Course 1
COURSE 1 · FOUNDATIONS OF GAME THEORY
Contents
Every tutorial in the course, topic by topic. Within a topic the tutorials are listed alphabetically.
← Back to Course 1 — Foundations of Game Theory
Topic 1 · Foundations
Equilibrium concepts, information, repeated games, complexity — the formal core. Topic page →
| Title | Description |
|---|---|
| Backward induction and the centipede game paradox | Implement backward induction for the centipede game in R, demonstrate the dramatic gap between theory and observed play, and analyse how finite vs infinite horizons shape rational behaviour. |
| Common knowledge and its paradoxes | Formalise common knowledge using Aumann’s partition model in R, implement the muddy children puzzle to show how public announcements create common knowledge, and analyse how failures of common knowledge lead to coordination failures, speculative trade, and the electronic mail game paradox. |
| Correlated equilibrium — coordination through shared signals | Implement Aumann’s correlated equilibrium in R via linear programming, compare it with Nash equilibrium, and visualize how a public correlating device expands the set of achievable payoffs. |
| Dominant strategies and iterated elimination of dominated strategies | Define strict and weak dominance in normal-form games, implement iterated elimination of dominated strategies (IEDS) in R, and show how rationality assumptions progressively simplify strategic analysis. |
| Epsilon-equilibrium and approximate Nash equilibria | Implement epsilon-Nash equilibrium in R, analyse when approximate solutions suffice for practical predictions, compute the epsilon required for various games, and connect to computational complexity of exact Nash equilibrium finding. |
| Extensive-form games and subgame-perfect equilibrium | Formalize sequential games as extensive-form game trees in R, implement backward induction to find subgame-perfect equilibria, and visualize the difference between Nash equilibria and credible equilibria. |
| Folk theorem with perfect monitoring | Prove and visualize the folk theorem for infinitely repeated games: any individually rational, feasible payoff can be sustained as a subgame-perfect equilibrium when players are sufficiently patient. |
| Mixed-strategy Nash equilibrium in 2×2 games | How to compute a mixed-strategy Nash equilibrium for a 2×2 normal-form game, with R and Python implementations and interactive visualizations. |
| Rationalizability and iterated dominance — the epistemic foundation of game theory | Define rationalizable strategies as those surviving iterated deletion of never-best-responses, implement the algorithm in R for arbitrary games, and show the relationship to IESDS and common knowledge of rationality. |
| Signaling games and perfect Bayesian equilibrium | Model strategic communication under asymmetric information in R, compute perfect Bayesian equilibria with separating and pooling strategies, and visualize how informed players signal their types. |
| The Nash bargaining solution | Derive and implement Nash’s axiomatic bargaining solution in R, compute the Nash product for two-player bargaining problems, and visualize the Pareto frontier with the bargaining outcome. |
| The revelation principle — why truth-telling mechanisms suffice | Prove and implement the revelation principle in R, showing that any outcome achievable by any mechanism can be replicated by a direct mechanism where truthful reporting is optimal, with applications to auctions and public goods provision. |
| Trembling Hand Perfect Equilibrium | Implement Selten’s trembling hand perfection refinement to eliminate unreasonable Nash equilibria by requiring robustness to small probability mistakes, with applications to coordination games and weakly dominated strategies. |
| Zero-sum games and the minimax theorem | Prove and implement von Neumann’s minimax theorem for two-player zero-sum games in R, compute minimax strategies via linear programming, and visualize the security level geometry. |
Topic 2 · Classical Games
Prisoner’s Dilemma, Hawk-Dove, oligopoly, bargaining, coordination — the canon. Topic page →
| Title | Description |
|---|---|
| Bank runs as a coordination game — the Diamond-Dybvig model | Model bank runs as a coordination game using the Diamond-Dybvig framework in R, showing the Stag Hunt structure, deposit insurance as equilibrium selection, and contagion dynamics with N depositors. |
| Battle of the Sexes — coordination with conflicting preferences | Analyse the Battle of the Sexes game formally in R, find all Nash equilibria (two pure + one mixed), and visualize the coordination problem with best-response correspondences. |
| Chicken / Hawk-Dove — brinkmanship and anti-coordination | Analyse the Chicken (Hawk-Dove) game in R, computing pure and mixed Nash equilibria, exploring the evolutionary stable strategy, and connecting to models of escalation and conflict. |
| Matching Pennies — Theory Meets Experimental Evidence | Implement the classic zero-sum matching pennies game with its unique mixed Nash equilibrium, then simulate experimental play using bounded rationality models including epsilon-greedy, fictitious play, and reinforcement learning to study convergence and exploitable patterns in R. |
| Matching Pennies — the simplest zero-sum game | Analyse Matching Pennies as the canonical two-player zero-sum game with a unique mixed-strategy Nash equilibrium, implementing minimax strategies and expected payoff calculations in R. |
| The Colonel Blotto game — resource allocation across battlefields | Analyse the Colonel Blotto game in R, exploring optimal resource allocation across multiple battlefields, mixed-strategy equilibria, and applications to political campaigns and military strategy. |
| The El Farol Bar Problem | Simulating Arthur’s (1994) El Farol Bar problem with heterogeneous prediction strategies, analysing attendance oscillations, and connecting to minority games and market ecology. |
| The Prisoner’s Dilemma — formal setup and dominant strategy analysis | Define the one-shot Prisoner’s Dilemma formally, prove that mutual defection is the unique dominant-strategy equilibrium, and explore how payoff parameters affect the tension between individual and collective rationality. |
| The Sealed-Bid War of Attrition | Analyse the sealed-bid war of attrition as an all-pay second-price auction, derive the symmetric Bayesian Nash equilibrium with exponential bidding strategies for uniform private values, and compare revenue and surplus with the standard all-pay auction. |
| The Stag Hunt — coordination, trust, and risk dominance | Analyse the Stag Hunt as a coordination game with Pareto-ranked equilibria in R, comparing payoff dominance and risk dominance, with basin-of-attraction analysis. |
| The game of Chicken and the art of brinkmanship | Analyse the Chicken game in R, compute all Nash equilibria including the mixed-strategy equilibrium, model brinkmanship as a commitment device, and apply to nuclear deterrence, traffic merging, and corporate competition. |
| The iterated Prisoner’s Dilemma — Axelrod’s tournaments | Replicate Robert Axelrod’s classic 1980 computer tournaments for the iterated Prisoner’s Dilemma in R, with round-robin scoring, strategy comparison, and evolutionary dynamics. |
Topic 3 · Decision Theory
Expected utility, prospect theory, ambiguity, sequential and group decision. Topic page →
| Title | Description |
|---|---|
| Ambiguity aversion and the Ellsberg paradox | Implement the Ellsberg urn experiment in R, model ambiguity aversion using maxmin expected utility and alpha-MEU, and show how ambiguity affects strategic decisions and games with unknown opponent types. |
| Expected Utility Theory: The Axiomatic Foundations | Implementing the von Neumann-Morgenstern axioms in R, verifying them with lottery objects, demonstrating the Allais and Ellsberg paradoxes, and calibrating CRRA and CARA utility functions via MLE. |
| Expected utility theory and the von Neumann–Morgenstern axioms | Derive expected utility from the VNM axioms in R, demonstrate the representation theorem with numerical examples, and visualize risk attitudes through utility function curvature. |
| Reference dependence and loss aversion in games — how prospect theory reshapes equilibria | Recompute Nash equilibria when players evaluate payoffs through prospect theory’s lens of reference dependence and loss aversion, and visualize how equilibrium structure shifts in the Stag Hunt as loss aversion varies. |
| Savage’s Subjective Expected Utility: From Axioms to Beliefs | Implement Savage’s seven axioms of subjective expected utility, derive subjective probabilities from revealed preferences, and demonstrate the Ellsberg paradox as a violation of the sure-thing principle. |
| The Allais paradox and expected utility violations | Demonstrate the Allais paradox as a systematic violation of the independence axiom in expected utility theory, implement EU calculations in R, and explore prospect theory’s probability weighting as an explanation. |
Topic 4 · Evolutionary Game Theory
Replicator dynamics, ESS, Moran processes, spatial games, co-evolution. Topic page →
| Title | Description |
|---|---|
| Evolutionarily stable strategies (ESS) | Define and compute evolutionarily stable strategies for symmetric 2x2 games, apply to Hawk-Dove, Stag Hunt, and Prisoner’s Dilemma, and visualize invasion dynamics and fitness landscapes. |
| Hawk-Dove and the war of attrition — escalation in evolutionary games | Model the continuous-time war of attrition as an extension of the Hawk-Dove game in R, derive the mixed ESS with exponential waiting times, and visualise how resource value and fighting costs shape escalation behaviour. |
| Moran process in finite populations | Implement and analyze the Moran process for frequency-dependent selection in finite populations, computing fixation probabilities analytically and via simulation for 2-strategy evolutionary games. |
| Replicator dynamics for Rock-Paper-Scissors with deSolve | Derive and simulate the replicator dynamics for the Rock-Paper-Scissors game in R using deSolve, visualizing cyclic orbits on the 2-simplex. |
| Replicator-mutator dynamics — evolution with mutation | Extend the standard replicator equation with a mutation matrix Q, implement the ODE system using deSolve in R, and show how mutation stabilises the interior equilibrium of Rock-Paper-Scissors. |
| Spatial Evolutionary Games: Cooperation on Grids and Networks | Implement evolutionary dynamics on a 2D lattice where agents play the Prisoner’s Dilemma with neighbours, showing how spatial structure enables cooperator clusters to survive even when defection dominates in well-mixed populations. |
Topic 5 · Cooperative Game Theory
Shapley value, core, nucleolus, coalition formation, cost allocation. Topic page →
| Title | Description |
|---|---|
| Core stability in cooperative games — unblocked allocations and the Bondareva-Shapley theorem | Define the core of a cooperative game, check non-emptiness via linear programming and the Bondareva-Shapley theorem, compute core allocations for classic games, and visualize the core as a region in the simplex for three-player games. |
| Hedonic coalition formation games | Implement hedonic coalition formation with individual and Nash stability concepts in R, applied to a roommate matching example with 6 players. |
| Rubinstein’s alternating-offers bargaining game | Implement Rubinstein’s infinite-horizon alternating-offers model in R, derive the unique subgame-perfect equilibrium as a function of discount factors, and visualise how patience determines bargaining power. |
| The Shapley value — fair division of cooperative surplus | Derive the Shapley value axiomatically, implement it for arbitrary coalitional games in R, and visualize marginal contributions and fairness properties with interactive examples. |
| The Shapley-Shubik power index for weighted voting games | Measure each voter’s true power in weighted voting games using the Shapley-Shubik index, and discover why voting weight and voting power are fundamentally different. |
| The bankruptcy game and the Talmudic rule | Implement the contested garment principle and the Talmudic bankruptcy rule in R, show their connection to the nucleolus of cooperative games, and compare with proportional and constrained equal awards division methods. |
| Voting power indices — when weight does not equal power | Compute the Banzhaf and Shapley-Shubik power indices for weighted voting games in R, apply them to the EU Council, UN Security Council, and US Electoral College, and visualize the divergence between voting weight and true voting power. |
Topic 6 · Behavioral Game Theory
Bounded rationality, QRE, level-k, social preferences, learning. Topic page →
| Title | Description |
|---|---|
| Fehr-Schmidt inequality aversion: modelling fairness in games | Implement the Fehr-Schmidt model of inequality aversion, where players experience disutility from both advantageous and disadvantageous inequality, and apply it to the ultimatum, dictator, and public goods games. |
| Level-k thinking and cognitive hierarchy — bounded rationality in strategic games | Implement the level-k model and Camerer-Ho-Chong cognitive hierarchy for the p-beauty contest, simulate heterogeneous populations, and visualize convergence to Nash equilibrium as strategic sophistication increases. |
| Quantal response equilibrium — bridging rationality and noise | Implement quantal response equilibrium (QRE) in R, derive the logit equilibrium for 2x2 games, trace the QRE correspondence from uniform randomisation to Nash equilibrium, and fit QRE to experimental data from coordination and asymmetric games. |
| The public goods game with punishment — sustaining cooperation through costly enforcement | Implement the public goods game with and without costly punishment in R, simulate contribution dynamics with heterogeneous agent types, and reproduce the Fehr and Gaechter finding that punishment sustains cooperation where standard theory predicts free-riding. |
| The ultimatum game — fairness, rejection, and bounded rationality | Analyse the ultimatum game in R, contrasting the subgame-perfect equilibrium prediction with experimental evidence on fairness norms, and implementing inequality aversion models that explain rejection behaviour. |
Topic 7 · Network Games
Strategic network formation, diffusion, influence, centrality as power. Topic page →
| Title | Description |
|---|---|
| Braess’s paradox — when adding roads makes traffic worse | Implement Braess’s paradox in R, compute Wardrop equilibrium flows on networks with selfish routing, demonstrate how adding capacity can increase total latency, and compare the price of anarchy across network topologies. |
| Congestion games and potential functions | Implement Rosenthal’s congestion games with Monderer-Shapley potential functions, demonstrate the existence of pure Nash equilibria, and visualise Braess’s paradox in a traffic network. |
| Key player identification in network games | Identify the most influential players in network games using Ballester-Calvo-Zenou intercentrality, analysing network effects on equilibrium play, optimal targeting policies, and applications to crime and R&D networks. |
| Public goods provision on networks | Public goods games played on networks where contributions benefit neighbours: how network topology — star, circle, and random graphs — shapes equilibrium provision, free-riding by central agents, and aggregate welfare through best-response dynamics on adjacency matrices. |
| Strategic network formation — who links with whom? | Implement the Jackson-Wolinsky connections model in R, analyse pairwise stability and efficiency of network structures, and visualise the tension between stable and efficient networks using igraph. |
| Two-sided markets and platform competition | Model platform markets as two-sided network games with cross-side externalities, analyzing pricing structure, platform competition, tipping, and multihoming. |
Topic 8 · History of Game Theory
From Cournot 1838 through von Neumann, Nash, Selten, to modern Nobel laureates. Topic page →
| Title | Description |
|---|---|
| From von Neumann to Nash — a mathematical timeline of game theory | Interactive timeline of key results from the 1928 minimax theorem through Nash’s 1950-51 equilibrium existence proof, with R visualizations of each milestone’s mathematical contribution to game theory. |
| Nash’s existence proof — historical context and intuition | A historical and mathematical account of John Nash’s 1950 proof that every finite game has a mixed-strategy equilibrium, with visual demonstrations of best-response correspondences, fixed points, and connections to computational complexity. |
| Nash’s existence theorem: fixed-point proof and computation | Trace Nash’s 1950 existence proof for equilibria in finite games using Kakutani’s fixed-point theorem, implement the best-response correspondence in R, visualise the fixed-point mapping for 2×2 games, and connect the proof structure to modern equilibrium computation. |
| Selten’s trembling hand perfection — historical and technical | A historical and mathematical account of Reinhard Selten’s 1975 trembling hand perfect equilibrium refinement, with R implementations that verify perfection by constructing sequences of completely mixed strategy profiles converging to candidate equilibria. |
| Von Neumann’s minimax theorem: proof and implementation | A rigorous treatment of von Neumann’s 1928 minimax theorem for two-person zero-sum games, including a constructive proof via linear programming duality, a full R implementation using lpSolve, and interactive visualizations of the minimax solution surface. |
| Zermelo’s Theorem – The First Formal Result in Game Theory | Explore Zermelo’s 1913 theorem on determined games, implementing backward induction for Nim and simplified chess endgames in R, with Sprague-Grundy theory and Grundy value computation. |