Author: Ricardo Gibert
Date: 14:17:59 04/24/05
Go up one level in this thread
On April 24, 2005 at 17:01:11, John Merlino wrote: >On April 24, 2005 at 16:49:55, Ricardo Gibert wrote: > >>On April 24, 2005 at 16:45:12, John Merlino wrote: >> >>>On April 24, 2005 at 16:40:26, Ricardo Gibert wrote: >>> >>>>On April 24, 2005 at 16:28:06, Vincent Lejeune wrote: >>>> >>>>>On April 24, 2005 at 15:05:41, John Merlino wrote: >>>>> >>>>>>On April 24, 2005 at 14:31:52, Jack Lad wrote: >>>>>> >>>>>>>[d] 6nr/p2rkp1n/1pQRq1pp/1N2P3/8/4BN1P/PPP2PP1/2K4R w >>>>>>> >>>>>>>1.Nc7 or Nfd4? >>>>>>> >>>>>>>Or can they both mate in the same number of moves? >>>>>> >>>>>>Playing through the lines, it appears that: >>>>>> >>>>>>1.Nc7 is a Mate in 9. >>>>>> >>>>>>1.Nfd4 is a Mate in 11. >>>>>> >>>>>>jm >>>>> >>>>>3 best moves from ChessMaster 10 (AMD@3000+) >>>>> >>>>>Time Depth Score Positions Moves >>>>>9:38 1/13 Mate09 233251425 1.Nc7 Rxc7 2.Qxc7+ Kf8 3.Rxe6 Kg7 >>>>> 4.Nh4 Ng5 5.Bxg5 Ne7 6.Bf6+ Kh7 >>>>> 7.Qxe7 Rf8 8.Qxf8 fxe6 9.Qg7# >>>> >>>>You need to be careful with 1.Nc7. If CMX is anything like Ruffian, it will get >>>>a mate score, lose it, get it again only to lose it again while shifting from a >>>>best response to it of Rxc7, Nf8, Qf5 and Ngf6 all the while. Ruffian had a >>>>tough time with 1.Nc7. It took 5 times longer to reach the same depth of 14 that >>>1.Rxe6 got and ultimately lost the mate score. >>> >>>Chessmaster will never lose a mate score for the side to move. Once it reports a >>>mate, the only thing it can do is possibly report a shorter mate. >>> >>>Therefore, my "playing with the lines" (which you comment on in another post) >>>was merely to ensure that there was no shorter mate than the ones that The King >>>found relatively quickly. >>> >>>jm >> >>Okay. I did not know that CM enforced the property of never losing a mate score. >>Interesting. > >In my opinion, no engine should report a mate score and then say "oops, I was >wrong, sorry...." :-) > >jm I agree it is a nice property, but rather difficult to maintain if a program makes agressive use of forward pruning. The TT can cause problems too.
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.