Curtis Rollo
msg tools
Hey guys,

My name is Curtis, and in my spare time, I love boardgames and coding. So sometimes, in my spare time, I would re-create board games in code (python) and think about how AI can find the best strategy.

But I'm just a hobbyist, and don't know that much about Neural Networks and Machine Learning. Nor for that matter, how to tweak mathematical algorithms to max efficiency. I'm amazed when someone like Code Bullet can make AI learn to play Flappy Bird on their own. Can I do something like that, except with board games?

Anyone have any similar interests? I'd be happy to meet people who have similar ideas and we can talk things out.

Currently, working on a program that can play Dice Hospital.

Thanks,
Curtis
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Kevin B. Smith
United States
Lake Forest Park
Washington
flag msg tools
designer
Avatar
mbmbmbmbmb
I would love to get into projects along those lines. Sadly, I don't think I'm motivated enough to free up the time to work on stuff like that.

The Keldon Race for the Galaxy implementation amazed me when it came out (2009). You can get the source code. I haven't looked at it.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Thomas Stevenson
United States
Columbus
Ga
flag msg tools
badge
mbmbmbmbmb
One approach that works for a lot of games is to come up with way to score the game state and then evaluate every possible move and pick the one the nets the biggest increase. That work well for me with "Empires of the Middle Ages" and pretty well for "Concordia" although I had to add a long If then else block on what resources to gather when no move generates victory points for the later.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.