Manual

Standard Rules

Each player has nine pieces (hence the name, Nine Men's Morris) which are placed and moved on the line crossings of the board. Whenever three pieces of the same color are placed in a straight row, a mill is closed and one opponent piece may be removed. The goal of the game is to reduce the opponent to only two pieces (such that he cannot form a mill anymore), or to surround the opponent pieces in such a way that there are no valid moves for the opponent.

The game proceeds in three distinct phases (opening, midgame, and endgame). Unlike chess, these phases are distinguished by special rules for each phase.

Opening - Setting pieces

The white player begins. Each player places one piece on an unoccupied position on the board in turns. If a mill is closed by setting a piece, the player may take one of the opponent's pieces. Once all pieces are set, the midgame starts.

Midgame - Moving pieces

Each player moves one piece along the lines to a free, neighboring position. Again, if the move results in closing a mill, one of the opponent's pieces may be removed. Note that a player must move a piece in each turn. If there is no legal move, the player has lost.

Endgame - Flying

If a player has only three pieces left, he my jump (or fly) with one piece to any unoccupied position instead of moving only along the board lines.

Rule Details and Game Variations

Some of these rules are often interpreted differently, such that a variety of rule variants exist. This game tries to support most of them. In particular, the following rule variations are supported:

AI algorithm

The AI algorithm is a standard alpha-beta search in a NegaMax implementation using iterative deepening. It employs a transposition table to quickly find previously computed positions.

A special feature is the automatic learning capability: whenever the computer wins or loses a game, it will prefer to obtain or avoid similar situations in the future. This results in a better long-term motivation, since the computer will not make the same mistake twice and the gameplay will be more randomized.

The evaluation function is still quite basic and consists of four parts: