Author: Heiner Marxen
Date: 10:40:44 05/21/03
Go up one level in this thread
On May 21, 2003 at 12:07:58, leonid wrote:
>On May 20, 2003 at 10:46:39, Art Basham wrote:
>
>>If it is so "easy" for computers to solve,
>>then why has no one posted that their program
>>announced mate in 10 at move one?
>>That is what i wanted to know...:-)
>
>On Celeron 600Mhz. LLchess mate solver by selective search, 10 moves deep, mate
>found in 0 seconds. This signify that time is less than 0.055 sec.
>
>Actual position have very good branching factor, even for brute force search. By
>brute force I went only 8 moves deep and it took for program less than 2 minutes
>to respond.
>
>Leonid.
>
>
>>Thanks anyway...!
>>art
>>
>>
>>On May 19, 2003 at 11:01:52, Roy Eassa wrote:
>>
>>>On May 19, 2003 at 10:18:55, Art Basham wrote:
>>>
>>>>[D]3r3k/p4Qp1/7p/1b1Bp3/4Pp2/7R/4nqPP/4R2K w - -
>>>>
>>>>Nice ending...! See if your best program will ann. mate at move one!
>>>
>>>
>>>I think this is a good position for a HUMAN to test him/herself with.
>>>
>>>I think it's too easy for most modern programs/PCs.
>>>
>>>I enjoyed solving it myself. :-)
Even a full brute force search is feasable here: Chest solves it in 19.7 mins
on an Athlon 1500+ (128 MB hash):
PV: Rxh6+ gxh6 Qf6+ Kh7 Qe7+ Kh8 Qxd8+ Kg7 Qg8+ Kf6 Qf8+ Kg5 Qg7+ Kh5 Bf7+ Kh4
Qxh6+ Kg4 Be6#
Rxh6+ gxh6 Qf6+ Kh7 Qe7+ Kh8 Qxd8+ ...+6
Kg6 Qf7+ ...+4
Interestingly, the last level ate most of the time:
# 5 0.01s 2kN [ 2.49] 1.37 175- 0
# 6 0.03s [ 3.00] 6kN [ 3.36] 1.73 367- 0
# 7 0.13s [ 4.33] 28kN [ 4.79] 1.81 2153- 0
# 8 1.54s [ 11.85] 324kN [ 11.72] 1.90 24634- 0
# 9 14.33s [ 9.31] 3246kN [ 10.02] 2.32 241594- 0
# 10 1180.25s [ 82.36] 259973kN [ 80.08] 2.90 17547590- 14348357
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.