Author: Vasik Rajlich
Date: 06:16:44 01/20/06
Go up one level in this thread
On January 19, 2006 at 10:56:03, Robert Hyatt wrote: >On January 19, 2006 at 08:32:33, Vasik Rajlich wrote: > >>On January 19, 2006 at 03:40:33, Tord Romstad wrote: >> >>>On January 18, 2006 at 22:19:30, Jay Urbanski wrote: >>> >>>>On January 18, 2006 at 07:32:07, Vasik Rajlich wrote: >>>> >>>>>On January 17, 2006 at 21:34:23, Steve Maughan wrote: >>>>> >>>>>>Vasik, >>>>>> >>>>>>>I was thinking to have just a "spin" parameter where the tester indicates the >>>>>>>opponent's rating. (Or rating difference ..) All engines could in principle >>>>>>>expose such a parameter. >>>>>> >>>>>>Do you know that the UCI specification already has this feature? The >>>>>>UCI_Opponent is a standard option that also send the opponents ELO. While it's >>>>>>part of the spec I don't know of any GUI that actually sends this info - not >>>>>>even the Shredder Classic GUI. >>>>>> >>>>>>Regards, >>>>>> >>>>>>Steve >>>>> >>>>>Steve, >>>>> >>>>>thanks for the info. In this case, if I implement this feature, it will be >>>>>through this command. Testers may need to set it manually (if the GUI doesn't). >>>>> >>>>>Vas >>>> >>>> >>>>As someone who uses engines mostly on ICS servers, I would love to see this >>>>feature implemented. >>> >>>You could easily implement this yourself by making some small additions >>>to PolyGlot, I think. I am fairly sure XBoard already sends the opponent's >>>name, rating and title. >>> >>>>Now if there were only also a way to indicate whether >>>>your opponent is human or a (C)omputer :) >>> >>>There is. Have a look at the UCI specification: >>> >>>* <id> = UCI_Opponent, type string >>>With this command the GUI can send the name, title, elo >>>and if the engine is playing a human or computer to the >>>engine. >>>The format of the string has to be [GM|IM|FM|WGM|WIM|none] [<elo>|none] >>>[computer|human] <name> >>>Examples: >>>"setoption name UCI_Opponent value GM 2800 human Gary Kasparov" >>>"setoption name UCI_Opponent value none none computer Shredder" >>> >>>Tord >> >>Thanks for the info. For Rybka 1.2, I will use the rating data, but (probably) >>not the other data. I'm not really even sure how the human/computer info could >>be taken advantage of. I suppose some theoretically drawn endgames could be >>given positive evaluations. Anyway it would be pretty hard to get test data. >> >>Vas > > >I will add that I put this into xboard years ago. Along with code to supply the >actual name of the opponent, and his rating along with the program's rating, >which I used (still use) to influence draw decisions and the contempt factor. Aha. Now it just needs to be put into Polyglot and we're all set :) Vas
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.