Author: Michel Langeveld
Date: 12:01:46 10/05/99
Go up one level in this thread
It's not my habit to post message like this, but have to say: BRAVO!!! In theory Crafty 17 with the CAP data will be the best crafty of ever! By the way: Is it possible to let crafty analyze and ignore certain kind of moves?? has this to be done with adding values in the hashtable? Kind regards, Michel Langeveld On October 05, 1999 at 12:02:05, Robert Hyatt wrote: >On October 05, 1999 at 11:19:57, James Robertson wrote: > >>On October 05, 1999 at 10:37:30, Robert Hyatt wrote: >> >>>On October 05, 1999 at 07:25:44, Steve McRiley wrote: >>> >>>>I see that BOB uses v. 17.0 at ICC now. What caused the jump from v 16.x? >>>> >>>>/McRiley >>> >>> >>>The most substantial change is a new book format to take advantage of the >>>C.A.P. data from Dan. I don't know exactly how this data is going to work >>>out in helping with the opening book, but the book format changed to make >>>room for the new data. Rather than have 16.19 use the old format, and 16.20 >>>use the new, and continually fill my mailbox with questions about why my >>>book won't work? I am just changing the major version number to solve the >>>problem. :) >>> >>>There are other changes as well, including the new pawn majority/candidate >>>stuff a particular GM has been harping on for several months. :) >>> >>>Bob >> >>I suppose the most pressing question now is: "does it make crafty stronger?" >> >>James > > >Way too soon to tell. It gives very deep search results that can be used to >choose or avoid a book line. Out of the nearly 800K CAP results, almost 200K >matched positions in my book (based on enormous) and were saved. But as to how >it affects the overall play, this will take time to analyze, and it will take >more time to perfect the usage of this new info... > >But it is in and working in a rudimentary way already...
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.