Author: Drexel,Michael
Date: 11:30:05 10/18/05
Go up one level in this thread
On October 18, 2005 at 12:37:47, George Tsavdaris wrote: >On October 18, 2005 at 12:11:56, Drexel,Michael wrote: > >>[D]1n1r3k/1p1n4/2p5/8/3B4/8/4K3/R7 b - - 0 1 >>This is not supposed to be the Giveaway chess engine :) >> >>The endless number of Bugs in the GUI are not worth mentioning of course. > >Good discovery! The bug seems to be deeper i suspect, since Fritz 8 (i haven't >analysed with 9 because i don't have it) evaluates black's position as slightly >better -actually as almost drawish) even after Kg8 or Kh7. My Fritz 8 (8.0026) has no problem at all: 1n1r3k/1p1n4/2p5/8/3B4/8/4K3/R7 b - - 0 1 Analysis by Fritz 8: 5...Kh7 6.Kd3 c5 7.Th1+ Kg6 8.Tg1+ Kf5 9.Tf1+ Ke6 10.Te1+ Kd5 -+ (-4.18) Depth: 7/18 00:00:00 53kN 5...Kh7 6.Kd3 c5 7.Le3 Se5+ 8.Kc2 Sa6 9.Th1+ Kg6 10.Lf4 -+ (-4.28) Depth: 8/19 00:00:00 195kN 5...Kh7 6.Kd3 c5 7.Lb2 Se5+ 8.Ke4 Sd3 9.Lf6 Td6 10.Lc3 -+ (-4.34) Depth: 9/25 00:00:00 525kN 5...Kh7 6.Kd3 c5 7.Lb2 Se5+ 8.Kc2 Sbc6 9.Lxe5 Sxe5 10.Te1 Sc6 11.Kc3 -+ (-4.28) Depth: 10/28 00:00:00 997kN 5...Kh7 6.Kd3 c5 7.Lb2 Se5+ 8.Ke4 Sc4 9.Th1+ Kg6 10.Tg1+ Kf7 11.Tg7+ Ke6 12.Tg6+ Kd7 13.Lc3 -+ (-4.43) Depth: 11/41 00:00:02 2951kN 5...Kh7 6.Kd3 c5 7.Lb2 Se5+ 8.Kc2 Sd3 9.Th1+ Kg6 10.Tg1+ Kf5 11.Tf1+ Ke4 12.Lf6 Tc8 13.Kc3 c4 -+ (-4.47) Depth: 12/39 00:00:06 7664kN 5...Kh7 6.Kd3 b5 7.Kc3 c5 8.Le3 Te8 9.Th1+ Kg7 10.Tg1+ Kf7 11.Lxc5 Sxc5 -+ (-4.50) Depth: 13/45 00:00:22 26161kN 5...Kg8! -+ (-4.53) Depth: 13/45 00:00:35 41905kN > So i suspect Fritz 9 would evaluate Kg8 or Kh7 as around 0.38 so it plays Ne5 >as this looks to Fritz as drawish too. That means it doesn't understand the >position. What if you force the moves 1...Kh7 2.Rh1+ Kg6 3.Rg1+ Kf7 4.Rf1+ Ke6 >+- >Does Fritz 9 shows now a big advantage for black? If my suspicions are correct >and with a comparison with F8 evaluation, i believe that even in this position >that occured, Fritz 9 will believe that it is still a draw. > > So now Frans Morsch has to try to understand what is the "problem" behind this >position, because as such position exists, there may be (many) others with the >same bad behaviour for Fritz 9 and this is the actual problem..... Of course there are many other such positions. For example: [D]3r3k/2pp4/1pn5/8/8/8/4KB2/R7 w - - 0 5 Black has a nice endgame. Analysis by Fritz 9: 5.Kd2 Tg8 6.Kc3 Tg2 7.Ta2 Kh7 8.Kc4 Se5+ -+ (-2.43) Depth: 7/18 00:00:00 84kN 5.Lg3! -+ (-2.42) Depth: 7/18 00:00:00 96kN 5.Lg3 d6 6.Tc1 Sd4+ 7.Kd3 c5 -+ (-2.33) Depth: 7/18 00:00:00 103kN 5.Lg3 Te8+ 6.Kd2 Se5 7.Kc3 d6 8.Ta2 c5 9.Th2+ Kg7 10.Tg2 -+ (-2.52) Depth: 8/21 00:00:00 228kN 5.Kd2! -+ (-2.52) Depth: 8/22 00:00:00 343kN 5.Kd2! -+ (-2.36) Depth: 8/22 00:00:00 365kN 5.Kd2 Tg8 6.Th1+ Kg7 -+ (-2.34) Depth: 8/22 00:00:00 388kN 5.Kd2 Tg8 6.Th1+ Kg7 7.Tg1+ Kf7 8.Txg8 Kxg8 9.Kc3 d6 10.Kc4 Se5+ 11.Kc3 -+ (-2.44) Depth: 9/21 00:00:00 523kN 5.Kd2 Se5 6.Th1+ Kg7 7.Kc3 c5 8.Tg1+ Kf7 9.Tf1 -+ (-2.54) Depth: 10/24 00:00:01 1670kN 5.Lg3! -+ (-2.53) Depth: 10/31 00:00:02 2300kN 5.Lg3 d6 6.Tc1 Sd4+ 7.Kd3 c5 8.Lh4 Tb8 9.Lf6+ Kh7 10.Lxd4 cxd4 11.Tc7+ Kg6 12.Kxd4 -+ (-2.44) Depth: 10/31 00:00:02 2491kN 5.Lg3 d6 6.Tc1 Sd4+ 7.Kd3 c5 8.Lh4 Tb8 9.Lf6+ Kh7 10.Lxd4 cxd4 11.Tc7+ Kg6 12.Kxd4 -+ (-2.44) Depth: 11/35 00:00:04 4756kN 5.Lg3 Tc8 6.Tc1 Kg7 7.Kd3 Ta8 8.Lxc7 -+ (-2.48) Depth: 12/40 00:00:10 13079kN 5.Lg3 Tc8 6.Tc1 Kg7 7.Kd3 Kf7 8.Kd2 Ke6 9.Te1+ Kd5 10.Kc3 d6 11.Td1+ Ke6 -+ (-2.48) Depth: 13/48 00:00:23 29996kN 5.Lg3 Tc8 6.Tc1 Kg7 7.Kd3 Kf7 8.Kc4 d6 9.Kd5 Se7+ 10.Kc4 Tg8 11.Lf4 Sg6 -+ (-2.56) Depth: 14/49 00:00:55 72318kN You would think it is even better to have an extra bishop in the position: [D]3r3k/2pp1b2/1pn5/8/8/8/4KB2/R7 w - - 0 5 Analysis by Fritz 9: 5.Kd2 Sd4 6.Ta6 Te8 7.Lg3 Te2+ 8.Kd1 d6 µ (-0.80) Depth: 7/18 00:00:00 183kN 5.Kd2 Sd4 6.Ta7 Sb5 7.Ta6 d5 8.Le3 d4 µ (-0.82) Depth: 8/20 00:00:00 348kN 5.Kd2 Sd4 6.Ta7 c5 7.Le3 b5 8.Kd3 Kg7 9.Tb7 µ (-0.86) Depth: 9/22 00:00:00 878kN 5.Kd2 Se5 6.Ta7 Sc4+ 7.Kc3 c5 8.Lg3 d5 9.Tb7 Kh7 10.Lf4 µ (-0.90) Depth: 10/24 00:00:03 3253kN 5.Kd2 Se5 6.Ta7 Sc4+ 7.Kc3 c5 8.Lh4 d5 9.Tb7 Kg7 10.Lg5 Kg6 µ (-0.94) Depth: 11/26 00:00:07 8182kN 5.Kd2 Se5 6.Lg3 c5 7.Lf4 d5 8.Ta7 Kh7 9.Tb7 Tb8 10.Tc7 Kg6 µ (-0.99) Depth: 12/35 00:00:25 29673kN Surprise,surprise Now the position is rather drawish. We humans can still learn a lot from the silicon monsters ;) Michael >After 1...Kh7 2.Rh1+ Kg6 3.Rg1+ Kf7 4.Rf1+ Ke6 +- >[D]1n1r4/1p1n4/2p1k3/8/3B4/8/4K3/5R2 w - - 0 5 > >> >>Michael
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.