Author: A. Steen
Date: 20:12:44 11/20/05
Go up one level in this thread
On November 20, 2005 at 22:54:06, Timothy J. Frohlick wrote: >Mr. Steen, > >Fruit is a good program. Chess is a game based on deception. Whoever can lure >the opponent into a bad position will win. Fruit is good at being a deceiver. >Is Fruit demon-possessed? No! Fruit does spice up the playing of chess and is a >most entertaining product. I still wish that Chessbase would include a special >Fritz/Fruit edition on DVD in the near future....as the spanish would say >Fruit y Fritz. > >TJ Mr TJ, Fruit is an excellent program against other engines. Engine vs engine, I think it is the strongest one available to the public, no exceptions. I don't know if I have given some other impression (surely not?). But the "luring" strategy would only work to explain the pattern OVER DIFFERENT GAMES (not in the same game, that is a different matter and I readily concede that Fruit is a lurer!) if the other engine too had some form of opponent detection/learning coding in it that operated between games. Right? :) I don't think that is the case. :) So, what is going on? All these Fruit vs EngineX matches which start with something like Fruit 2.2.1 0½01½0½10½0 and then the tide turns the other way. Please research the Fruit 2.2.x results (against S9/F9 competition) posted here in the last month, you will find few exceptions! Wven my Fruit vs F5.32 matches started off the same way (F5.32, for what it is, does well against Fruit at fast (at most 10+5) TCs - scores almost 40%). Best, A.S. >On November 20, 2005 at 22:45:33, A. Steen wrote: > >>All but one engine match (9 so far) I have played involving FRUIT 2.2/2.2.1 and >>a near competitor has resulted in FRUIT starting off badly and then catching up >>(and surpassing) later. >> >>This also seems to apply to almost all engine matches played by others reported >>here in talkchess.com, even in the slow TC SSDF ones reported here by Mr >>Hedlund. >> >>So it isn't just my trial FRUIT, or something funny about my setup. >> >>The thread just below this one ie, >>http://www.talkchess.com/forums/1/message.html?463252 >>has another example of what I am talking about. >> >>1 Fruit 2.2.1 0½0½0½½011111½½½½1½1½½10½010½0½½1½1½½½11 22.5/40 >>2 Fritz 9 1½1½1½½100000½½½½0½0½½01½101½1½½0½0½½½00 17.5/40 >> >>Is it a coincidence? The numbers aren't big enough to prove anything, >> >>Or is there some subtle learning going on, RAM only, nothing on file in the >>FRUIT directory? So when a new match begins after a shutdown of the chess GUI >>or engine or computer, it starts off afresh (or assumes a new opponent?) >> >>Please tell me I am talking "Nonsense".... could Fruit 2.2.x be already doing >>some form of "opponent-detection" that is mooted for a future Fritz? >> >>If so, FRUIT is even fruitier than I thought! :) >> >>Best, >> >>A.S.
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.