Author: Martin
Date: 04:26:29 02/19/00
Go up one level in this thread
On February 18, 2000 at 17:51:41, John Merlino wrote: >> >>1 Rb5+ Nb6 (Ka7 is mate in 2) 2 Bxb6 Rh2+ >>3 Kg4! (not Kg3) Rxh1 4 Bg1+ Kc7 5 Rb1! >>The bR is caught. Advancing the h pawn >>only brings the wK closer to the rook. >>5... Kd7 6 Re1! >>Now the bK gets short of white squares. >>6... Kc6 7 Rd1 Kb5 8 Rc1 Ka4 9 Rb1. >> >>>Did your program find all moves? >>> >>>Martin > >While I agree that most programs will not see the solution in a reasonable >amount of time, I believe they WILL find it given ENOUGH time. Chessmaster, for >instance, WILL see the solution (within a few seconds) if you plug in the first >few moves for both sides. Does it really "see" the solution? An advantage of 4-5 pawns (it's the material advantage of just one piece) shows that it doesn't. The plus score should go beyond +7. From the above position, however, a typical chess >program will need to calculate (in my estimation) about 12 or 13 plies before >discovering the winning combination. This is quite a lot of positions to see, >given all of the pieces still on the board. Yes, I agree, if the task is to go deep enough, there are certainly many positions to calculate. But to see that the rook gets caught requires 9 plies. So it's also a question of "knowledge"... > >jm > >p.s. I'm letting CM6000 run on the position at this time. It took 19 minutes to >get to a depth of 9/11 (and still prefers 1.Bd5, as it did from the outset), so >it could be QUITE some time before it finds the win. I'll let you know the >results if I have enough patience.... ;-) (-: Fritz also prefers Bd5 for a long time; Hiarcs recommends Rb5+ immediately but without seeing the real advantage (see above). Martin
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.