Due date: Wednesday, October 8, 2025.

Ex. 1. State Benefit Heuristic

Download the following Python modules implementing a D&D like game with one human player and multiple NPCs, like in Homework 2.
GM2.py
Player2.py

We will continue the implementation of an intelligent NPC for this game. For this homework, in the class Player, in the function play, the selected action and the opponent to attack if the action is 4 are obtained by a call to the function strategy_choice. For now, this function just makes a random choice of action and opponent, which is done now by the function with the same name. Replace that function call with a call to a function you define implementing your chosen strategy.  The goal is to implement an informed search for the best action by defining a heuristic that estimates the benefit of a state and chooses an action that attempts to optimize the outcome.

After the deadline, we will collect all the heuristics and have a tournament where they compete with each other. The winner gets 3 extra credit points. The second place gets 2 points, the third place 1 point, and the first loser 1 point. 

Changes to the rules defined in homework 2 and other tournament details:

• In a battle, the losing player loses not just 1 health point, but a number of points equal to the ceiling of half of the difference between the generated attacks.
• The GM now detects when a single player is left alive and ends the game.
• A self-attack is admissible. The player attacking itself can lose health points but gain intelligence, which will help in future actions. A massive loss of health can lead to the player being eliminated. The action can also result in a draw, which has no effect on the player. 
• The Player class stores an id number idn for the players to make it easier to figure out which player did what in each round. 
• The tournament will start with 4 groups of 4 NPCs playing against each other. The winner of each group will participate in a second round that will determine the first, second, and third places.
• The first loser will be determined by the number of rounds it took for it to lose. In case of a tie, another round can choose the first loser among them.
• In cases of ties, player stats can also be used: the player's strength plus twice the player's intelligence plus their home base level. 
• To make sure each competition group eventually ends, we will limit the number of action rounds to 50. 
• Only code submitted on time can participate in the competition.

Turn in:

Submit the modified file Player2.py to Canvas, Assignments, Homework 6.