Author: Dann Corbit
Date: 14:21:24 11/27/01
Go up one level in this thread
On November 27, 2001 at 17:15:07, ALI MIRAFZALI wrote: >Earlier Mr Nemeth had the following win against FRITZ6 >1.e4 d5 2.exd5 h5 3.Bb5+ c6 4.dxc6 Nxc6 5.Nf3 Nh6 6.Nc3 g6 7.0-0 Bg4 8.h3 e6 >9.hxg4 hxg4 10.Ne5 Qh4 11.g3 Qh3 12.Nd5 Ng8 13.Bxc6+ bxc6 14.Nc7+ Ke7 15.Nd5+ >cxd5 16.Nxg6+ fxg6 17.d4 Qh1#. > The problem with this game is move number 6;that is it impossible to >get Fritz6 to actually play 6.Nc3 ;it always plays 6.d4.Since Machines are >very deterministic ;this kind of testing remains valid. False. Every good chess program has randomness built into it. They also have book learning. Very likely the Fritz he was using learned to avoid that choice for some reason. Every machine is different, every time control is different, every version of the program is different. >There was also another >game posted which ended in a win for black after move 23.It began with >1.e4 d5 2.exd5 h5 3.Bb5+ c6 4.dxc6 Nxc6 5.Nf3 g6 6.0-0 .Once again it is >impossible to see how fritz6 chooses 6.0-0 since 6.d4! is the move that is >actually played. In your setting. You seem to be under the dillusion that chess programs are like light-switches. Does your chess program play 1. e4 every single time? I'll bet it does not. So if that can change, what about other moves? >It is intersting to note that many other software programs >also choose 6.d4 in this particular position.However the most shocking example >in my opinion is the game given the Title "THE MORNING DANCE" it is played >at a time control of Blitz in 30 for each side.It begins with >1.e4 e5 2.d4 exd4 3.Qxd4 Nc6 4.Qe3 Nf6 5.Nc3 Bb4 6.Bd2 0-0 7.Bc4 Bxc3 >8.Bxc3 Nxe4 9.0-0-0 Nxc3 10.Qxc3 Qg5+ 11.Rd2 Qxg2 12.Bd5 Qf1+ 13.Rd1 Qxf2 >14.Nh3 Qf5 15.Rhf1 Qe5 16.Qc4 Qe3+ 17.Kb1 Qxh3?? (black lost on move 26) >17....Qxh3??? is aBlunder the kind that bug free programs dont make. Every program has bugs. >I do not >own fritz7 so a direct testing like in the previous games was not available to >me. In other words, you cannot reproduce his games. >The position was tested with the program ZARKOV 2.6 an old program from >1990 A program Vastly inferior to any version of Fritz (Current ssdf rating >only 1833) .Not even for a SPLIT SECOND does the program even Consider >17...Qxh3?? (the same kind of results were also obtained for other programs >too numerousto list here).About another game that begins with >1.d3 d5 2.c3 e5 3.g3 Nf6 4.h4 Bd6 5.Bg2 0-0 6.Nh3 c6 7.Bg5 h6 8.e4 hxg5 >9.hxg5 Bg4 10.f3 Bxh3 11.Rxh3 Nfd7 12.f4 Qb6?? perhaps silence is the best >response ;but let me give one example:Chessmaster 8000 considers 12...Qb6 >for exactly 1 second before switching to 12...g6.It is Obvious that no >serious chess program would make a move like 12...Qb6?? in an actual game. Before you make a serious accusation and call someone a liar, I suggest that you get your facts straight. You have no idea what you are talking about. I have definitely seen top chess games blunder. Commical, whimsical, absurd blunders. If you look over enough chess games (especially at fast pace) you will find plenty yourself. There seems to be some problem with the F7 MMX version. I think it may be buggy.
This page took 0.01 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.