Author: Dann Corbit
Date: 17:44:15 08/26/02
Go up one level in this thread
On August 26, 2002 at 20:36:20, Mike Byrne wrote: >On August 26, 2002 at 20:34:38, Mike Byrne wrote: > >>On August 26, 2002 at 19:38:19, Arturo Ochoa wrote: >> >>>On August 26, 2002 at 19:11:27, Mike Byrne wrote: >>> >>>>On August 26, 2002 at 17:48:27, Arturo Ochoa wrote: >>>> >>>>>On August 26, 2002 at 17:45:04, Mike Byrne wrote: >>>>> >>>>>>I would like to see two different programs with and without opening books. I >>>>>>think the difference between unlike programs would be less. >>>>>> >>>>>>Mike >>>>> >>>>>Hello: >>>>> >>>>>Not at all. What it is being proved here is how a chess program can increased >>>>>its level using a book. >>>>> >>>>>The most realiable prove is doing it against the same engine without book. >>>>> >>>>>A mixture of different engine doesnt prove anything. >>>> >>>>That's not true , I just played Crafty and Yace both with books in a 100 game >>>>match - ..I can play one against the other - one without book - that would give >>>>me some data on the impact of books. >>>> >>> >>>The problem here is what Engines to choose in order to get reliable data. >>> >>>Engine plus Book plus Engine without Book against what? >>>Amateurs? Which Amateurs? All the Amateurs? >>>Comercials? Which Comercials? All the Comercials? >>> >>>If you dont choose all the programs, why would you choose Crafy Vs Yace instead >>>Crafty Vs Fritz or Gandalf? >>> >>>The idea of to Match the Engine + Book Vs the Engine without book is to validate >>>how the Engine increased its level with an opening book. >>> >>>It can happen that Yace without Book can easily beat Crafty withtout Book but >>>what does it prove? >> >>That's my point, you not even reading what I am saying - I just played 100 games >>both with book, now I (could) play Yace WITHOUT book vs Crafty WITH book, then >>flipflop, play crafty WITH book and yace WITHOUT book - don't try to read in >>between the lines, just read the lines and you'll be fine. > >of course in my haste , I errored in typing my thought - the above shoudl read: > >That's my point, you not even reading what I am saying - I just played 100 games >both with book, now I (could) play Yace WITHOUT book vs Crafty WITH book, then >flipflop, play crafty WITHOUT book and yace WITH book - don't try to read in >between the lines, just read the lines and you'll be fine. But what has been shown until you have: Cr+book verses Ya+book Cr-book verses Ya+book Cr+book verses Cr-book Ya+book verses Ya-book ? The reason I ask is: Any of the books can be defective. Hence, it is going to take a big pile of games before you can find something out. However, if you play an individual program with its book against the exact same program but without its book, you will eliminate all except one variable. Naturally, my experiment answers only one question: Under the conditions of the experiment, does program X play better against itself with or without a book? The question you answer also includes as many other programs as you try. Both questions are interesting.
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.