Author: Robert Henry Durrett
Date: 13:27:24 06/03/02
Go up one level in this thread
On June 03, 2002 at 14:46:26, Jorge wrote: >On June 03, 2002 at 14:29:25, Dante Rosati wrote: > >>some claim that the lunsen opening book is superior to powerbook. Is there any >>basis for this? Has anyone run games with the two opening books against each >>other? >> >>Dante > >I have not done any testing with the Lunsen book, but I will say this much, a >powerbook is over 500 MBs, which is huge. I use powerbook quite frequently in >comp to comp matches. They usually last over 20 moves. IMO, I don't think Lunsen >book is better. > >regards, >jorge If Fritz Power Book [PB] is "huge" when compared to Lunsen's Book [LB], then PB is probably going to be preferred by most, just because of the size difference. If we can get past that, then what questions remain to be considered? Maybe one question of interest would be whether or not lines in LB which are not in PB might be superior in some sense. To answer that question, a lot of work would still need to be done to get a definitive answer, but the task might not be as big a task as might have been had the two books been approximately the same size. This may be easier to discuss this topic if the equivalent game collections are discussed. For each opening book there is an equivalent collection of lines which can be created from the book. With CB8, it just takes a few punches at the keyboard to produce the equivalent collections of lines. [Some people call these truncated games.] To make this perfectly clear, the collection of lines would produce the **original** book if used later for making a book. At first thought, it might be assumed that what we would be looking for is all of the LB lines which are not in PB. Upon reflection, it is seen that what we really need is a collection of all positions where LB deviates from PB. There may be some very elegant way to get this information quickly by using CB8 or some equivalent chess database manager. Unfortunately, I don't know any elegant ways. The way I see it, it's going to take some good old-fashioned "grunt work." Here's one way that the collection of positions could be obtained. [Please, everybody, don't groan all at once when you see I'm using a spreadsheet to do this task.] The LB collection of lines, i.e. truncated games, could first be entered into the spreadsheet and all made bold red. [I am leaving out steps. Please forgive. Each move requires one cell.] Then the PB collection of lines would be entered into that same spreadsheet, but all in black unbolded. Then all of the lines are sorted in alphabetical order. Easy to do in a spreadsheet and impossible to do in CB8. One would then use highlighting to mark the desired positions. [Actually, highlight the LB move which comes after the position.] Then delete all rows which are irrelevant. These rows to be deleted would be of two types: First, delete LB lines which are duplicated by PB lines. Then delete all remaining irrelevant PB lines. I will stop here, since it should be easy to visualize what remains to be done. :) [If you can play blindfold chess, then you should be able to see this spreadsheet.] Anyway, as noted by other bulletins, reaching a verdict as to whether an LB move is better than a PB move is going to be non-trivial. Have fun! Bob D.
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.