bb8-- > +- (1.69) Depth: 17/20 00:00:00 125kN >58.Bb8-- > ± (1.38) Depth: 18/19 00:00:00 153kN >58.Bb8-- > ± (1.38) Depth: 18/19 00:00:00 165kN >58.Bb8-- > ± (1.06) Depth: 19/19 00:00:00 193kN >58.Bb8-- > ± (1.06) Depth: 19/24 00:00:00 317kN >58.Bb8-- > ± (0.75) Depth: 20/20 00:00:00 359kN >58.Bb8-- > ± (0.75) Depth: 20/26 00:00:00 476kN >58.Bb8-- > ² (0.44) Depth: 21/21 00:00:00 547kN >58.Bb8-- > ² (0.44) Depth: 21/28 00:00:00 596kN >58.Bb8! > ± (0.75) Depth: 22/31 00:00:02 1365kN, tb=2 >58.Bb8-- > ² (0.44) Depth: 23/25 00:00:02 1451kN, tb=2 >58.Bb8-- > ² (0.44) Depth: 23/35 00:00:02 1882kN, tb=10 >58.Bb8! > ± (0.75) Depth: 24/35 00:00:04 3460kN, tb=17 >58.Bb8-- > ² (0.44) Depth: 25/27 00:00:05 3717kN, tb=17 >58.Bb8-- > ² (0.44) Depth: 25/37 00:00:06 4703kN, tb=29 >58.Bb8-- > = (0.13) Depth: 26/36 00:00:07 5661kN, tb=38 >58.Bb8-- > = (0.13) Depth: 26/38 00:00:09 7092kN, tb=53 >58.Kd4! > = (0.16) Depth: 26/45 00:00:36 24625kN, tb=182 > >(Lee, Pensacola,Fl 28.02.2001) > > >and the next move..... > >Spassky,B - Byrne,R >1B1b4/p7/Pp2k3/1P3p2/2K2P1p/7P/8/8 b - - 0 1 > >Analysis by Fritz 6: > >58...Bf6-- > +- (1.75) Depth: 1/2 00:00:00 >58...Bf6-- 59.Bxa7 > +- (2.25) Depth: 1/3 00:00:00 >58...Bf6 59.Bxa7 > +- (2.25) Depth: 2/4 00:00:00 >58...Bf6 59.Bxa7 Bd8 > +- (2.50) Depth: 3/5 00:00:00 >58...Be7! > +- (2.31) Depth: 3/5 00:00:00 >58...Be7 59.Bxa7 Bc5 60.Bb8 > +- (2.25) Depth: 4/7 00:00:00 >58...Be7 59.Bxa7 Bc5 60.Bb8 Kd7 61.a7 > +- (2.22) Depth: 5/9 00:00:00 >58...Be7 59.Bxa7 Bc5 60.Bb8 Be3 61.a7 > +- (2.34) Depth: 6/8 00:00:00 >58...Be7-- > +- (2.66) Depth: 7/10 00:00:00 1kN >58...Be7-- > +- (2.66) Depth: 7/10 00:00:00 2kN >58...Ke7! > +- (2.62) Depth: 7/10 00:00:00 2kN >58...Ke7! 59.Bxa7 Bc7 60.Kd5 Kd7 > +- (2.34) Depth: 7/11 00:00:00 3kN >58...Ke7-- > +- (2.66) Depth: 8/12 00:00:00 4kN >58...Ke7-- > +- (2.66) Depth: 8/12 00:00:00 4kN >58...Bf6! > +- (2.62) Depth: 8/12 00:00:00 5kN >58...Kd7! > +- (2.59) Depth: 8/12 00:00:00 7kN >58...Kd7! 59.Kd5 Kc8 60.Bxa7 Bc7 > +- (2.41) Depth: 8/12 00:00:00 7kN >58...Be7! > +- (2.37) Depth: 8/12 00:00:00 7kN >58...Be7-- > +- (2.69) Depth: 9/11 00:00:00 9kN >58...Be7-- 59.Bxa7 Kd7 60.Bxb6 Kc8 61.Bc5 Bf6 62.b6 Kb8 63.Bd6+ > +- (4.16) Depth: 9/15 00:00:00 15kN >58...Kd7! > +- (4.12) Depth: 9/15 00:00:00 15kN >58...Kd7! 59.Kd5 Bc7 60.Bxc7 Kxc7 61.Ke5 Kd8 62.Kxf5 Ke7 63.Ke5 > +- (2.69) Depth: 9/15 00:00:00 17kN >58...Kd7-- > +- (3.00) Depth: 10/13 00:00:00 21kN >58...Kd7-- > +- (3.00) Depth: 10/13 00:00:00 21kN >58...Kd7 59.Kd5 Kc8 60.Bxa7 Bc7 61.Ke6 Bxf4 62.Bxb6 Kb8 63.Kxf5 > +- (3.28) Depth: 11/15 00:00:00 38kN >58...Kd7 59.Kd5 Kc8 60.Bxa7 Bc7 61.Ke6 Bxf4 62.Bxb6 Bg3 63.Kxf5 > +- (3.31) Depth: 12/17 00:00:00 58kN >58...Kd7 59.Kd5 Kc8 60.Bxa7 Bc7 61.Ke6 Bxf4 62.Bxb6 Kb8 63.Kxf5 > +- (3.41) Depth: 13/18 00:00:00 90kN >58...Kd7-- > +- (3.72) Depth: 14/19 00:00:00 139kN >58...Kd7-- > +- (3.72) Depth: 14/19 00:00:00 143kN >58...Kd7-- > +- (4.03) Depth: 15/19 00:00:00 231kN >58...Kd7-- > +- (4.03) Depth: 15/19 00:00:00 236kN >58...Ke7! > +- (4.00) Depth: 15/21 00:00:00 299kN >58...Ke7 59.Kd5 Kd7 60.Ke5 Kc8 61.Bxa7 Bc7+ 62.Kxf5 Kd7 63.Kg4 > +- (3.94) Depth: 16/20 00:00:00 364kN >58...Ke7! > +- (3.62) Depth: 17/25 00:00:00 484kN >58...Ke7-- > +- (3.94) Depth: 18/19 00:00:00 557kN >58...Ke7-- > +- (3.94) Depth: 18/24 00:00:01 612kN >58...Kd7! > +- (3.91) Depth: 18/24 00:00:01 643kN, tb=1 >58...Kd7-- > +- (4.22) Depth: 19/25 00:00:01 951kN, tb=3 >58...Kd7-- > +- (4.22) Depth: 19/27 00:00:01 1021kN, tb=4 >58...Ke7! > +- (4.19) Depth: 19/29 00:00:01 1139kN, tb=5 >58...Ke7-- > +- (4.50) Depth: 20/27 00:00:02 1631kN, tb=5 >58...Ke7-- > +- (4.50) Depth: 20/27 00:00:02 1753kN, tb=5 >58...Kd7! > +- (4.47) Depth: 20/27 00:00:02 1872kN, tb=8 >58...Kd7-- > +- (4.78) Depth: 21/29 00:00:04 3112kN, tb=25 >58...Kd7-- > +- (4.78) Depth: 21/33 00:00:04 3420kN, tb=28 >58...Ke7! > +- (4.75) Depth: 21/33 00:00:05 3692kN, tb=28 >58...Ke7 59.Kd5 Kd7 60.Ke5 Kc8 61.Bxa7 Bc7+ 62.Kxf5 Kd7 63.Kg5 > +- (5.00) Depth: 22/31 00:00:08 6007kN, tb=36 >58...Kd7! > +- (4.97) Depth: 22/31 00:00:09 6818kN, tb=69 >58...Kd7-- > +- (5.28) Depth: 23/31 00:00:12 9161kN, tb=108 >58...Kd7-- > +- (5.28) Depth: 23/38 00:00:15 11288kN, tb=112 >58...Ke7! > +- (5.25) Depth: 23/38 00:00:16 12004kN, tb=112 >58...Ke7! 59.Kd5 Kd7 60.Ke5 Kc8 61.Bxa7 Bc7+ 62.Kxf5 Kd7 63.Kg6 > +- (5.06) Depth: 23/38 00:00:17 12612kN, tb=118 > >(Lee, Pensacola,Fl 28.02.2001) > >Is this common ? It is why I don't like searching from the back of a game to the front. You store position X in the hash table and say <Y.YY score. When You look it up, you fail low, you lower alpha to -infinity, and now this entry is no good. And you can't see deep enough to see why it failed low. You repeat this the next iteration, and the next, ad infinitum. Because you can never search deep enough to see how bad it is, you only have the <Y.YY value and use that to trigger a fail-low each time you hit it...
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.