Author: Heiner Marxen
Date: 11:35:47 05/19/01
Go up one level in this thread
On May 19, 2001 at 08:13:18, leonid wrote: >On May 18, 2001 at 23:56:01, Heiner Marxen wrote: > >>On May 17, 2001 at 13:34:19, leonid wrote: >> >>>Hi! >>> >>>If you would like to solve one mate then try this: >>> >>>[D]qr3rqB/3k3B/N1qqq1Nb/Q1n1n2Q/1qQ1Qq1Q/2RPR2Q/Q2K3Q/1q3q1b w - - >>> >>>Please indicate your result. >>> >>>Thanks, >>>Leonid. >> >>Hmm, well. Since you didn't say that it is "very easy", not even mentioned >>a simple "easy", I should have expected that: there is no mate in 10. >>Chest on K7/600 with 350 MB hash needed 1.8 hours for this. The effective >>branching factor from depth 9 to 10 is around 6, so I'm not sure that >>I can go further than depth 11. To be continued... > >Hi! > >I already wrote my message when my connection was broken. Now will be very >brief. This is position in 11 since you looked alread all 10 moves. Yes, according to Chest you are right. On K7/600 with 350 MB hash, burning 49980.47 sec (833.0 min) (13.9 hrs) Chest finds a unique key move. Here is just the PV: Nxc5+ Qbxc5 Nxe5+ Qfxe5 Qhxe6+ Qdxe6 Qxe6+ Qcxe6 Qc7+ Qcxc7 Rxc7+ Kxc7 Q5xe5+ Qxe5 Qhxe5+ Kb6 Q5d4+ Kb5 Qd7+ Kb6 Bd4# >I went only first 8 moves and it took 1 hour 56 min. My branching was not that >good as your. It was very stable between 4 and 7 moves and was 9.5. Only between >7 and 8 it fell slightly to 8.3. Chest timing for branching factor: # 3 0.03 0.96 76- 0 # 4 0.20 1.13 653- 0 # 5 1.44 1.35 3947- 0 # 6 9.30 1.63 26007- 0 # 7 68.47 2.07 209752- 0 # 8 296.66 2.82 917079- 0 # 9 1097.25 4.13 3622460- 3426 # 10 6444.29 5.03 25330293- 16582427 # 11 48443.30 5.45 202160784- 193412883 4-> 5: 7.200 5-> 6: 6.458 6-> 7: 7.362 7-> 8: 4.332 8-> 9: 3.698 9->10: 5.873 10->11: 7.517 Beetween 7 and 8 my BF is also much smaller. At the end it rises, again, what is not uncommon near a solution. Overall my BF is much smaller than 9.5. >Will try today to read somewhat from your program description for fun. Yesterday >when I went to read few lines (it is not code but description that I am looking >for) I had the impression that it was written on other planet. Nothing in >commune! I hope you like reading from another planet :-) > Nevertheless, very often I have the impression that our programs react >in similar way for similar positions. Sure, basically both are mate provers, so they have a lot in common, regarding the functionality (even if the sources do not look similar). Have a nice weekend! 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.