Wednesday, July 28, 2004
Hi, this is Patrick Kellogg. That war paper can be found on the web at (http://www.patrickkellogg.com/school/papers/war/index.htm). Thanks for checking it out.
The research I'm currently most fascinated with is evolutionary programming, which is a lot like genetic algorithms, but it's the actual *algorithm* that is modified each generation. I'm hoping I can have a computer find better statistical measures for "good" war hands than I could do myself. However, beyond buyng a bunch of books, I've been too busy writing other code lately.
I never did a similar statistical test for solitaire, and I really wanted to. I wanted to answer the question of how many games were "winnable", how to recognize winnable set-ups before playing, and whether an optimal strategy exists. Oh well, maybe in the future! Take care, and happy coding...
For many solitaire games, it is NP-complete (i.e. computationally intractable) to determine a priori whether the game is winnable. However, in most solitaire games (unlike War) there is a little bit of strategy involved -- places where you have choices what play to make next -- that combinatorially explodes the size of the solution space.Post a Comment
TRS-80 BASIC interpreter
Analysis of The Game of Chips
Man-Bag Buying Guide
Home ImprovementCustom Home Ideas
MiscellanyAmazon.com Wish List
Old StuffSteiner Trees
D.C.-area Software Companies