When do Prices
Coordinate
Markets?
Aaron Roth
Based on joint work with:
Justin Hsu, Jamie Morgenstern, Ryan Rogers, and Rakesh Vohra
Prices are remarkable
Markets are decentralized.
Individuals observe prices, and buy bundles of
goods that optimize thei
NETS 412: Algorithmic Game Theory
February 23, 2017
Lecture 12
Lecturer: Aaron Roth
Scribe: Aaron Roth
Stable Matchings
In this lecture, well consider a model of 1950s dating. Although this is the metaphor we will use,
stable matchings are an extremely us
NETS 412: Algorithmic Game Theory
March 28 and 30, 2017
Lecture 16+17
Lecturer: Aaron Roth
Scribe: Aaron Roth
Approximation in Mechanism Design
In the last lecture, we asked how far we can go beyond the VCG mechanism when we want to optimize
an objective
NETS 412: Algorithmic Game Theory
April 20, 2017
Lecture 23
Lecturer: Bo Waggoner
Scribe: Bo Waggoner
Proper Scoring Rules and Prediction Markets
Today we look at the question: How can we incentivize an agent or group of agents to make an
accurate predict
NETS 412: Algorithmic Game Theory
February 2, 2017
Lecture 6
Lecturer: Aaron Roth
Scribe: Aaron Roth
Zero Sum Games and the MinMax Theorem
In this lecture we study zero sum games, which have very special mathematical and computational
properties. They are
NETS 412: Algorithmic Game Theory
March 2, 2017
Lecture 13
Lecturer: Aaron Roth
Scribe: Aaron Roth
Walrasian Equilibrium
In this lecture, we bring money into the picture, while still thinking about a matching like problem.
Suppose we have:
1. We have m go
NETS 412: Algorithmic Game Theory
February 16, 2017
Lecture 10
Lecturer: Aaron Roth
Scribe: Aaron Roth
The Price of Anarchy and Stability
Up until now, we have been focussing on how agents playing together in a game, in a decentralized
manner, might arriv
NETS 412: Algorithmic Game Theory
February 21, 2017
Lecture 11
Lecturer: Aaron Roth
Scribe: Aaron Roth
Truthful, Pareto Optimal Exchange Without Money
This lecture begins the second half of the course: up until now, we have studied the behavior of
individ
NETS 412: Algorithmic Game Theory
January 24, 2017
Lecture 3
Lecturer: Aaron Roth
Scribe: Aaron Roth
When do Best Response Dynamics Converge?
In this lecture, we ask how far we can go beyond congestion games while still being certain that best
response dy
February 14, 2017
NETS 412: Algorithmic Game Theory
Lecture 9
Lecturer: Aaron Roth
Scribe: Aaron Roth
Achieving No Swap Regret
Recall that we argued last lecture that in any game, if every player plays according to the polynomial
weights algorithm, then
NETS 412: Algorithmic Game Theory
March 21 2017
Lecture 14
Lecturer: Aaron Roth
Scribe: Aaron Roth
Auction Design
Last lecture we studied pricing equilibria. In this lecture, we continue our study of money as a means
of exchange, from the perspective of m
NETS 412: Algorithmic Game Theory
February 9, 2017
Lecture 8
Lecturer: Aaron Roth
Scribe: Aaron Roth
Correlated Equilibria
Consider the following two player traffic light game that will be familiar to those of you who can drive:
STOP
GO
STOP
(0,0)
(1,0)
G
NETS 412: Algorithmic Game Theory
January 12, 2017
Lecture 1
Lecturer: Aaron Roth
Scribe: Aaron Roth
Basic Definitions
In this class we introduce some of the basic definitions we will be using throughout the semester. First,
what is a game?
Definition 1 A
NETS 412: Algorithmic Game Theory
April 4, 2017
Lecture 18
Lecturer: Aaron Roth
Scribe: Aaron Roth
Profit Maximization, Digital Goods, and the Random Sampling Auction
In previous lectures, we have extensively studied auction design for welfare maximizatio
NETS 412: Algorithmic Game Theory
January 26 and 31, 2016
Lecture 4 and 5
Lecturer: Aaron Roth
Scribe: Aaron Roth
The Polynomial Weights Algorithm
In the last several lectures, we have seen several games in which best response dynamics inevitably
converge
NETS 412: Algorithmic Game Theory
March 23, 2017
Lecture 15
Lecturer: Aaron Roth
Scribe: Aaron Roth
Auction Design in Single Parameter Domains
Last lecture, we saw the VCG mechanism, which has a tremendous number of nice features, and achieves
them all P
NETS 412: Algorithmic Game Theory
February 7, 2017
Lecture 7
Lecturer: Aaron Roth
Scribe: Aaron Roth
Convergence of No Regret Dynamics to Equilibrium in Separable
Multi-player Zero Sum Games
Last class we saw that two-player zero sum games are special. Am
NETS 412: Algorithmic Game Theory
April 11, 2017
Lecture 20
Lecturer: Aaron Roth
Scribe: Aaron Roth
Dynamic Pricing: Profit Maximization From Bandit Feedback
In the last lecture, we thought about running an auction in the online setting, in which buyers a
NETS 412: Algorithmic Game Theory
January 17, 2017
Lecture 2
Lecturer: Aaron Roth
Scribe: Aaron Roth
Congestion Games
In general, to represent an n player game in which each player has k actions, we need k n numbers just
to encode the utility functions. C
NETS 412: Algorithmic Game Theory
April 13, 2017
Lecture 21
Lecturer: Aaron Roth
Scribe: Aaron Roth
Mechanism Design via Differential Privacy
In this class, we continue using digital goods auctions with valuations vi [0, 1] as a case study for
another tec
NETS 412: Algorithmic Game Theory
April 6, 2015
Lecture 19
Lecturer: Aaron Roth
Scribe: Aaron Roth
Profit Maximization in Online Auctions
In this lecture, well bring the class full circle. Well consider a variant of the problem we considered
in the last l
Algorithmic Game Theory: Problem Set 3
Due online via GradeScope before the start of class on Tuesday, February 21
Aaron Roth
Remember you can work together on problem sets, but list everyone you worked with, and everyone turn
in their own assignment. Ask
Algorithmic Game Theory: Problem Set 1
Due online via GradeScope before the start of class on Tuesday, January 24
Aaron Roth
Collaboration on problem sets is ok, but list everyone you worked with, and everyone must turn in their
own assignment. Ask questi
Algorithmic Game Theory: Problem Set 2
Due online via GradeScope before the start of class on Tuesday, February 7
Aaron Roth
Remember you can work together on problem sets, but list everyone you worked with, and everyone turn
in their own assignment. Ask
Algorithmic Game Theory: Problem Set 4
Due online via GradeScope before the start of class on Tuesday, March 14
Aaron Roth
Remember you can work together on problem sets, but list everyone you worked with, and everyone turn
in their own assignment. Ask qu
Algorithmic Game Theory: Problem Set 6
Due on Tuesday, April 18
Aaron Roth
Remember you can work together on problem sets, but list everyone you worked with, and everyone turn
in their own assignment.
Efficient Implementation of VCG (15 pts)
Consider an a
Algorithmic Game Theory: Problem Set 5
Due on Tuesday, April 4
Aaron Roth
Remember you can work together on problem sets, but list everyone you worked with, and everyone turn
in their own assignment.
Walrasian Equilibrium (20 pts)
1. We saw in lecture 13
MKSE Summary
1. Network is artificial and is the aggregate of all hobbies/interests/activities pairwise connections
2. Internet Router:
a. Points are physical machines (PC, Ipad),
b. Links are physical wire or wireless,
c. Interaction is electronic (signa
The scaling laws of human travel
Identity and search in social networks
Navigation in a small world
An Experimental Study of Search in Global Social Networks
An experimental study of the small world problem
Random people selected in Boston and Nebraska t
Structural Properties of Networks
- Distance between two nodes in different connected components is infinity.
- Diameter of a network is the average difference between pairs
- Hubs - high degree individuals
- Network Structure vs Network Dynamics vs Netwo