Modify to allow user to specify that Rybka will play the nth best move, assuming that the nth choice is "reasonably playable."
For example, the user might specify that Rybka will play the 4th best move. Then Rybka would do so ONLY IF doing so would not be a gross blunder. "Gross Blunder" might be user defined as a move where the difference between the score for the nth move choice was not more than m rating points below that for the best move choice. The user should also be given an option to select the value of m.
For example: User selects n=3, m=0.25 pawns. Then if the third best choice was evaluated to be 0.5 pawns worse than the best choice, Rybka would automatically go to the second best move. If the second best move were also more than 0.25 pawns worse than the best choice then Rybka would play the first choice move in the given position.
The motivation for this suggested change is to help learners play against intelligently selected nth choice moves. The beginner might begin with n=5 and m=1 pawn.
The only doubt I have is whether or not this can be implemented at the GUI level without some modification or other accommodation by the engine.
What GUI are you using?
I do my detailed middlegame and endgame analysis using ChessBase 11 with your Rebka engine running as an "analysis engine." I also use hard copy endgame books as a guide for endgame analysis. [Of course, I use endgame tablebases.]
I do my opening repertoire expansion analyses also using ChessBase 11 with your Rebka engine running as an "analysis engine" except I examine ten lines simultaneously and let it run at least eight hours for each opening position. After that, I examine promising lines more and compare to what GMs are doing in MegaBase.
I will forward my product improvement suggestion to the ChessBase people.
Perhaps your GUI people would be interested in this idea too?
The whole idea was to help people who have trouble finding suitable human opponents. Most "internet humans" seem to want to play bullet chess and not "real" chess. Finding someone to play with reasonable time limits can be frustrating for strong amateurs. It is no fun for a strong amateur to play against beginners because the games are not sufficiently challenging. There are always plenty of beginners and weak amateurs to play against but the stronger players seem to want bullet chess.
> I will forward my product improvement suggestion to the ChessBase people.
> Perhaps your GUI people would be interested in this idea too?
Good luck, your suggestion will probably be more useful than the Mate-O-Meter or Hotness features.
>Perhaps your GUI people would be interested in this idea too?
I use Shredder Classic, they haven't added features in a long time so I don't think it'd happen.
If you meant the general you, then it's possible you can get it implemented into Aquarium (they already have some implementation of human emulation) more easily than you can get it implemented into Chessbase, so it's worth a try. For free chess GUIs, you can suggest the feature for Winboard in the Computer Chess subforum.
I started looking at Aquarius but must admit that I find it very confusing. It is not clear what Aquarius is! Also, I am not sure that my suggestion would be well received by them because the purpose of Aquarius seems more complex than just playing chess.
The fact is that I am a bit too cynical. I know that this is irrational but my feeling is that most people do not like change. My guess is that I will get a negative response. My suggestion would require some expenditure of resources and may be viewed as a hassle by the programmers. :-(
Nevertheless, if they do implement the suggestion, I will probably start playing a lot more against chess-playing programs.
Yes, the GUI is very confusing but if you only want to deal with your feature you could ignore everything else.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill