[Kde-games-devel] AI library in the gamelibs?

Ian Wadham iandw.au at gmail.com
Sun Jan 5 02:11:42 UTC 2014


On 05/01/2014, at 3:01 AM, Inge Wallin wrote:

> On Saturday, January 04, 2014 12:06:01 Ian Wadham wrote:
> > Secondly, I wonder if the above approaches could provide a better
> > AI player in KJumpingCube (Einstein?). ATM the algorithm is basic
> > MiniMax, preceded by an algorithm to pick a set of "likely" moves.
> > Kepler AI is the original algorithm and Newton AI is one I added.
>  
> Interesting. I have completely missed this game.  I cloned the repo and had a quick look at the source code.  And I would say that yes, it could definitely produce a better AI.  Even if it didn't add anything to the evaluation, it would still improve things by letting you search deeper by using more efficient search code.

Thank you for your interest, Inge.

> I didn't see any use of alpha-beta in the KJC code. Did I miss it?  That thing in itself would let you search almost twice as deep in the same time.

There's a TODO about it in file ai_main.cpp, but that is as far as I got ...

> [1] http://chessprogramming.wikispaces.com/Alpha-Beta
> [2] http://chessprogramming.wikispaces.com/Transposition+Table

Thanks for the links.

Cheers, Ian W.



More information about the kde-games-devel mailing list