Author: Chessfun
Date: 16:22:38 11/23/99
Go up one level in this thread
On November 23, 1999 at 08:14:14, Mark Andreoli wrote: > >Hello Gentlemen; > > > This is a letter I posted to Heiko Mikala the other day!! It is exactly what >you're talking about! I'm not sure I can trust analysis anymore From Fritz 5 > (see below) especially after it beats itself using my move in the critical >position! It contradicts itself-see below comments will be greatly >appreciated!!! > > > >> >> Thanks so much for your expertise! anyways here goes with my >> comments! >> >> 1.e4 d6 2.d4 g6 3.f4 Bg7 4.Nf3 Nd7 5.Bd3 c5 6.c3 Ngf6 7.e5 >> Nd5 >> >> 8.OO cxd4 9.cxd4 dxe5 10.fxe5 OO 11.Qe1 Qb6 12.Kh1 Nb4 >> >> Ok here I was in my over the board game. I could chicken out and >> play Qe2 or go for the gold (or try anyways) with : >> >> 13. Bc4 !? (here F5 gives me a -2.00 afterall I am losing a rook) >> Nc2 >> >> 14. Qh4 h5 ( my opponent wants to prevent Bh6,Ng5 etc./He can >> always take the rook) 15. Nc3 Qb4 !? 16. Bd3 Nxa1 17. Nd5 Qa4 >> 18.Nxe7+ Kh8 19.Nxg6+ fxg6 20.Bxg6 Kg8 21. Qxh5 Rd8 22. Bh7+ Kf8 >> 23. Ng5+ Nf6 24. Qf7 mate... >> >> >> Ok, If you let engine play against engine (F5 vs F5) from the >> critical point in the game 13. Bc4 , Fritz 5 goes on to win against itself >> using my move with time controls of 40 moves @ 3 hours, 4 hours, or even >> 7.5 hours everytime!!! >> >> It starts playing a little different (try it) on move 21.....Rf4 >> 22.Bxf4 (I guess F5 plays Rf4 because he "sees the handwriting on the >wall" >> mate is inevitable! (but not a very human like move)) >> >> F5 also plays 15....Qb4 (but his advantage starts dissipating >> little by little in the F5 vs. F5 engine analysis displayed next to each >> move).I also played F5 vs F5 using the move 15....e6 to keep the knight >out >> of d5 and "you guessed it" white goes onto win again with all slow time >> controls! >> >> Then I "click" on coach and "click" on the analysis mode.I have F5 >> start analyzing my game from the last move to the beginning of the game. >> >> A funny thing happens even at settings of .1 to.5 pawns/60 seconds >> to 800 seconds, IT DOES NOT GIVE ANY COMMENTS OR VARIATIONS TO HELP OUT >> BLACK (OR WHITE) IN THE TEXT FROM MOVE 21 going backwards to MOVE 15. ??? >> what's going on? As white starts to win it should be complete with >> variations and comments? >> >> This seems very strange because it will always assign around a -2.00 >> for my 13th move Bc4 however when playing against itself at slow time >> controls it goes on to win with my move! >> You would think it would at least give a few variations for >> white/black when "I start turning the tide " in whites favor. BUT IT >> DOESN'T. So in essence it is saying that my move 13.Bc4 is a blunder (but >it >> is not because F5 beats F5 using my move) and it doesn't do any analysis >> going backwards for either side moves 21-15. >> >> So here you have it. Fritz 5 is contradicting itself! My 13th move >> is bad for F5 in all types of analysis but even F5 wins with it? (but why >> doesn't it give any comments in the analysis either going from move 21 to >> 15?? ) >> >> This is the first time Fritzie has let me down... I would really >> like to get to the bottom of this ? Maybe Hiarcs,Mchesspro, will analyze >my >> game move 21 to 15 better than Fritzie? That seems hard to believe. Thanks >> so much and let me know what you come up with ! Maybe my friend Fritz has >> been put in "TILT" :?) >> >> This is a simple depth problem clearly if Bc4 wins then the depth to see it is very deep, I personally have my doubts that Bc4 wins but without analysis as yet. This is also the reason for homework at GM level, that in analysis a move can be studied to extreme depths which at game time controls solutions are difficult to work out. This though does seem to me that black's position after Nc2 is very playable, please publish the game Fritz v Fritz from move 13. on. Thanks.
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.