Jump to content
  • Advertisement
Sign in to follow this  
manolopm

AI for cards games?

This topic is 4818 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

I wan't to write a multiplayer card game. I want to make AI players, but I don't know how to make it. I was thinking in minimax, alphabeta and somethings like that but it seems not to be a good solution for cards games. Any Idea? Url? Document? Help? Thanks in advance

Share this post


Link to post
Share on other sites
Advertisement
Which card game are you thinking of? I think AI for just about every card game I can think of is going to have to be probablistic in nature.

Whichever card game you decide on, try to think about what a strategy for that game actually entails. For example, with Texas Holdem, the strategy is simply your betting strategy because you have no choice on which cards to pick.

In the case of holdem, you could calculate the probability that you would win with your hand at each betting stage (a simple matter once you get the math down), and run that probability through some sort of function to determine your bet (or whether to fold). This function would have to be at least somewhat random in nature, or your AI would have no "poker face." You would also have to make an estimate of the opponents confidence of their hands through a mixture of probability means as well as perhaps applying a Bayesian model to their betting to determine a pattern. The function to determine your bet would be derived from weighing risk versus reward - you wouldn't want to bet a losing hand very often, but you don't want to bet a winning hand too strongly if your opponent did not appear confident in their hand. I haven't tried to implement anything such as this, so I have no idea how it would work in practice, but I imagine it would do pretty decently against beginner to intermediate players.

Share this post


Link to post
Share on other sites
manolopm,
I've some boots that play a four player trick-tacking card game. They play against human people with reasonable good results. I've been working on them for quite a long time (first ones, 10 years ago) ...

I have been trying to use expectiminimax to improve its current euristics algorithms, but for now results are worse. They use minimax "after" game to analyze its own results.

Wich game are you working on? (just a guess: mus? :) :) )

Send me a private if you are interested in details.

Thanks

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

We are the game development community.

Whether you are an indie, hobbyist, AAA developer, or just trying to learn, GameDev.net is the place for you to learn, share, and connect with the games industry. Learn more About Us or sign up!

Sign me up!