Scribd
Upload
96 views

Computer Network Assignment Help: Problems

This document provides solutions to 5 problems related to game theory and network effects. Problem 1 involves deriving the best response correspondence for a game with network effects where the value of purchasing a product depends on the fraction of the population that purchases. Problem 2 analyzes another market with network effects. Problems 3 and 4 analyze equilibrium concepts in different game theoretic models - a local network game with strategic substitutes and a prisoner's dilemma game. Problem 5 examines equilibria in a 2-player, 2-stage game and determines the highest welfare attainable in a subgame perfect equilibrium for the infinitely repeated version of the game.

Uploaded by

Computer Network Assignment Help
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
96 views

Computer Network Assignment Help: Problems

This document provides solutions to 5 problems related to game theory and network effects. Problem 1 involves deriving the best response correspondence for a game with network effects where the value of purchasing a product depends on the fraction of the population that purchases. Problem 2 analyzes another market with network effects. Problems 3 and 4 analyze equilibrium concepts in different game theoretic models - a local network game with strategic substitutes and a prisoner's dilemma game. Problem 5 examines equilibria in a 2-player, 2-stage game and determines the highest welfare attainable in a subgame perfect equilibrium for the infinitely repeated version of the game.

Uploaded by

Computer Network Assignment Help
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 7

For any Assignment related queries, Call us at : -  

+1 678 648 4277


You can mail us at : - [email protected] or
reach us at : - www.computernetworkassignmenthelp.com/

Computer Network Assignment Help


Problems:
Problem 1: Suppose there is a unit mass of consumers who can purchase a product at some cost c.
Each consumer i has a type vi drawn independently from a continuous distribution F on [0,1]. The
good exhibits network effects. In particular, if a fraction x of the population purchases, then the
value of purchasing to consumer i is

The value of not purchasing is zero.


(a) Suppose h is nonnegative, continuous, and strictly increasing on [0, 1]. Derive the best
response correspondence for this game—that is, fixing h(x), what fraction of the population xˆ
wishes to purchase? Prove that an equilibrium exists.
(b) Suppose h(x) = x and the cumulative distribution function F for vi takes the form

for some α > 0 and 0 < γ < β. Derive the best response correspondence. For what parameter
values does there exist an equilibrium in which consumers purchase the product?
Problem 2: Consider another market with network effects. There is a unit mass of potential
consumers who can purchase a product at some fixed price p with 0 < p <
Each consumer i has a private value vi drawn independently from a uniform distribution on [0,
1]. If a fraction x of the population purchases the product, the consumer’s payoff from
purchasing is
(a) Intuitively describe the kind of externalities g(x) is capturing. Can
you give a real life example that would fit?
(b) Characterize the set of equilibria.
(c) Which equilibria are stable? Why?
(d) Is social welfare maximized in any of the equilibria? Explain.
Problem 3: Consider the local network game with strategic substitutes from
the lecture slides. Each player i chooses an action xi ≥ 0 and earns the payof

Problem 4: Consider the following version of the Prisoner’s dilemma game:


Assume b is such that b0 (1) = k. Suppose G is a circle graph with four players.
Compute the set of equilibria.

In an infinitely repeated version of this game, with discount rate δ, can you
construct a subgame perfect equilibrium in which the players trade off cooperating
and defecting on one another? That is, on the equilibrium path, in period 1 the
action profile is (C, D), in period 2 it is (D, C), in period 3 it is (C, D), and so on.
How high must δ be for this to be an equilibrium? How does welfare in this
equilibrium compare to the equilibrium with cooperation in every period (when this
is an equilibrium)?
Problem 5: Consider the following game:

(a) In a one-shot play of this game, what are the pure strategy equilibria?
(b) Suppose the game is played twice with no discounting (δ = 1). What is
the highest welfare that can be obtained in a subgame perfect
equilibrium, and what is the equilibrum?
Solution
Problem 1

(b) Omitted.

Problem 2
(a) It describes a good that a consumer want some people to possess but not
many. For example, a party venue that gets better with more attendance, but gets
worse when it is too crowded. The value vi measures how much player i likes to
party. The value p is a cover charge for the club; if it is too high there is no
equilibrium with postivie attendance.
(d) Suppose consumers with values higher than 1 –x purchase
the good. Then the social welfare is

This is maximized at x = 1/4. Therefore, no equilibrium attains


the social optimum.

Problem 3
The circle graph with four players have the adjacency matrix of

Recall that agent i’s best response function is

First, consider an equilibrium where everyone is active. The condition is


which is impossible. Third, consider an equilibrium with two active agents.
By the same exercise, we know it is impossible to have agents 1 and 2 active.
For the case with agents 1 and 4 active, we have

This yields (x1, x2, x3, x4) = (1, 0, 0, 1) as long as δ ≥ 1/2. Note also that its
rotation (0, 1, 1, 0) is also an equilibrium. Finally, we can verify that there is no
equilibrium with one or zero active agent. Thus, there are two equilibria as
derived above.

Problem 4
Consider the equilibrium strategy in which player 1 plays C, D, C, D, . . . as long as
player 2 plays D, C, D, C, . . . , and vice versa. If the opponent deviates, then each
player commits to play D forever. In period 1, player 1’s anticipated payoff along the
given equilibrium path is
Thus, we need δ ≥ 1/6. It is easy to check that player 2 in period 1 (or player 1
in period 2) has no incentive to deviate if δ ≥ 1/6. Also, it is easy to see that if
either has deviated (so they are in an off-path state), then there is no incentive
for either to deviate from playing D forever. Hence, the given strategies
constitute an equilibrium if δ ≥ 1/6.

Problem 5
(a) The pure strategy equilibria are (B, B) and (C, C).

(b) In the second period, the highest payoff attainable is 1 at (B, B) since it is
the last period. In the first period, the highest possible payoff is 3 at (A,
A). I argue that payoff 3 in the first stage is attainable. Consider the
strategy in which a player takes A in the first period, and dependeing on
the opponent’s action in the first period, the player determines the
second-stage action; in particular, he takes B in the second period if the
opponent took A in the first period, and takes C otherwise. The pair of
this strategy earns a payoff of 4, while if one deviates, one can at most
obtain 4 - 1 = 3. Therefore, there is no incentive to deviate and hence it is
a subgame perfect equilibrium. Thus, the highest welfare attainable in a
SPE is 2(3 + 1) = 8.

You might also like