Author: leonid
Date: 15:47:16 07/29/01
Go up one level in this thread
On July 29, 2001 at 15:29:36, Heiner Marxen wrote: >On July 27, 2001 at 21:01:50, Heiner Marxen wrote: > >>On July 27, 2001 at 19:44:05, Uri Blass wrote: >> >>>On July 27, 2001 at 19:39:35, Uri Blass wrote: >>> >>>>On July 27, 2001 at 10:53:28, leonid wrote: >>>> >>>>>Hi! >>>>> >>>>>If you are ready, you can solve this problem: >>>>> >>>>>[D]1N1k1N1B/q1q1q1q1/1rRbRr2/Q1nBn1Q1/1Q3Q2/n1Q1Q1n1/1Q1Q1Q2/n2K2nb w - - >>>>> >>>>>Please indicate your result... even if your program fail. >>>>> >>>>>Thanks, >>>>>Leonid. >>>> >>>>chesmaster6000(PIII850 ss=10) >>>>solution Rexd6+ >>>>3:56 mate in 11 >>>>5:47 mate in 10 >>> >>>and after 10:49 Rxc7 is mate in 9 >>> >>>Uri >> >>Chest has "problems" with this problem: to find "no mate in 7" needed already >>103 minutes. With an effective branching factor of 15+ the next depth will >>need more than a full day. :-( >>If there is no mate in 8 it is out of reach for Chest. >> >>To be continued... :-) >> >>Heiner > >Ha! According to Chest this is a mate in 8 :-) >But what a huge effort to spend: 162992 sec = 2716.5 min = 45.3 hrs = 1.9 days >on a K7/600 with 350 MB hash. For a mate in 8 is kind of a record. Thanks, Heiner! Mine had so bad branching factor here that it had no chance for going so far. Even 5 moves already took 3 min and 24 sec. It is great for me to know this position minimal number for possible future verification. My old positions (positions that I wrote when I did my mate solver) are already useless. They see very often mate between 5 and 7 moves deep. Then it was long search for mine old 80386 25Mhz. Cheers, Leonid. >PV: Qexc5 Bf3+ Bxf3 Ne4 Rcxd6+ Qed7 Rxd7+ Qgxd7 Bxf6+ Nxf6 Qfxf6+ Kc8 Qd8+ Qxd8 >Qgxd8# > >Surprisingly, the key move is not a checking one. It is a capture move, >but it does not even take one of those queens, but rather a knight. :-O > >Compared with other problems, the effective branching factor is quite bad: > > depth time EBF[T] EBF[N] hash-factor ># 1 0.00s 0kN 0.87 1- 0 ># 2 0.00s 0kN 1.00 1- 0 ># 3 0.02s 1kN [ 20.33] 0.95 99- 0 ># 4 0.61s [ 30.50] 33kN [ 53.84] 1.05 1640- 0 ># 5 18.04s [ 29.57] 888kN [ 27.04] 1.23 76121- 0 ># 6 359.95s [ 19.95] 17453kN [ 19.66] 1.57 1623694- 16 ># 7 6184.85s [ 17.18] 319730kN [ 18.32] 1.94 26133013- 17385126 ># 8 162992.70s [ 26.35] 4618561kN [ 14.45] 1.91 673926679- 665178778 > >Also, the speed up by the hash table could be better. With a recall rate >of 7.5% the factor stays below 2. >Sort of a "killer problem" for Chest. Wow! > >Cheers, >Heiner
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.