Ezio Marchi
Director of Instituto de Matemática Aplicada San Luis, Universidad Nacional de San Luis, Ejército de los Andes 950, San Luis, Argentina
Correspondence to: Ezio Marchi, Director of Instituto de Matemática Aplicada San Luis, Universidad Nacional de San Luis, Ejército de los Andes 950, San Luis, Argentina.
Email: |  |
Copyright © 2024 The Author(s). Published by Scientific & Academic Publishing.
This work is licensed under the Creative Commons Attribution International License (CC BY).
http://creativecommons.org/licenses/by/4.0/

Abstract
In game theory there are many discussions about the concept of rationality. Theoretically these discussions are concerned with an appropriate mathematical concept which derives from the concept itself. In this paper we present a new way of looking rationality which is concerned with an external view of the rules of games. We prove the existence of these new friendly equilibrium and perturbed friendly equilibria.
Keywords:
Equilibria, Friendly
Cite this paper: Ezio Marchi, Friendly Equilibria, Journal of Game Theory, Vol. 13 No. 2, 2024, pp. 32-36. doi: 10.5923/j.jgt.20241302.02.
1. Results
The most important equilibrium of non-cooperative games is the Nash equilibrium point; the perfect equilibrium of Selten and proper equilibrium of Myerson are also considered important. Most authors of non-cooperative games study applications of these three equilibria. However, there was no plan to consider friends. In this work we emphasize this last concept that can be applied to problems of sociology, economics, mathematics, etc. In the future it can be thought that the existence of this friendly equilibria can be applied to the aforementioned balances. We prove the existence.Here we are going to introduce the general tools of the game. Consider an n-person game in normal form
where the strategy sets are non-empty and finite. The mixed extension is
Now for a player consider a finite sequence of players
which we call the “successor friends” of player
. Then from an intuitive point of view it is natural to assume that the player
is “rational”: whenever his payoff is the same he is going to act in such way that first he will maximize the payoff of his first successor friend
. In the case that the payoff of this successor friend
is equal again he will act to maximize the payoff of his second successor friend
and so on until the last
.Consider for
the sets
Formally we define a friendly equilibrium point if
.Theorem 1: Given any game
and
is upper semi-continuous for each
, there always exists a friendly equilibrium.Proof: given any point
consider the set of points
since the expectations are multi-linear then each
is non-empty convex and compact. Moreover, by the continuity of the expected functions and the hypothesis such,
are upper semi-continuous. Therefore using Kakutani’s fixed point applied to
there exists a fixed point
. Such point is a friendly equilibrium (q.e.d.).As a simple observation we would like to point out that in the case that
for each
our new notion coincides with the fundamental Nash equilibrium points.It is interesting to remark that the existence of a friendly equilibrium is equivalent to the existence of a solution
of the non-linear system
where
denotes the discrete support of
. The proof of the equivalence is rather easy and is left to the reader.
2. Perturbed Friendly Equilibria
Now we will extend the concept of friendly equilibria to the theory of perturbed equilibria. For this, we will follow the approaches of Selten (1975) and Myerson (1978).Consider a point
such that it satisfies
Theorem 2: A point is a friendly equilibrium point if and only if it fulfills B) with
for each
.Proof: We prove it by induction on
. For the case
for each
, this is very well known. Assume that it is valid for
, then we will show that it is also valid for
.For this, consider that
for each
. Suppose that for 
and
Let
be the point defined as
then
which is impossible since
and
is identified with
.Inversely, consider a point
such that
for each
, then
only if
and
therefore
for each
.Then following the ideas of Selten (1975), we define a perfect friendly equilibrium as follows: given
, a point
is called a
-perfect friendly equilibria if
A perfect friendly equilibrium is a limit point of a sequence
where
is an
-perfect friendly rational equilibria. Similarly following the ideas of Myerson (1978) we define
-proper friendly equilibria as
Clearly an
-proper friendly equilibrium is an
-perfect friendly equilibrium. A point
is called a proper friendly equilibria if
is a limit point of a sequence
where
is an
-proper friendly equilibrium.When
the
’s corresponding concepts become the perfect and proper ones respectively.Theorem 3: Every normal form game possesses at least one proper friendly equilibrium if
,
is upper semi-continuous.Proof: In order to prove this existence theorem let us observe that for given positive integers
it always holds true
This can be proven by induction on
, which we skip. Therefore, it is clear that
Now given
, define
where
is the cardinality of
, and let
For
define the correspondence
from
by:
We have that
is closed and convex for every
and the mapping
is upper semi-continuous. Now we will prove that
. Consider
for all
and
. Clearly
Let the point
Clearly
.Now if
then
and  | (1) |
 | (2) |
since
and the observation given above.Now if
then all
Then performing the same operation as before instead in, in, it is easy to see that
Let
be the n-tuple
. Then it satisfies the condition of the Kakutani Fixed Point Theorem (Kakutani, 1941) and the fixed point of
is an
-proper friendly equilibrium. Now for
each
we have an
-proper friendly equilibrium, thus making
, it is always possible to find a converging subsequence since the space is compact. The limit point of such subsequence is a friendly equilibrium point. The proof is complete, then the theorem follows.As a simple consequence of this result we haveCorollary: Every normal form game has at least one perfect friendly equilibrium.Finally we would like to say that if the
do not belong to
and are given externally, we have a external new notion of equilibrium. And we would like to emphasize that the previous material might be generalized accordingly for the continuous case. In this situation the upper semi-continuity condition appears more natural.
ACKNOWLEDGMENTS
I am great full for the collaboration of Miss Gimena Páez in editing and taping this paper.
Note
1. This paper was written in the year 1990 in a project supported by CONICET.
References
[1] | Kakutani, S. (1941). A generalization of Brouwer fixed point theorem. Duke Math. J. 451- 459. |
[2] | Marchi, E. (1991-1992). -subgame perfect equilibrium. |
[3] | Myerson, R. B. (1978). Refinements of the Nash equilibrium concept. Int. Game Theory 7, 73-80. |
[4] | Selten, R. (1975). Reexamination of the perfectness concept for equilibrium points in extensive games. Int. J. Game Theory 4, 25-55. |
[5] | Van Damme (1987). Stability and perfection of Nash Equilibria. Springer-Verlag, Berlin – New York. |