Imputation (Game Theory)¶

A payoff allocation in a transferable-utility cooperative game that distributes the grand coalition's entire worth while giving every player at least its stand-alone coalition value.

Version: v3 · 2026-09-06 · History
Domain-specific #: 2048
Origin domain: economics
Subdomain: cooperative game theory
Aliases: Game-theoretic imputation, Cooperative-game imputation, Imputation set

Core Idea¶

For a transferable-utility coalitional game ((N,v)), an imputation is a payoff vector \(x\in\mathbb R^N\) satisfying two baseline admissibility conditions:

\[ \sum_{i\in N}x_i=v(N) \quad\text{and}\quad x_i\ge v(\{i\})\ \text{for every }i\in N. \]

The first condition is efficiency: the grand coalition's available worth is fully distributed. The second is individual rationality: no player receives less than it can guarantee by acting alone.^[1] An imputation is therefore not yet a prediction or a uniquely fair answer. It is the feasible negotiating set remaining after waste and individually unacceptable allocations are removed.

That set supplies the domain on which stronger cooperative-game solution concepts operate. The core adds resistance to deviations by every coalition; the Shapley value applies marginal-contribution axioms;^[2] the nucleolus lexicographically reduces coalition excesses.^[3]

Structural Signature¶

A finite player set (N).
A transferable scalar payoff.
A characteristic function (v(S)) assigning worth to each coalition.
A grand-coalition worth (v(N)).
Singleton fallback values (v({i})).
A payoff vector with one component per player.
Efficiency: all grand-coalition worth is allocated.
Individual rationality for every player.
A possibly empty, singleton, or multidimensional imputation set.
A baseline domain for stronger solution concepts.
Separation between admissibility and selection.
Explicit sign conventions when the game allocates costs rather than benefits.

What It Is Not¶

It is not statistical imputation, which fills missing observations. It is not merely any feasible payoff vector: a pre-imputation may be efficient while failing individual rationality. It is not automatically in the core, because a subgroup may be able to block an allocation even when no individual can.^[4]

It is also not synonymous with the Shapley value or nucleolus. Those rules select a particular vector under additional principles, and their selected vector need not satisfy every other solution concept's requirements.

Scope of Application¶

Imputations are used in cooperative cost and surplus sharing, coalition formation, joint ventures, public projects, logistics, insurance, voting-power analysis, and dynamic cooperative games. The definition assumes transferable utility: one common payoff scale can be redistributed among players.

For non-transferable utility, externalities among coalitions, uncertain coalition values, or participation constraints richer than stand-alone worth, the baseline must be generalized. Cost games also require a consistent reversal of inequality and efficiency conventions.

Clarity¶

Publish (N), the characteristic function or at least (v(N)) and all singleton values, the payoff sign convention, and both tests. Calling a vector “an imputation” because it divides a total is incomplete. Likewise, individual rationality must be evaluated against each player's modeled outside option, not a vague fairness judgment.

Manages Complexity¶

The definition reduces an unconstrained continuum of payoff vectors to an affine hyperplane clipped by participation half-spaces. This geometrical object can be inspected for emptiness, dimension, extreme points, and intersections with the core. It cleanly separates the first question—what agreements are minimally acceptable?—from the harder question—which acceptable agreement should be chosen?

Abstract Reasoning¶

Specify the coalitional game and transferable payoff unit.
Compute the grand coalition's worth.
compute each singleton's stand-alone worth.
Form the efficiency hyperplane.
Intersect it with every individual-rationality constraint.
Test whether the imputation set is nonempty.
If selection is required, apply a declared stronger solution concept.
Check coalition stability separately rather than inferring it from individual rationality.
For a dynamic game, verify that continuation allocations remain acceptable over time.

Knowledge Transfer¶

The portable pattern is define the minimally admissible allocation set by combining full-budget use with a floor at each participant's outside option. It transfers to negotiated budgets, consortium revenue sharing, and resource pooling. The proposed immediate parent is Allocation.

Examples¶

Two producers. If (v({1})=5), (v({2})=5), and (v({1,2})=15), every vector ((x_1,x_2)) with (x_1+x_2=15) and both components at least five is an imputation.

Empty set. If the sum of singleton values exceeds (v(N)), no efficient allocation can meet every individual floor; the grand coalition is not viable under the model.

Core distinction. With three or more players, an imputation can satisfy each singleton while a pair can jointly obtain more than their combined assigned payoffs. The pair then blocks it, placing the vector outside the core.^[4]

Structural Tensions¶

Full distribution versus retaining slack or reserves.
Individual participation versus coalition stability.
Minimal admissibility versus normative fairness.
Transferable utility versus multidimensional preferences.
Static agreement versus time-consistent continuation.
Modelled outside options versus real bargaining power.

Structural–Framed Character¶

Budget balance, lower bounds, feasible-set intersection, and selection from an admissible region are structural. Players, coalitions, characteristic functions, transferable utility, and cooperative solution concepts are constitutive domain machinery. The identity is therefore domain-specific.

Structural Core vs. Domain Accent¶

The portable core is exhaust shared value + respect each participant's fallback floor. The domain accent is a transferable-utility characteristic-function game and the formal vocabulary of coalition worth.

Allocation is the proposed immediate parent. Cooperation, Pareto Efficiency, Fairness, Bargaining Power, and Constraint are related. Statistical Imputation is only a homonym and must remain separate.

The prospective queue contains one strict edge to prime:allocation. No live DAG mutation is authorized.

Relationships to Other Abstractions¶

Current abstraction Imputation (Game Theory) Domain-specific

Parents (1) — more general patterns this builds on

Imputation (Game Theory) is a kind of Allocation Prime

Allocation is the proposed immediate parent.

Hierarchy path (1) — routes to 1 parentless root

Imputation (Game Theory) → Allocation → Scarcity → Constraint

Neighborhood in Abstraction Space¶

Imputation (Game Theory) sits in a sparse region of the domain-specific corpus (91^st percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Fair Division & Cooperative Power (7 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Missing-data imputation.
Pre-imputation.
Core allocation.
Shapley value.
Nucleolus.
Nash bargaining solution.
Any equal split without reference to the characteristic function.

References¶

[1] John von Neumann and Oskar Morgenstern, Theory of Games and Economic Behavior, 2^nd ed. (Princeton University Press, 1947), foundational treatment of cooperative games, imputations, and solution sets. registry ↩

[2] Lloyd S. Shapley, “A Value for n-Person Games,” in H. W. Kuhn and A. W. Tucker, eds., Contributions to the Theory of Games II (Princeton University Press, 1953), pp. 307–317, doi:10.1515/9781400881970-018. registry ↩

[3] David Schmeidler, “The Nucleolus of a Characteristic Function Game,” SIAM Journal on Applied Mathematics 17, no. 6 (1969): 1163–1170, doi:10.1137/0117107. registry ↩

[4] Donald B. Gillies, “Some Theorems on n-Person Games,” PhD dissertation, Princeton University (1953), introducing the core as a coalitional-stability refinement. registry ↩a ↩b