Scribd
Upload
8 views58 pages

Chapter 2

Chapter 2 discusses games with imperfect information, particularly focusing on Bayesian games where players lack complete knowledge about others' preferences. It introduces key concepts such as maxminimization, rationality, and strict Nash equilibrium, emphasizing how players form beliefs and make strategic decisions based on expected payoffs. The chapter illustrates these concepts through examples and the iterative elimination of irrational choices to identify rationalizable outcomes.

Uploaded by

MaX GaMeR
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
8 views58 pages

Chapter 2

Chapter 2 discusses games with imperfect information, particularly focusing on Bayesian games where players lack complete knowledge about others' preferences. It introduces key concepts such as maxminimization, rationality, and strict Nash equilibrium, emphasizing how players form beliefs and make strategic decisions based on expected payoffs. The chapter illustrates these concepts through examples and the iterative elimination of irrational choices to identify rationalizable outcomes.

Uploaded by

MaX GaMeR
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 58

Chapter 2

Games with Imperfect Information


 Games with imperfect information are those where player lacks full knowledge of the
game state at all times. This contrasts with perfect information games like chess,
where all past moves and positions are known to all players. Imperfect information often
introduces elements of uncertainty, deception, and probability.

 In this chapter we describe the model of a “Bayesian game”, which generalizes the
notion of a strategic game to allows us to analyze any situation in which each player is
imperfectly informed about some aspect of her environment relevant to her choice of an
action.
Example : Consider a variant of the situation modeled by BoS in which player 1 is unsure
whether player 2 prefers to go out with her or prefers to avoid her, whereas player 2,
as be fore, knows player 1’s preferences.

Solution:
The expected payoffs of player 1 for the four possible pairs of
actions of the two types of player 2
Finding the Best Strategy for Player 1
 Checking the Nash Equilibrium

The problem claims (B, (B, S)) is the equilibrium.


 Understanding Bayesian Games (Strategic
Games with Imperfect Information)
• A Bayesian game is a type of strategic game where players have imperfect
information about some aspects of the game, such as the preferences or types of other
players. The term "Bayesian" comes from the idea that players must form beliefs about
unknown factors using probabilities.

 Key Concepts in Bayesian Games


 Maxminimization

• The maxminimization (or maximin) criterion is a conservative strategy in game theory used
when you have no knowledge of your opponents' actions or strategies. It assumes that your
opponents will act in a way that minimizes your payoff, and your goal is to maximize the worst-
case outcome.

• How Maxminimization Works


 For each possible action you can take, identify the worst possible payoff you might receive. This is
found by assuming your opponent(s) will play the strategy that hurts you the most.
 Compare these worst-case payoffs across all your available actions.

 Choose the action that gives the highest of these worst-case payoffs.
Applying the Maxmin Criterion:

1.Find the worst-case payoff for each of your actions:

1. If you choose A, the worst case is -2 (opponent picks Y).

2. If you choose B, the worst case is 1 (opponent picks X).

2.Pick the action that gives the best worst-case payoff:

1. The worst-case for A is -2.

2. The worst-case for B is 1.

3. Since 1 is better than -2, you should choose B.


• This passage discusses a fundamental question in game theory: what outcomes in a strategic game are consistent
with players' analyses of each other's rational behavior? While Nash equilibrium captures a steady state where
players have learned each other's actions over time, this approach focuses on rational deduction—how players
anticipate their opponents’ moves based on preferences and reasoning.
• Rationality: A player chooses a strategy that maximizes expected payoff, given their belief about what the
other players will do.
• Probabilistic Beliefs: Each player forms a belief μᵢ about the strategies of others and assigns probabilities to
different actions they might take.
• Expected Utility Maximization: Given a mixed strategy αᵢ, a player's expected payoff is Uᵢ(αᵢ, a₋ᵢ),
determined by their own strategy and the opponents’ actions.

• This reasoning leads to Bayesian games and rationalizability—concepts that extend beyond Nash equilibrium by
considering beliefs and iterative reasoning about rational choices.
1. Rationality and Probabilistic Beliefs

Each player forms a belief about the strategies of others.


• A belief is a probability distribution over the opponents' possible actions.
• Players maximize expected payoff based on these beliefs.
• If Player 1 believes that Player 2 will play A with probability 0.9 and B with probability
0.1, Player 1’s expected payoffs are:
• Playing A: (0.9 × 3) + (0.1 × 0) = 2.7
• Playing B: (0.9 × 2) + (0.1 × 1) = 1.9
• Since 2.7 > 1.9, Player 1 will choose A.
• If Player 2 reasons similarly, she will also choose A.
• The outcome (A, A) is consistent with rational belief formation.
2. Rationalizability: Iterative Elimination of Irrational Choices

• Player 1 analyzes Player 2:


• If Player 2 is rational, she will never play Right, because choosing Left always gives her at least as much
payoff (1 vs. 0, 2 vs. 1).
• So Player 1 eliminates Right from his belief.
• Now Player 1’s decision is between Up (2) and Down (0), so he chooses Up.
• The outcome (Up, Left) is rationalizable.
• This process is called iterative elimination of strictly dominated strategies—it refines the set of plausible
outcomes.
Strict Nash Equilibrium
• A Strict Nash Equilibrium is a strategy profile where each player’s strategy is strictly better than any deviation, meaning
no player can improve their payoff by unilaterally changing their strategy.
Checking for Strict Nash Equilibrium:
A pair (s1,s2) is a Strict Nash Equilibrium if neither player benefits from deviating to another strategy.
1.(Small, Small) = (5,5)
1. If Beetle 1 switches to Large, it gets 8 instead of 5 (better).
2. If Beetle 2 switches to Large, it gets 8 instead of 5 (better).
3. Not a Strict Nash Equilibrium, because both players prefer to deviate.

2.(Small, Large) = (1,8)


1. Beetle 1 can switch to Large and get 3 instead of 1 (better).
2. Beetle 2 is already getting 8, so it does not change.
3. Not a Strict Nash Equilibrium, because Beetle 1 prefers to deviate.
3. (Large, Small) = (8,1)
1. Beetle 1 is already maximizing its payoff, so it does not deviate.
2. Beetle 2 can switch to Large and get 3 instead of 1 (better).
3. Not a Strict Nash Equilibrium, because Beetle 2 prefers to deviate.
4. (Large, Large) = (3,3)
4. If Beetle 1 switches to Small, it gets 1 instead of 3 (worse).
5. If Beetle 2 switches to Small, it gets 1 instead of 3 (worse).
6. Strict Nash Equilibrium, because neither player benefits from deviating

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy