Tải bản đầy đủ

Chapter 3

Chapter 3
Static Games of Incomplete Information


Normal-Form Representation of Static Bayesian
Games
Bayesian Games: Most popular model for describing situations with incomplete information,
developed by Harsanyi (1967).

Recall the Battle of Sexes.
Pat

Chris

Opera

Fight

Opera

2 ,1


0 ,0

Fight

0 ,0

1 ,2


Suppose that Chris and Pat are not sure of each other’s payoff.
In particular, suppose that the following situation;
Privately known by Chris
Opera

Chris

Opera

2+tc , 1
0 , 0

Fight

Pat

Fight

0 , 0
1 , 2+tp

The Battle of the Sexes with Incomplete Information

TC : the set of possible tc
Tp : the set of possible tp

Privately known by Pat
Type sets



Following Harsanyi (1967), we assume that the timing of the static game above is
as follows.

(1) Nature (God) draws a type vector t=(tc , tp) according to the prior
probability distribution p(t), which is assumed to be common
knowledge among them.
(2) Nature reveals tc to Chris but not to Pat, and reveals tp to Pat but
not to Chris.
Given her type tp, Pat computes the belief (probability)

pp(tc | tp) using Bayes’ rule: 𝑝𝑝 𝑡𝑐 |𝑡𝑝 =

𝑝(𝑡𝑐 ,𝑡𝑝 )

𝑝(𝑡𝑝 )



𝑝(𝑡𝑐 ,𝑡𝑝 )

𝑡𝑐 ∈𝑇𝑐

𝑝(𝑡𝑐 ,𝑡𝑝 )

.

Given his type tc, Chris computes the belief (probability)
pc(tp | tc) using Bayes’ rule: 𝑝𝑐 𝑡𝑝 |𝑡𝑐 =

𝑝(𝑡𝑝 ,𝑡𝑐 )
𝑝(𝑡𝑐 )



𝑝(𝑡𝑝 ,𝑡𝑐 )

𝑡𝑝 ∈𝑇𝑝

𝑝(𝑡𝑝 ,𝑡𝑐 )

.

(3) Chris and Pat simultaneously choose actions, Opera or Fight, and
then payoffs are received.


Definition The normal-form representation of an n-player
static Bayesian game specifies the players’ action spaces A1,…,An,
their type spaces T1,…,Tn,
their beliefs p1,…,pn, and their payoff functions u1,…,un.
Player i’s type, ti, is privately known by player i, and is a member
of the set of possible types Ti. Each type vector t=(t1,…,tn)
determines i’s payoff function ui(a1,…,an ; t) where (a1,…,an) is a
profile of all players’ actions.
Player i’s belief pi(t-i | ti) describes i’s uncertainty about n-1 other
players’ possible types, t-i, given i’s own type, ti.
We denote this game by G={A1,…,An ; T1,…,Tn ; p1,…,pn ; u1,…,un}.


Definition In the static Bayesian game G={A1,…,An ; T1,…,Tn ;
p1,…,pn ; u1,…,un}, a strategy for a player is a function si(ti), where for
each type ti in Ti, si(ti) specifies the action from the feasible set Ai
that type ti would choose if drawn by nature.
Definition In the static Bayesian game G={A1,…,An ; T1,…,Tn ; p1,…,pn ;
u1,…,un}, the strategies s*=(s*1,…,s*n) are (pure-strategy) Bayesian Nash
equilibrium if for each player i and for each of i’s type ti in Ti, s*i(ti)
solves


max σ𝑡−𝑖 ∈𝑇−𝑖 𝑢𝑖 𝑠1∗ 𝑡1 , … , 𝑠𝑖−1
𝑡𝑖−1 , 𝑎𝑖 , 𝑠𝑖+1
𝑡𝑖+1 , … , 𝑠𝑛∗ 𝑡𝑛 ; t 𝑝𝑖 (𝑡𝑖−1 |𝑡𝑖 )

𝑎𝑖 ∈𝐴𝑖

.


The Battle of the Sexes with Incomplete Information
Pat
Privately known by Chris
Opera

Opera

2+tc , 1

Fight

0 , 0

Chris

Fight

0 , 0
1 , 2+tp
Privately known by Pat

Suppose that tc and tp are independent
draws from a uniform distribution on [0,x].


In terms of the abstract static Bayesian game in normal form
G={Ac,Ap;Tc,Tp; pc,pp;uc,up};
・ the action spaces are Ac=Ap={Opera,Fight},
・ the type spaces are Tc=Tp=[0,x],
・ the beliefs are pc(tp)=pp(tc)=1/x for all tc and tp.
Task 1 Let’s construct a pure-strategy Bayesian Nash equilibrium in
which
・ Chris plays Opera if tc exceeds a critical value c, and plays Fight
otherwise;
・ Pat plays Fight if tp exceeds a critical value p, and plays Opera
otherwise.


Task 2 Find the probability that Chris plays Opera, and the probability
that Chris plays Fight in such a equilibrium.

Task 3 (an interpretation of mixed-strategy Nash equilibria)
Show that as x approaches zero, that is, as the incomplete information
disappears, those probability above (i.e., the players’ behavior in this
pure-strategy Bayesian Nash equilibrium of the incomplete information
game) approaches their behavior in the mixed-strategy Nash
equilibrium in the original game of complete information.



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Tải bản đầy đủ ngay

×