Introduction | Prisoner's Dilemma | Iterated Prisoner's Dilemma | The Game | Results

Part 1: Introduction

For this Master's thesis, 'emergent cooperative gameplay' is when players choose to cooperate in an otherwise non-cooperative game.

The project had two goals:

This Master's project is half game and half data. The Gamer's Dilemma is a 2D two-player game that implements a system derived from game theory in order to influence the choices players make.

Background

Most cooperative games beat players over the head with hard rules for cooperation - players are put on teams and pit against other teams, they share the same goals (i.e., defend the flag), and unilateral actions costs you and your team points, while actions that help the team give the individual extra points. These tactics work extremely well.

This project, though, is concerned with creating a system where there are no explicit rules on cooperating with or competing against the other players and then seeing if that system actually works - if, over time, players choose to cooperate.

A system already exists that does this, though it's typically used in mathematics, economic theory, political science, and even evolutionary theory: the Prisoner's Dilemma. Players are given two choices - cooperate with the other person or rat them out by defecting - and based on the combined choices of two players a different outcome plays out.

When playing a single round of the Prisoner's Dilemma, humans will defect (compete) against each other about 70% of the time, often to their own detriment. But when playing multiple rounds of the Prisoner's Dilemma - known as the Iterated Prisoner's Dilemma - player strategies curve up towards cooperation over time.

The Prisoner's Dilemma is essentially a system that creates a naturally occurring - or emergent - cooperative strategy.

Unfortunately, the Prisoner's Dilemma is normally only played in a very sterile environment with none of the bells and whistles you see in a game marketed for entertainment. The Prisoner's Dilemma is not fun. The Iterated Prisoner's Dilemma is downright tedious.

The challenge, then, was to take the system of the Prisoner's Dilemma and implement it in a video game, add dozens of variables (like flashy graphics, timers, movement, and enemies!), make it fun, but still get the underlying system to influence players to change their strategies to cooperate with one another.

What the Project Does

For this project, I created a two-player 2D game called "The Gamer's Dilemma". In a series of fast rounds, players had two choices: send out a bomb to destroy the enemies converging upon them, or redeem that bomb for points to add to their score. Each player's objective was to get the highest possible score within sixty rounds.

While they play, the game records all kinds of data, such as they choice they make each round, their score, and their controls, and used that data to figure out if the project was a success and identify what new variables affected player choices the most.

Continue to Part 2: The Prisoner's Dilemma