I would like to use neural network method for approximating Nash Equilibrium for general games (especially card games)... What would be some standard package I could use? Thank you.


It sounds like you should be looking into Gambit which is a GNU licensed package for exactly that sort of thing:

  • Open Source - GNU GPL Licence
  • Cross Platform Windows, Mac & Linux
  • GUI for exploring strategies & outcomes
  • Python API for implementing complex games, etc.
  • Command-line tools
    • gambit-enumpure: Enumerate pure-strategy equilibria of a game
    • gambit-enumpoly: Compute equilibria of a game using polynomial systems of equations
    • gambit-enummixed: Enumerate equilibria in a two-player game
    • gambit-gnm: Compute Nash equilibria in a strategic game using a global Newton method
    • gambit-ipa: Compute Nash equilibria in a strategic game using iterated polymatrix approximation
    • gambit-lcp: Compute equilibria in a two-player game via linear complementarity
    • gambit-lp: Compute equilibria in a two-player constant-sum game via linear programming
    • gambit-liap: Compute Nash equilibria using function minimization
    • gambit-simpdiv: Compute equilibria via simplicial subdivision
    • gambit-logit: Compute quantal response equilbria
    • gambit-convert: Convert games among various representations
  • the problem is, the game I am interested in is extremely complex, just like texas holdem... Which i think those method should not work...
    – user40780
    Oct 24 '15 at 7:41
  • Have you tried it costs nothing Oct 24 '15 at 16:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.