Comprehensive Game Theory Learning Roadmap

This comprehensive roadmap guides you from foundational concepts through cutting-edge research in game theory. The field uniquely combines rigorous mathematics with practical applications across economics, computer science, biology, and social sciences. Start with core equilibrium concepts, implement algorithms to solidify understanding, and gradually explore specialized topics aligned with your interests.

1. Structured Learning Path

Phase 1: Mathematical Foundations (2-3 months)

Probability and Statistics

  • Probability spaces, random variables, distributions
  • Conditional probability and Bayes' theorem
  • Expected value, variance, and moments
  • Joint distributions and independence
  • Law of large numbers and central limit theorem

Linear Algebra

  • Vectors, matrices, and systems of equations
  • Eigenvalues and eigenvectors
  • Linear transformations and rank
  • Matrix games and linear programming connections

Calculus and Analysis

  • Optimization: unconstrained and constrained
  • Lagrange multipliers and KKT conditions
  • Convex analysis basics
  • Fixed-point theorems (Brouwer, Kakutani)

Discrete Mathematics

  • Set theory and relations
  • Graph theory basics
  • Combinatorics and counting
  • Logic and formal proofs

Microeconomics Fundamentals

  • Utility theory and preferences
  • Risk aversion and expected utility
  • Consumer and producer theory
  • Market equilibrium concepts

Phase 2: Core Game Theory (3-4 months)

Strategic Form Games (Normal Form)

  • Game representation: players, strategies, payoffs
  • Dominant and dominated strategies
  • Iterated elimination of dominated strategies (IEDS)
  • Best response and best response correspondence
  • Nash equilibrium: definition and existence
  • Pure vs. mixed strategy Nash equilibrium
  • Computing Nash equilibria (2x2, nxn games)
  • Rationalizability and common knowledge

Extensive Form Games

  • Game trees and perfect information
  • Backward induction and subgame perfection
  • Subgame perfect Nash equilibrium (SPNE)
  • Imperfect information and information sets
  • Behavioral strategies vs. mixed strategies
  • Kuhn's theorem on equivalence
  • Credible threats and commitment

Applications and Classic Games

  • Prisoner's Dilemma and cooperation problems
  • Coordination games (Stag Hunt, Battle of the Sexes)
  • Hawk-Dove game and war of attrition
  • Matching pennies and zero-sum games
  • Public goods games
  • Tragedy of the commons
  • Voting games and majority rule

Zero-Sum Games

  • Minimax theorem
  • Value of the game
  • Optimal strategies
  • Connection to linear programming
  • Von Neumann's minimax theorem
  • Rock-Paper-Scissors and symmetry

Phase 3: Advanced Equilibrium Concepts (2-3 months)

Refinements of Nash Equilibrium

  • Trembling hand perfect equilibrium
  • Proper equilibrium
  • Sequential equilibrium
  • Perfect Bayesian equilibrium
  • Forward induction
  • Stability and ESS (Evolutionarily Stable Strategy)

Repeated Games

  • Finitely repeated games
  • Infinitely repeated games and discounting
  • Folk theorems (feasibility, individual rationality)
  • Trigger strategies and punishment
  • Renegotiation-proof equilibria
  • Reputation effects
  • Automaton representation

Bayesian Games (Incomplete Information)

  • Types, beliefs, and common prior assumption
  • Bayesian Nash equilibrium
  • Mechanism design preliminaries
  • Auctions as Bayesian games
  • Signaling games
  • Screening games
  • Perfect Bayesian equilibrium in signaling

Coalitional Game Theory

  • Characteristic function form
  • Core, Shapley value, nucleolus
  • Transferable vs. non-transferable utility
  • Bargaining sets and stable sets
  • Convex games
  • Market games and assignment problems
  • Voting power indices (Shapley-Shubik, Banzhaf)

Phase 4: Specialized Topics (3-4 months)

Mechanism Design and Auctions

  • Revelation principle
  • Incentive compatibility (IC) and individual rationality (IR)
  • Direct mechanisms vs. indirect mechanisms
  • Vickrey-Clarke-Groves (VCG) mechanism
  • Revenue equivalence theorem
  • Optimal auction design (Myerson)
  • First-price, second-price, all-pay auctions
  • Combinatorial auctions
  • Spectrum auctions
  • Implementation theory

Evolutionary Game Theory

  • Replicator dynamics
  • Evolutionarily stable strategies (ESS)
  • Adaptive dynamics
  • Population games
  • Stochastic stability
  • Learning in games
  • Moran process
  • Spatial games and networks

Algorithmic Game Theory

  • Computational complexity of Nash equilibrium (PPAD-completeness)
  • Approximate equilibria
  • Price of anarchy and price of stability
  • Network routing games
  • Selfish routing and Braess's paradox
  • Congestion games and potential games
  • Online learning and regret minimization

Cooperative Game Theory

  • Matching theory (stable matching, Gale-Shapley)
  • Market design applications
  • Fair division problems
  • Cake cutting algorithms
  • Coalition formation
  • Network formation games
  • Bargaining theory (Nash, Rubinstein, Kalai-Smorodinsky)

Differential Games

  • Continuous-time strategic interactions
  • Open-loop vs. closed-loop strategies
  • Linear-quadratic differential games
  • Pursuit-evasion games
  • Dynamic programming approach
  • Hamilton-Jacobi-Isaacs equations

Phase 5: Modern Applications (2-3 months)

Multi-Agent Systems

  • Agent-based modeling
  • Emergence and self-organization
  • Coordination mechanisms
  • Multi-agent reinforcement learning (MARL)
  • Communication and protocols
  • Distributed systems and consensus

Network Games

  • Games on graphs
  • Network formation and stability
  • Influence and diffusion games
  • Social network analysis
  • Networked markets
  • Contagion and cascades

Algorithmic Trading and Markets

  • Market microstructure
  • High-frequency trading strategies
  • Liquidity provision games
  • Order book dynamics
  • Automated market makers (AMMs)
  • Decentralized finance (DeFi) mechanisms

Political Economy and Voting

  • Social choice theory
  • Voting paradoxes (Condorcet, Arrow's impossibility)
  • Strategic voting
  • Gerrymandering and districting games
  • Campaign finance games
  • Legislative bargaining

Behavioral Game Theory

  • Bounded rationality
  • Level-k reasoning and cognitive hierarchy
  • Quantal response equilibrium (QRE)
  • Prospect theory in games
  • Fairness and reciprocity
  • Experimental game theory results
  • Neuroeconomics foundations

2. Major Algorithms, Techniques, and Tools

Algorithms for Computing Equilibria

Nash Equilibrium Algorithms

  • Support enumeration (Lemke-Howson for 2-player)
  • Simplicial subdivision methods
  • Continuation methods
  • Best response dynamics
  • Fictitious play and variants
  • Gradient-based methods
  • Regret matching
  • Counterfactual regret minimization (CFR)
  • Linear complementarity problem (LCP) formulation
  • Polynomial approximation schemes (PTAS)

Correlated Equilibrium

  • Linear programming formulation
  • Swap regret minimization
  • Internal regret algorithms
  • Correlated Q-learning

Evolutionary Dynamics

  • Replicator equation simulation
  • Moran process Monte Carlo
  • Wright-Fisher model
  • Best response dynamics
  • Logit dynamics
  • Pairwise comparison dynamics

Mechanism Design Tools

  • Automated mechanism design (AMD)
  • Virtual valuations computation
  • Revenue optimization algorithms
  • Incentive compatibility checking
  • VCG payment computation

Coalition Formation

  • Partition function computation
  • Core checking algorithms
  • Shapley value calculation
  • Nucleolus computation (LP-based)
  • Stable matching algorithms (Gale-Shapley, top trading cycles)

Auction Algorithms

  • Hungarian algorithm (assignment)
  • Ascending/descending price auctions
  • Combinatorial auction solvers (CPLEX, Branch-on-Bids)
  • Winner determination problem (WDP) solvers
  • Bidding strategies (truthful vs. strategic)

Computational Techniques

Game Tree Analysis

  • Minimax algorithm
  • Alpha-beta pruning
  • Monte Carlo Tree Search (MCTS)
  • Upper Confidence Bound for Trees (UCT)
  • Expectimax for stochastic games
  • Best-first minimax search

Learning in Games

  • Q-learning and multi-agent Q-learning
  • Policy gradient methods (REINFORCE)
  • Actor-critic methods
  • Independent learners vs. joint action learners
  • No-regret learning (Hedge, EXP3)
  • Mirror descent in games

Approximation Methods

  • Nash equilibrium approximation (ε-Nash)
  • Quasi-polynomial time algorithms
  • Heuristic methods (genetic algorithms, simulated annealing)
  • Mean field games (large population limits)
  • Sampling-based methods

Optimization Techniques

  • Linear programming for zero-sum games
  • Quadratic programming for equilibria
  • Mixed-integer programming for discrete games
  • Variational inequalities
  • Fixed-point iterations
  • Complementarity formulations

Software Tools and Libraries

Python Libraries

  • Nashpy (2-player games, support enumeration)
  • Gambit (extensive game analysis, GUI)
  • QuantEcon (economic modeling, repeated games)
  • OpenSpiel (RL and game playing)
  • PettingZoo (multi-agent RL environments)
  • Game Theory Explorer
  • PyNFG (network form games)
  • EGTA (empirical game-theoretic analysis)

MATLAB/Octave

  • Game Theory Toolbox
  • SGT (Stochastic Game Theory Toolbox)
  • Custom implementations for research

R Packages

  • GameTheory
  • GTDesign (experimental design)
  • CoopGame (cooperative games)

Specialized Software

  • Gambit (comprehensive game theory software)
  • AGS (Action Graph Games)
  • GAMUT (game generator)
  • Poker solver tools (PioSOLVER, GTO+)
  • Auction simulation platforms

General Purpose

  • SageMath (symbolic computation)
  • Julia packages (Games.jl, StrategicGames.jl)
  • Mathematica/Wolfram
  • CPLEX, Gurobi (optimization backend)

3. Cutting-Edge Developments

AI and Multi-Agent Learning

Deep Reinforcement Learning for Games

  • AlphaGo/AlphaZero architecture for perfect information games
  • DeepStack and Libratus for poker (imperfect information)
  • OpenAI Five for complex team games (DOTA 2)
  • AlphaStar for real-time strategy (StarCraft II)
  • Neural Fictitious Self-Play (NFSP)
  • Policy Space Response Oracles (PSRO)

Multi-Agent Reinforcement Learning Theory

  • Convergence guarantees in multi-agent settings
  • Opponent modeling and theory of mind
  • Centralized training, decentralized execution (CTDE)
  • QMIX and value decomposition networks
  • Multi-agent actor-critic methods (MADDPG)
  • Communication learning in cooperative settings

Large-Scale Game Solving

  • Abstraction techniques for massive games
  • Counterfactual regret minimization variants (CFR+, Linear CFR, DCFR)
  • Neural network function approximation in CFR
  • Solving games with 10^160+ states
  • Subgame solving and safe resolving

Mechanism Design and Markets

Automated Mechanism Design

  • Machine learning for mechanism design
  • Deep learning for revenue optimization
  • Neural auction design
  • Differentiable economics
  • End-to-end learning of market mechanisms

Blockchain and Decentralized Mechanisms

  • Consensus mechanisms (proof-of-work, proof-of-stake)
  • Smart contract game theory
  • Maximal extractable value (MEV) games
  • Tokenomics and incentive design
  • DAO governance mechanisms
  • Flash loan attacks as equilibrium deviations

Matching Markets

  • Dynamic matching algorithms
  • Online matching with stochastic arrivals
  • Thick market design
  • Refugee matching systems
  • School choice improvements
  • Kidney exchange optimizations

Theoretical Advances

Computational Complexity

  • PPAD-completeness refinements
  • Hardness of approximation results
  • Communication complexity in distributed games
  • Query complexity of equilibrium finding
  • Beyond worst-case analysis

Bounded Rationality Models

  • Rational inattention theory
  • Entropy regularization in games
  • Quantal response equilibrium extensions
  • Cognitive hierarchies in large games
  • Satisficing and aspiration-based behavior

Learning Theory Convergence

  • Last-iterate convergence in games
  • Finite-time convergence guarantees
  • No-regret learning in non-convex games
  • Convergence to Nash vs. correlated equilibria
  • Time-average vs. last-iterate behavior

Interdisciplinary Applications

Cybersecurity Games

  • Network defense strategies
  • Adversarial machine learning as games
  • Intrusion detection game theory
  • Deception in security (honeypots)
  • Moving target defense
  • Cyber insurance markets

Climate Change and Sustainability

  • International climate agreements as games
  • Carbon pricing mechanisms
  • Common-pool resource management
  • Energy market design
  • Sustainable supply chain games

Epidemiology and Public Health

  • Vaccination games and herd immunity
  • Social distancing as coordination problem
  • Information spread vs. disease spread
  • Resource allocation during pandemics
  • Behavioral interventions design

Autonomous Systems

  • Self-driving car interactions
  • Drone swarm coordination
  • Human-robot interaction games
  • Traffic management as game
  • Multi-robot task allocation

Social Media and Online Platforms

  • Attention economy games
  • Content recommendation as mechanism design
  • Fake news spread as signaling game
  • Platform competition (two-sided markets)
  • Influencer economics
  • Echo chamber formation

4. Project Ideas by Level

Beginner Projects (1-2 weeks each)

Project 1: Classic Game Analyzer

Build a tool to analyze 2x2 and 3x3 games. Find all pure strategy Nash equilibria, dominated strategies, and mixed strategy equilibria. Visualize best response correspondences.

Project 2: Prisoner's Dilemma Tournament

Implement a tournament of strategies for iterated Prisoner's Dilemma (Tit-for-Tat, Always Defect, Pavlov, etc.). Analyze which strategies perform best and why.

Project 3: Rock-Paper-Scissors Variants

Create variations of RPS (Rock-Paper-Scissors-Lizard-Spock, weighted payoffs). Compute mixed strategy Nash equilibria. Build a bot that plays optimally.

Project 4: Nim and Combinatorial Games

Implement Nim and analyze winning strategies. Extend to other combinatorial games (Chomp, Hex). Use backward induction to find optimal play.

Project 5: Auction Simulator

Simulate first-price, second-price, and all-pay auctions. Compare revenue and efficiency across auction formats. Analyze bidding strategies for different valuations.

Project 6: Public Goods Game

Create a simulation of the public goods game with varying contribution levels and free-riding. Visualize how cooperation evolves with different parameters.

Project 7: Coordination Game Experiments

Implement Battle of the Sexes and Stag Hunt games. Create interactive visualizations showing multiple equilibria and coordination problems.

Intermediate Projects (2-4 weeks each)

Project 8: Poker Bot (Simplified)

Build a bot for simplified poker (Kuhn poker or Leduc Hold'em). Implement CFR algorithm to find approximate Nash equilibrium strategies. Visualize strategy profiles.

Project 9: Repeated Game Analyzer

Analyze infinitely repeated games with different discount factors. Compute the set of feasible and individually rational payoffs. Verify folk theorem conditions.

Project 10: Voting System Comparisons

Implement multiple voting systems (plurality, Borda count, approval, ranked-choice). Analyze strategic manipulation and Condorcet paradoxes with simulated elections.

Project 11: Stable Matching Simulator

Implement Gale-Shapley algorithm for stable matching. Simulate college admissions or residency matching. Analyze incentive compatibility and strategic misrepresentation.

Project 12: Network Formation Game

Model strategic network formation where players choose connections. Analyze stable networks under different cost-benefit structures. Visualize network evolution.

Project 13: Congestion Game Simulation

Simulate traffic routing as a congestion game. Compute user equilibrium vs. social optimum. Calculate price of anarchy for different network topologies.

Project 14: Mechanism Design Tool

Create a tool for designing simple mechanisms (single-item auctions, public project decisions). Check incentive compatibility and implement VCG payments.

Project 15: Evolutionary Game Dynamics

Simulate replicator dynamics for various games (Hawk-Dove, Prisoner's Dilemma). Visualize phase portraits and identify evolutionarily stable strategies.

Advanced Projects (1-2 months each)

Project 16: Multi-Agent RL Environment

Build a custom multi-agent environment and train agents using MARL algorithms (IQL, QMIX, MADDPG). Compare convergence to Nash equilibrium vs. other outcomes.

Project 17: Combinatorial Auction Solver

Implement a combinatorial auction for complex allocation problems (spectrum, logistics). Solve winner determination problem. Compare different payment rules.

Project 18: Signaling Game Analyzer

Analyze separating and pooling equilibria in signaling games (job market, education signaling). Implement perfect Bayesian equilibrium finder for finite types.

Project 19: Algorithmic Trading Game

Create a limit order book simulation where multiple agents compete. Implement market-making strategies and analyze equilibrium spreads and depth.

Project 20: Coalition Formation Platform

Build a platform for analyzing coalition formation in cooperative games. Compute core allocations, Shapley value, and stability concepts for various games.

Project 21: Bargaining Protocol Simulator

Implement Rubinstein bargaining and variations. Analyze how discount factors, outside options, and impatience affect equilibrium outcomes.

Project 22: Security Game Application

Model a security domain (airport screening, network defense) as a Stackelberg game. Solve for defender's optimal mixed strategy commitment. Visualize coverage.

Project 23: Online Learning in Games

Implement various no-regret learning algorithms (Hedge, EXP3, multiplicative weights). Analyze convergence to correlated vs. Nash equilibrium in different game classes.

Research-Level Projects (2-4 months each)

Project 24: Deep CFR Implementation

Implement Deep CFR for large-scale imperfect information games. Apply to custom poker variants or other domains. Compare with tabular CFR and analyze scalability.

Project 25: Mean Field Game Solver

Model a large-population game using mean field approximation. Solve using PDE methods or neural networks. Compare with finite-population equilibria.

Project 26: Automated Mechanism Design with ML

Use neural networks to design mechanisms for specific objectives. Implement end-to-end differentiable auction learning. Test on allocation problems with complex constraints.

Project 27: Blockchain Consensus Analysis

Model blockchain consensus as a game (longest chain, BFT). Analyze incentive compatibility and security under different attack scenarios. Quantify centralization risks.

Project 28: Bounded Rationality Experiments

Implement level-k reasoning and quantal response equilibrium models. Conduct human-subject experiments or agent simulations. Compare predictions with behavior.

Project 29: Multi-Agent Communication Learning

Train agents to develop communication protocols emergently. Use neural networks with discrete or continuous communication channels. Analyze evolved languages and compositionality.

Project 30: Fair Division with Strategic Agents

Design and implement fair division algorithms for strategic agents (cake cutting, rent division). Prove or test incentive compatibility. Compare efficiency-fairness tradeoffs.

Project 31: Adversarial ML as Games

Frame adversarial machine learning as a two-player game. Implement attacks and defenses as strategies. Analyze equilibria and provide robustness certificates.

Project 32: Dynamic Matching Market

Build a continuous-time matching market with arrivals and departures. Implement thick market design principles. Optimize for welfare under uncertainty.

Project 33: Network Game with Learning

Create a network game where agents learn about network structure and payoffs. Implement distributed learning algorithms. Analyze emergence of coordination and information diffusion.

Project 34: Behavioral Game Theory Framework

Build a comprehensive framework incorporating multiple behavioral biases (fairness, loss aversion, probability weighting). Predict behavior in various games and validate empirically.

Recommended Learning Resources

Textbooks

Foundational

  • "Game Theory" by Fudenberg & Tirole (comprehensive graduate text)
  • "A Course in Game Theory" by Osborne & Rubinstein (mathematical approach)
  • "Game Theory: An Introduction" by Tadelis (modern, applied focus)
  • "Essentials of Game Theory" by Leyton-Brown & Shoham (AI perspective)

Specialized Topics

  • "Algorithmic Game Theory" edited by Nisan et al. (computational aspects)
  • "The Theory of Learning in Games" by Fudenberg & Levine
  • "Auction Theory" by Krishna
  • "Matching, Mechanism Design, and Computation" by Nisan et al.
  • "Evolutionary Games and Population Dynamics" by Hofbauer & Sigmund

Applied and Behavioral

  • "Behavioral Game Theory" by Camerer
  • "Thinking Strategically" by Dixit & Nalebuff (popular introduction)
  • "The Art of Strategy" by Dixit et al. (accessible applications)

Online Courses

  • Stanford's Game Theory (Coursera) by Jackson, Shoham, Leyton-Brown
  • Yale's Open Course on Game Theory (Ben Polak)
  • MIT OCW: Game Theory with Applications
  • AGT (Algorithmic Game Theory) by Tim Roughgarden
  • Mechanism Design and Auctions (online courses)

Research Resources

  • Journal of Economic Theory
  • Games and Economic Behavior
  • International Journal of Game Theory
  • ACM Conference on Economics and Computation (EC)
  • Neural Information Processing Systems (NeurIPS, multi-agent track)

Interactive Tools

  • Game Theory Explorer (web-based)
  • Veconlab (experimental economics)
  • MobLab (classroom experiments)
  • Zocalo (prediction markets and mechanisms)