The whole elephant
Game theory has become a common language for strategic interaction in computer science, communication networks, energy systems and markets, control, and artificial intelligence. Most textbooks, however, approach it either from economics or from a single application domain, so students from different backgrounds often learn one part of the subject well and miss the rest, like the blind men in the parable who each describe the elephant from the one part they touch.
This book brings the many aspects of game theory together in one framework and one notation: solution concepts and the existence, uniqueness, and refinement of equilibria; computation; incomplete information and mechanism design; bargaining, coalition, and matching games; network, routing, and population games; learning and evolutionary dynamics; and dynamic games in discrete and continuous time.
Dynamic games at the center
Real interactions are rarely one-shot. They unfold over many rounds, in environments that change over time, and what a player knows, and when, can change the nature of an equilibrium. Repeated, Markov, differential, stochastic, Stackelberg, and mean-field games are a core part of the book, not an advanced supplement.
For multi-agent systems and control
Students of multi-agent systems, now often including AI agents that learn through repeated interaction, need this dynamic view. So do students of systems and control, for whom a dynamic game is a control problem with several controllers, each pursuing its own objective. Hence the subtitle.
Shaped in the classroom
The lectures grew out of game theory courses taught for several decades at Illinois and for more than a decade at NYU, to students from engineering, computer science, mathematics, economics, and beyond. Every chapter closes with further notes and reading, and with exercises.
Thirty chapters in seven parts
Part I is the core of the book, and the other parts can be taught in different combinations. Part I is available now as a free sample: fundamentals through correlated equilibrium, with the Preface and the full contents, 270 pages. Three appendices review optimal control and the mathematical background and collect the notation.
IFundamentals of Game Theoryin sample
- 1Introduction to Game Theory
- 2Matrix Games
- 3Nash Equilibrium: Concepts and Definition
- 4Nash Equilibrium: Properties and Computation
- 5Extensive-Form Games
- 6Stackelberg (Leader-Follower) Games
- 7Equilibrium Refinement
- 8Correlated Equilibrium
IIContinuous-Kernel Nash and Stackelberg Games
- 9Infinite and Continuous-Kernel Games
- 10Continuous-Kernel Stackelberg Games
- 11Quadratic and Network Games
IIIGames with Incomplete Information
- 12Bayesian Games
- 13Auction Games
- 14Revenue-Optimal Auction Design
- 15Mechanism Design Theory
IVDynamic and Repeated Games
- 16Extensive-Form Dynamic Games
- 17Principal-Agent Games
- 18Repeated Games
- 19Markov Games
- 20Dynamic Infinite Games
VDifferential and Stochastic Games
- 21Differential Games
- 22Dynamic Stackelberg Games
- 23Stochastic Differential Games
- 24Mean-Field Games
VICooperative and Matching Games
- 25Nash Bargaining
- 26Coalition Games
- 27Matching Games
VIIAdvanced Topics
- 28Routing Games
- 29Evolutionary Games
- 30Learning in Games
A·B·CAppendices
- AElements of Optimal Control Theory
- BMathematical Background
- CNotation
Slides, code, and corrections
This is the book's companion website, for material that does not fit between the covers. Items are posted here as they become ready, and further supplementary material will be added over time.
Lecture slides
Slides for each lecture, following the book chapter by chapter, for students and for instructors teaching from it.
Code for the figures
The scripts used to compute and draw the book's plots, so that readers can reproduce them and vary the parameters.
Corrections
A list of corrections to the book, updated as errors are found. None are posted yet. Please send any error you find to qz494@nyu.edu.
Sample chapters
Part I, Chapters 1 to 8, from the introduction through correlated equilibrium. Download the PDF (4.2 MB).
Where this material is taught and discussed
Instructors who use the book in a course are welcome to write to us.
University of Illinois Urbana-Champaign
The game theory courses taught by Tamer Başar, in which the lectures first took shape.
NYU Tandon · ECE-GY 6263 Game Theory
The graduate course taught by Quanyan Zhu. The Part I sample doubles as the course's opening chapters.
Dynamic games community
Executive Board of the International Society of Dynamic Games; Associate Editor of Dynamic Games and Applications (Quanyan Zhu).
GameSec
Chair of the 15th Conference on Decision and Game Theory for Security (New York, 2024), and program chair and organizer in earlier years (Quanyan Zhu).
Game theory at NYU
The theory in the book is the working language of my group: designing systems whose participants, human or artificial, pursue their own objectives, and proving what those systems will do.
Games and control of socio-technical systems
Game-theoretic control as a design discipline for systems where people, institutions, and machines interact.
- Revisiting Game-Theoretic Control in Socio-Technical Networks · Zhu and Başar, IEEE L-CSS 2025 · DOI
- Disentangling Resilience from Robustness · Zhu and Başar, IEEE Control Systems Magazine 2024 · DOI
- Cooperative and Noncooperative Paradigms for Game-Theoretic Control · Başar, Hayakawa, Ishii, and Zhu · arXiv
Security games and deception
Attacker–defender games, defensive deception, and automated penetration testing.
- MEGA-PT: A Meta-Game Framework for Agile Penetration Testing · Ge and Zhu, GameSec 2024, Best Paper · DOI
- Bilateral Cognitive Security Games under Stealthy Injection Attacks · Nguyen, Zhu, and Teixeira, CDC 2025 · DOI
- More on the deception project site
Learning, conjectures, and misspecification
What agents converge to when their models of the world, and of each other, are wrong.
Coalitions, bargaining, and mean-field limits
Cooperative game theory for large populations of agents that join, leave, split, and merge.
Mechanism and information design
Designing the rules, the contracts, and the signals rather than the actions.
- Stackelberg Risk Preference Design · Liu and Zhu, Mathematical Programming 2025 · DOI
- Transparent Tagging for Strategic Social Nudges on Misinformation · Yang, Li, and Zhu, IEEE TNSE 2025 · DOI
- Herd Accountability of Privacy-Preserving Algorithms: A Stackelberg Game Approach · Yang, Zhang, and Zhu, IEEE TIFS 2025 · DOI
Games for agentic AI
Language models and AI agents as strategic players: reasoning, pricing, and coordination.
- Reasoning and Behavioral Equilibria in LLM-Nash Games · Zhu · arXiv
- PACT: A Contract-Theoretic Framework for Pricing Agentic AI Services · Yang and Zhu, GLOBECOM 2025 · DOI
- More on the Agentic AI and AI Tokenomics sites
