Author: Aaron Tay
Date: 09:18:54 08/27/00
Go up one level in this thread
On August 27, 2000 at 09:28:05, Uri Blass wrote: >On August 27, 2000 at 09:10:57, Aaron Tay wrote: > >>On August 27, 2000 at 08:28:33, Uri Blass wrote: >> >>>On August 27, 2000 at 07:29:46, Aaron Tay wrote: >>> >>>>On August 26, 2000 at 21:39:27, Jorge Pichard wrote: >>>> >>>>>On August 26, 2000 at 18:17:49, Alexander Kure wrote: >>>>> >>>>>>On August 26, 2000 at 15:20:44, Sandro Necchi wrote: >>>>>> >>>>>>[...] >>>>>> >>>>>>>>Which program had the best opening book? Shredder 5.0 >>>>>> >>>>>>[...] >>>>>> >>>>>>Can't be, as Shredder had lost positions out of the opening against Rebel and >>>>>>Nimzo. >>>>>> >>>>>>Greetings >>>>>>Alex >>>>> >>>>>Two mediocre opening lines out of thousands of variations which is probably a >>>>>least an average of 8 variation to every opening lines is a small percentage, >>>>>which I' am sure that it will be fixed before the final release of shredder 5. >>>>>Now mention just one program that does not have any dubious lines ? >>>>> >>>>>Pichard. >>>> >>>>Hmm I suppose one of those programs that don't have any opening book at all! >>> >>>I think that shredder falled in the trap against nimzo because it had no >opening book in the relevant line. >> >>True. But it was the chess engine's fault for falling into the trap. Not because of a bad opening line! >>Say my opening book consisted only of one line 1 e4! >>After which my program loses to some "cooked" opening line ? Would you claim >>that 1 e4 is a dubious line? > >Yes. > >1.e4 without more moves is a dubious line for the engine if it falls into a >prepared trap after this line. >There are deep trap in the opening that programs cannot avoid in practical >games and if you play with no book or witha very small book your program can >fall into one of them, >Uri Hmm. I guess we have different definitions of "dubious lines" For me the line 1 e4 e5 2 Nf3 Nc6 3 Bb5 is not a dubious line. Say Crafty has only the above line. And nothing else. You seem to imply that the ruy lopez is a dubious line if Crafty(picking a example) on it's own (without opening book) later falls into the Noah Ark's trap for example.. I suppose you could blame the author of the opening book. But do you seriously expect the guy preparing the opening book to prepare for every possible trap? You could say he didn't do his home work, but to say that the ruy lopez is a dubious line is ridicious! So basically as far as i can see there are serious type of "opening book error" TYpe 1 - This is where the stored move is simply wrong. Usually the stored move is a tactical blunder that if the program was left to think on it's own, it could avoid.. This type of error, is relatively easy to check, there is the CAP project, Rebel allows you to analyse the opening book for errors etc etc Type 2 - Here the stored move is wrong because of positional reasons or it could be tactical beyond the range of the programs. Type 3 - This is when the opening book leads to a perfectly sound position. Unfortunately, the program just doesn'n know how to play it.. It doesn't lose straight away but drifts into a losing position.. Type 4 - As above, except that, this time because the opening book ends too early, and because the program doesn't know how to handle it, it falls into a opening trap set by another program with a deeper book. Are the "o called poor book opening errors"in the recently concluded WMCCC mainly of type 4?
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.