Abstract
This chapter studies the structure of games in partition function and according to an axiomatic point of view, we provide a global description of linear symmetric solutions by means of a decomposition of the set of such games (as well as of a decomposition of the space of payoff vectors). The exhibition of relevant subspaces in such decomposition and based on the idea that every permutation of the set of players may be thought of as a linear map, allow for a new look at linear symmetric solutions.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
The precise statement will be provided in Sect. 3.
- 2.
Notice that G can be identified with the set of real vectors indexed on the elements of EC, and so, \(\dim G = \left \vert EC\right \vert\).
- 3.
So, θ ⋅ w is the partition function game which would result from w if we relabeled the players by permutation θ.
- 4.
The formal statement will be found at the end of this section.
- 5.
Formally, if Y be a subspace of a vector space X, then Y is invariant (for the action of S n ) if for every y ∈ Y and every θ ∈ S n , we have that θ ⋅ y ∈ Y.
- 6.
That is, a subspace Y is irreducible if Y itself has no invariant subspaces other than {0} and Y itself.
- 7.
Here, \(\mathbf{0} = (0, 0,\ldots, 0) \in \mathbb{R}^{n}\).
- 8.
This seems like the natural inner product to consider, since intuitively G can be identified with \(\mathbb{R}^{EC}\).
- 9.
This type of games may be thought of as its counterpart for symmetric games in TU games.
- 10.
That is, r(S, Q) = 0 for every (S, Q) ∈ EC.
References
Albizuri, M.J., Arin, J., Rubio, J.: An axiom system for a value for games in partition function form. Int. Game Theory Rev. 7 (1), 63–72 (2005)
Bolger, E.M.: A class of efficient values for games in partition function form. J. Algebraic Discrete Methods 8 (3), 460–466 (1987)
Bolger E.M.: A set of axioms for a value for partition function games. Int. J. Game Theory 18 (1), 37–44 (1989)
Fulton, W., Harris, J.: Representation Theory: A First Course. Graduate Texts in Mathematics, vol. 129. Springer, New York (1991)
Hernández-Lamoneda, L., Juárez, R., Sánchez-Sánchez, F.: Dissection of solutions in cooperative game theory using representation techniques. Int. J. Game Theory 35 (3), 395–426 (2007)
Hernández-Lamoneda, L., Sánchez-Pérez, J. Sánchez-Sánchez, F.: The class of efficient linear symmetric values for games in partition function form. Int. Game Theory Rev. 11 (3), 369–382 (2009)
Hu, C.C., Yang, Y.Y.: An axiomatic characterization of a value for games in partition function form. SERIEs: J. Span. Econ. Assoc. 1 (4), 475–487 (2010)
Ju, Y.: The consensus value for games in partition function form. Int. Game Theory Rev. 9 (3), 437–452 (2007)
Kleinberg, N.L., Weiss, J.H.: Equivalent n-person games and the null space of the Shapley value. Math. Oper. Res. 10 (2), 233–243 (1985)
Kleinberg, N.L., Weiss, J.H.: Weak values, the core and new axioms for the Shapley value. Math. Soc. Sci. 12, 21–30 (1986)
Lucas, W.F., Thrall, R.M.: n-person games in partition function form. Nav. Res. Logist. Q. 10, 281–298 (1963)
Macho-Stadler, I., Pérez-Castrillo, D., Wettstein, D.: Sharing the surplus: an extension of the Shapley value for environments with externalities. J. Econ. Theory 135, 339–356 (2007)
Myerson R.B.: Values of games in partition function form. Int. J. Game Theory 6 (1), 23–31 (1977)
Pham, D.K., Norde, H.: The Shapley value for partition function games. Int. Game Theory Rev. 9 (2), 353–360 (2007)
Sánchez-Pérez, J.: Application of the representations of symmetric groups to characterize solutions of games in partition function form. Oper. Res. Decis. 24 (2), 97–122 (2014)
Acknowledgements
I thank the participants of the European Meeting on Game Theory (St. Petersburg, Russia, 2015) for comments, interesting discussions, and encouragement. I am also grateful to the Editors and two anonymous referees for their useful comments and suggestions. J. Sánchez-Pérez acknowledges support from CONACYT research grant 130515.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix
Appendix
The point of view we take in this work depends heavily on a certain decomposition of G (and \(\mathbb{R}^{n}\)) as a direct sum of subspaces, as well as on the Schur’s Lemma. A reference for this subject is [4]; nevertheless, we recall all the basic facts that we need.
The symmetric group S n acts on G via linear transformations (i.e., G is a representation of S n ). That is, there is a group homomorphism ρ: S n → GL(G), where GL(G) is the group of invertible linear maps in G. This action is given by:
for every θ ∈ S n , w ∈ G and (S, Q) ∈ EC.
The space of payoff vectors, \(\mathbb{R}^{n}\), is also an S n -representation:
In general, for any representation X of a finite group H, there is a decomposition
where the X i are distinct irreducible representations. The decomposition is unique, as are the X i that occur and their multiplicities a i .
This property is called “complete reducibility” and the extent to which the decomposition of an arbitrary representation into a direct sum of irreducible ones is unique is one of the consequences of the following:
Theorem 4 (Schur’s Lemma).
Let X 1 ,X 2 be irreducible representations of a group H. If T: X 1 → X 2 is H-equivariant, then T = 0 or T is an isomorphism.
Moreover, if X 1 and X 2 are complex vector spaces, then T is unique up to multiplication by a scalar \(\lambda \in \mathbb{C}\) .
The previous theorem is one of the reasons why it is worth carrying around the group action when there is one. Its simplicity hides the fact that it is a very powerful tool.
There is a remarkably effective technique for decomposing any given finite dimensional representation into its irreducible components. The secret is character theory.
Definition 5.
Let ρ: H → GL(X) be a representation. The character of X is the complex-valued function \(\chi _{X}: H \rightarrow \mathbb{C}\), defined as:
The character of a representation is easy to compute. If H acts on an n-dimensional space X, we write each element h as an n × n matrix according to its action expressed in some convenient basis, then sum up the diagonal elements of the matrix for h to get χ X (h). For example, the trace of the identity map of an n-dimensional vector space is the trace of the n × n identity matrix, or n. In fact, χ X (e) = dimX for any finite dimensional representation X of any group.
Notice that, in particular, we have χ X (h) = χ X (ghg −1) for g, h ∈ H. So that χ X is constant on the conjugacy classes of H; such a function is called a class function.
Definition 6.
Let \(\mathbb{C}_{class}(H) =\{ f: H \rightarrow \mathbb{C}\mid f\) is a class function on H}. If \(\chi _{1},\chi _{2} \in \mathbb{C}_{class}(H)\), we define an Hermitian inner product on \(\mathbb{C}_{class}(H)\) by
The multiplicities of irreducible subspaces in a representation can be calculated via the previous inner product. Formally, if \(Z = Z_{1}^{\oplus a_{1}} \oplus Z_{2}^{\oplus a_{2}} \oplus \cdot \cdot \cdot \oplus Z_{j}^{\oplus a_{j}}\), then the multiplicity Z i (irreducible representation) in Z is \(a_{i} = \left \langle \chi _{Z},\chi _{Z_{i}}\right \rangle\).
Proof (of Theorem 1).
First, \(\left \langle \chi _{G_{\lambda }^{k}},\chi _{\varDelta _{n}}\right \rangle\) and \(\left \langle \chi _{G_{\lambda }^{k}},\chi _{\varDelta _{n}^{\perp }}\right \rangle\) are the number of subspaces isomorphic to the trivial (Δ n ) and standard representation (Δ n ⊥ ) within G λ k, respectively.
We start by computing the number of copies of Δ n in G λ k:
Notice that \(\chi _{G_{\lambda }^{k}}(\theta )\) is just the number of pairs (S, Q) ∈ EC with \(\left \vert S\right \vert = k\) and λ Q = λ, that are fixed under θ ∈ S n .
Define
Then,
and so,
where,
Now, S n acts on the set Q λ and take Q ∈ Q λ . The orbit of Q under S n is
and the isotropy subgroup of Q is
By Lagrange theorem, we get
Notice that H = (S n ) Q acts on Q and takes S ∈ Q such that \(\left \vert S\right \vert = k\). The orbit of S under H is
Observe that \(\left \vert HS\right \vert = m_{k}^{\lambda }\) and the isotropy subgroup of S is
Again, by Lagrange theorem, we get
And therefore,
Now, we compute the multiplicity of Δ n ⊥ in G λ k. Since \(\mathbb{R}^{n} =\varDelta _{n} \oplus \varDelta _{n}^{\perp }\), then \(\chi _{\mathbb{R}^{n}} =\chi _{\varDelta _{n}} +\chi _{\varDelta _{n}^{\perp }}\ \Rightarrow \ \left \langle \chi _{G_{\lambda }^{k}},\chi _{\mathbb{R}^{n}}\right \rangle = \left \langle \chi _{G_{\lambda }^{k}},\chi _{\varDelta _{n}}\right \rangle + \left \langle \chi _{G_{\lambda }^{k}},\chi _{\varDelta _{n}^{\perp }}\right \rangle \ \Rightarrow \left \langle \chi _{G_{\lambda }^{k}},\chi _{\varDelta _{n}^{\perp }}\right \rangle = \left \langle \chi _{G_{\lambda }^{k}},\chi _{\mathbb{R}^{n}}\right \rangle - 1\).
Notice that \(G_{[1,1,\ldots,1]}^{1} \simeq \mathbb{R}^{n}\) (as a representation for S n ). Let us compute
For x ∈ S:
Without loss of generality, suppose \(\left \vert S\right \vert = k =\lambda _{1}\) and take the case \(m_{k}^{\lambda } = m_{\lambda _{1}}^{\lambda } = 1\). Here, \(M = H_{S} = \left \{\theta \in S_{n}\mid \theta (Q) = Q,\text{ }\theta (S) = S\right \}\) acts on S ∈ Q ∈ Q λ and take x ∈ S. The orbit of x under M is
and the isotropy subgroup of x is
By Lagrange theorem,
Thus,
Finally, without loss of generality, suppose \(\left \vert S\right \vert = k =\lambda _{1}\) and take the case m k λ > 1. Following the same line of reasoning as above, we obtain
In summary,
The next task is to identify such copies of Δ n and Δ n ⊥ inside G λ k. For that end, define the functions \(L_{\lambda,k}: \mathbb{R}^{n} \rightarrow G_{\lambda }^{k}\) by L λ, k (x) = x (λ, k). These maps are isomorphisms between U λ k and Δ n (similarly between \(\left \{x_{j}^{\left (\lambda,k\right )}\mid x_{j} \in \varDelta _{n}^{\perp }\right \}\) and Δ n ⊥ , for each j ∈ I λ, k ) since they are linear, S n -equivariant, and 1 − 1. Thus, inside of G λ k, we have the images of these two subspaces: U λ k = L λ, k (Δ n ) and V λ k = L λ, k (Δ n ⊥ ).
Finally, the invariant inner product \(\left \langle,\right \rangle\) gives an equivariant isomorphism, in particular must preserve the decomposition. This implies orthogonality of the decomposition.
Proof (of Proposition 2).
We start by computing the orthogonal projection of w onto U. Notice that {u λ k} is an orthogonal basis for U, and that \(\left \Vert u_{\lambda }^{k}\right \Vert ^{2} = m_{k}^{\lambda }\left \vert Q_{\lambda }\right \vert\).
Thus, the projection of w onto U is
and so,
Now, for each (λ, k) ∈ A n , we define \(f^{\lambda,k}: G \rightarrow \mathbb{R}^{n}\) as
where each f λ, k is S n -equivariant and observe that f [n], n(w) = w(N, {N})(1, …, 1). Let z ∈ Δ n ⊥ , then f λ, k(z γ, i, j (γ, i)) = 0 if λ ≠ γ or i ≠ k, whereas (by Schur’s Lemma) f λ, k(z λ, k, j (λ, k)) = z λ, k, j ∈ Δ n ⊥ .
Let \(p: \mathbb{R}^{n} \rightarrow \varDelta _{n}^{\perp }\) be the projection of \(\mathbb{R}^{n}\) onto Δ n ⊥ given by
This projection is equivariant, sends Δ n to zero and it is the identity on Δ n ⊥ .
Next, define L λ, k: G → Δ n ⊥ as L λ, k = p ∘ f λ, k. Observe that
since by equivariance, \(f^{\lambda,k}\left (U\right ) \subset \varDelta _{n}\) and \(f^{\lambda,k}\left (W\right ) = 0\). Moreover, f λ, k(z γ, i, j (γ, i)) = 0 if λ ≠ γ or i ≠ k. Then,
The value of the component \(\left (z_{\lambda,k,j}\right )_{i}\) follows from the fact that the last expression can be written as
Proof (of Proposition 3).
Let \(\varphi: G \rightarrow \mathbb{R}^{n}\) be a linear symmetric solution. From Proposition 2, w ∈ G decomposes as
where by linearity,
Now, \(\varphi _{i}\left (r\right ) = 0\) by Corollary 1 and Schur’s Lemma implies
for some constants {α(λ, k)∣(λ, k) ∈ A n } ∪{β(λ, k, j)∣(λ, k, j) ∈ B n }.
Set \(\left \vert (S,Q)_{\lambda,k}\right \vert = \left \vert (S,Q) \in EC: \left \vert S\right \vert = k,\lambda _{Q} =\lambda \right \vert\), then
Observe that
and
Thus
The result follows by setting \(\gamma \left (\lambda,k\right ) = \frac{\alpha \left (\lambda,k\right )} {\left \vert (S,Q)_{\lambda,k}\right \vert } +\mathop{\sum }\limits_{ j \in I_{\lambda,k}} \frac{j} {n}m_{j}^{\lambda -\left [\left \vert S\right \vert \right ]}\beta \left (\lambda,k,j\right )\) and \(\delta \left (\lambda,k,j\right ) = \frac{\alpha \left (\lambda,k\right )} {\left \vert (S,Q)_{\lambda,k}\right \vert } -\frac{k} {n}\beta \left (\lambda,k,j\right )\).
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Sánchez-Pérez, J. (2016). A New Look at the Study of Solutions for Games in Partition Function Form. In: Petrosyan, L., Mazalov, V. (eds) Recent Advances in Game Theory and Applications. Static & Dynamic Game Theory: Foundations & Applications. Birkhäuser, Cham. https://doi.org/10.1007/978-3-319-43838-2_12
Download citation
DOI: https://doi.org/10.1007/978-3-319-43838-2_12
Published:
Publisher Name: Birkhäuser, Cham
Print ISBN: 978-3-319-43837-5
Online ISBN: 978-3-319-43838-2
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)