The Core
From Theory
Notes for CS 8803 - Game Theory and Computer Science. Spring 2008
Scribe: Liam Mac Dermed
Lecturer: Yair Tauman
The Core of V: (Shapley and Gillies 1953)
- A set of payoff vectors (x1,...,xn) s.t.
- and
- and
if
for some S then S can object since S can do better by itself.
Note: The core is convex, since it is the solution of a set of linear constraints. For the same reason, it can contain zero, exactly one, or infinite vectors.
Example 1
A simple majority game with N = {1,2,3}
I.E. V(S) =
- 1 if
- 0 otherwise
suppose that
then from our definition of the core:
thus
so
but we know that V(1,2,3) = 1 so
.
Example 2
N = {1,2,3} where players 1 and 2 have a left glove while player 2 has a right glove.
V(S) =
- 1 if S = {1,3} or {2,3} or {1,2,3}
- 0 otherwise
Suppose (x_1, x_2, x_3) \in C(V)
therefore x2 = 0.
The same logic can be applied to show x1 = 0 so x3 = 1.
Both left gloves (player 1 and 2) are in competition with each other and for any positive allocation to the other player can offer player 3 a bigger piece of the pie. Thus player 3 will get everything. C(V) = (0,0,1).
Example 3
N = {1,2,3,4,5} where now players 1 and 2 have left gloves and players 3,4 and 5 have right gloves.
L = {1,2}
R = {3,4,5}
like above: C(V) = (1,1,0,0,0).
if |L| = |R| then they will all get 1/2.
The production of this material was supported in part by NSF award SES-0734780.
