Advanced Game Theory - PowerPoint PPT Presentation

1 / 18
About This Presentation
Title:

Advanced Game Theory

Description:

Commitment: Ensure you can credibly claim a share of that value ... Embarrassment. Inefficiency. New Zealand UHF License Auction ... – PowerPoint PPT presentation

Number of Views:31
Avg rating:3.0/5.0
Slides: 19
Provided by: joshu7
Category:

less

Transcript and Presenter's Notes

Title: Advanced Game Theory


1
Advanced Game Theory
  • Introduction

2
Two Dimensions of Game Theory
  • Strategy
  • Design

3
Strategy
  • Stand in shoes of other players
  • Key issues
  • Cooperation Realise opportunities to create
    value
  • Commitment Ensure you can credibly claim a share
    of that value
  • Coordination signal actions to others
  • Qualitative/Robustness Focus

4
Design
  • Stand in the shoes of agents
  • Key Issues
  • Incentives aligning agents interests with your
    goals
  • Information getting agents to reveal private
    information
  • Implementation practicalities of dealing with
    complexity
  • Quantitative/Details Focus

5
Housekeeping
  • Assessment
  • Syndicates
  • Project
  • Readings
  • Problem Sets
  • Web Page

6
Concept Review
  • Rational Choice
  • Statistical Concepts
  • Noncooperative Game Theory
  • Set-up payoffs, players, actions
  • Dominated strategies
  • Nash equilibrium
  • Backwards induction
  • Repeated interactions

7
Incentives
  • May desire a particular action
  • Engage in effort
  • Reveal information
  • Design issue is to design a mechanism that makes
    it a dominant strategy to undertake the desired
    action

8
Case
  • Corporate Raiding

9
Information
  • Have to act with incomplete information
  • If rational, recognise own ignorance and
    recognise ignorance of others
  • Think about what others know and dont know and
    what they know about what they know and dont
    know (e.g., Holmes and Moriarty)
  • Assume agents have common knowledge
  • An event is common knowledge among a group of
    players if each one knows it, if each one knows
    that the others know it, if each one knows that
    each other one knows that the others know it, and
    so on.
  • Common knowledge of public events, rationality
    etc.

10
Exercise
  • Hat Game

11
Common Knowledge in Hat Game
  • It is common knowledge that
  • Everybody can see two hats
  • Public pronouncements of ignorance
  • Knowledge of others reasoning
  • Later common knowledge that at least one hat is
    red (initially it was knowledge but not common
    knowledge)

12
Envelope Game
  • Two envelopes, each with money
  • Either 5, 10, 20, 40, 80 or 160
  • Common knowledge that one has twice as much as
    the other.
  • Give one to A and the other to B
  • Open private knowledge of contents
  • Give opportunity to switch and if both A and B
    want to they are allowed to
  • Unless have 160 looks like each is better off
    switching but how can that be?

13
Email Game
  • A and B are arranging meeting up in another city.
    They email each other about a location because if
    they fail to meet up they would be better off not
    traveling.
  • Suppose A emails B a location but there is a
    probability, p, the email is not received (there
    is no message sent that a message is not
    received). So A cant be sure B knows where to
    go.
  • A could ask for confirmation. But that
    acknowledgement may not get through. A and B know
    where to meet but B does not know if A knows that
    B knows.
  • Knowledge has increased but what about
    coordination?

14
Is Email Useful?
  • As choices
  • If travels regardless of acknowledgement, is
    ignoring that and are in situation B has sent no
    acknowledgment.
  • Should not travel since it was more likely
    original message was lost (p) than
    acknowledgement lost (p(1-p)).
  • Therefore, only travel if get acknowledgement
  • Bs choice
  • Should travel if got message but not otherwise.
  • Outcomes
  • Coordinate with probability (1-p)2
  • B travels alone with probability (1-p)p
  • No travel at all with probability p
  • Acknowledgement does not and never will help

15
Implementation
  • In design, the details are critical.
  • Failure to take into account all of the rules can
    lead to
  • Embarrassment
  • Inefficiency

16
New Zealand UHF License Auction
  • A simultaneous sealed-bid second price auction

17
Australian Satellite TV Auction
  • A sealed bid auction with no withdrawal penalty

Initial Winning Bid
Final Transaction Price
212,000,000
117,000,000
177,000,000
77,000,000
18
From here
  • Matching markets
  • Mechanism Design
  • Nonlinear Pricing
  • Auctions and Bidding
  • Voting and Institutions
Write a Comment
User Comments (0)
About PowerShow.com