Author: Richard Pijl
Date: 12:22:44 08/02/04
Go up one level in this thread
On August 02, 2004 at 13:38:40, Jorge Pichard wrote: >On August 02, 2004 at 12:42:16, Jorge Pichard wrote: > >>On August 02, 2004 at 12:28:11, Jorge Pichard wrote: >> >>>On August 02, 2004 at 11:10:35, Anthony Cozzie wrote: >>> >>>>On August 02, 2004 at 09:23:03, Peter Fendrich wrote: >>>> >>>>>On August 02, 2004 at 09:18:37, Anthony Cozzie wrote: >>>>> >>>>>>Poor Baron. We all know that KBN-K arises in _every_ game. Why, not being able >>>>>>to win that endgame must be worth almost 0.0001 elo. >>>>>> >>>>>>anthony >>>>> >>>>>OTOH it's quite easy to implement so the "achieved elo"/"devolpment time" >>>>>is not that bad... >>>>> >>>>>/Peter >>>> >>>>Of course, but Jorge acts like the sky is falling. >>>> >>>>anthony >>> >>>It only take a GM to know the opponent weakness to force X program to accept an >>>exchange. For instance the GM could have a significant losing game by having a >>>Rook and a pawn versus the program with two knights and two Bishops. But if the >>>GM human player knows that the program doesn't know how to Mate with a KBN vs K, >>>all that he has to do is force the exchange of his rook and pawn for one of his >>>bishop and knight, and the rest will end in a 50 moves draw :-) >>> >> >>PS: There could be different possibilities with different pieces combinations. >>Another possibility could be the GM human having a bishop and three pawns versus >>the program with two bishops + Knight and three pawns. If the GM force the >>program to exchange its three pawns and one of its bishop for his three pawns and >>one of his bishop. Or 2nd escenario the GM with a knight and three pawns versus >>program X with two knights a bishop and three pawns, again as long as the human >>GM knows he can force the program to exchange his three pawns and a knight for >>his three pawns and his knight. > > > >Here is another very common position, if it is white to move after the black >pawn is taken at d5 by program X, all that a human GM has has to do is take the >piece at d5 and the final position will end up with a King, bishop and knight >vesus a lonely king :-) > >[D]8/5q2/3k4/3p4/5N2/1N6/1K4B1/2B5 w - - 0 1 After the correction I made last night, this position is no problem. I did a little shootout between Baron without EGTB's and Baron with them (to play the best moves with the lone king) with 5 minutes on a P4-2.4. Although it didn't find the mate in the optimal number of moves, it was well within the 50 moves (in fact, in 31 moves after the exchange). Richard.
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.