Negamax is a variation on minimax which is an algorithm for determining the best move for a given player and game state in 2-player zero-sum game.

MiniMax.hs provides a game class MiniMaxable,

class MiniMaxable game where
  score :: (Real value) => Bool -> game -> value
  terminal :: game -> Bool
  allMoves ::  Bool -> game -> [game]

and 3 function, negamax, negamax', and negamaxalpha. negamax and negamax' are equivalent implementations of negamax, and negamaxalpha is my take on alpha/beta pruning with negamax.

If you can represent a 2-player zero-sum game in terms of the 3 functions required by MiniMaxable, the 3 version of negamax included in MiniMax.hs will tell you which moves are optimal.

The algorithms included here do not parameterize depth information and therefore will explore the full depth of the game tree. This means that if you have a particularly deep game tree (cough chess), the included algorithms will run for longer than you are willing to wait.