Author: Howard Exner
Date: 19:16:27 04/17/98
Go up one level in this thread
On April 17, 1998 at 07:57:59, Djordje Vidanovic wrote: >On April 17, 1998 at 04:21:03, Howard Exner wrote: > >>I wonder if anyone can confirm or disprove my guess about Fritz 5's >>assessment of king safety / kingside attacks. From observing numerous >>games it seems that Fritz has an agressive approach to weakening the >>opponents kingside and to aggresively open lines for active piece >>play against the opposing king. I've seen it play many brilliant moves >>that quickly open lines, leading to mate or big material gains. I've >>also seen this tactic backfire also (as in the recent KKUP2 games >>against Rebel and CM5500 - the first game that was cancelled). Does >>anyone else see this as a style of Fritz's play/programming? If true >>is this unique to Fritz? > >I think that I can go along with what you said. Your observations >regarding the style employed by Fritz 5 correspond wholly to what I've >seen so far. I too believe that these two facets of chess-playing -- >kingside attacks and king safety -- were of paramount importance to >Frans Morsch. What more can one ask at the moment? It is therefore >rather silly to be hairsplitting about Fritz's chess-playing strength, >when against humans it is so indecently strong: the other night, a >friend, about two hundred points stronger than I, an IM (official ELO >2445), got butchered by Fritz 5 at 4 m per game with 2 sec increment: >17.5 -2.5. The very same IM lost a 4-game blitz match vs Seirawan >2.5-1.5 and should have easily drawn it, but for the flag. Regardless >of the thousands of games I have played against different programs I >still have many, many problems with Fritz 5. All this due to the two >well worked out parameters in the subject line of your posting and my >reposting. Perhaps the successfulness of the implementation of these >two invokes the green-eyed monster of jealousy in some? These coupled >with the almost perfect interface and ease of use make out of Fritz 5 >and its modules an awesome computer chess system. When this agressive style becomes further perfected it will truly be a dangerous weapon particularily against human opposition. It seems a bit like playing the odds. ie: To implement such code you will know that at times the program will blindly follow the wrong path but that in the majority of cases it will hopefully work out for the good. It looks like with Fritz that the decision works out more than it fails. My view of programs in general seem to comply with the majority here. Namely that all programs contain both strong and weak points and that their styles of play are unique. It always amazes me to view the results of a test suite and notice the variance of solution times for the different programs.
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.