📈 economics

Null player neutrality in TU-games: Egalitarian and Shapley solutions

This paper introduces the axiom of null player neutrality to characterize a family of real linear combinations of the Shapley and equal division solutions in TU-games, extending the standard $\alpha$ -egalitarian Shapley values to arbitrary real coefficients while uniquely identifying the equal division solution when applied to nullifying players.

Original authors: J. C. Gonçalves-Dosantos, R. Martínez, J. Sánchez-Soriano

Published 2026-05-20

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: J. C. Gonçalves-Dosantos, R. Martínez, J. Sánchez-Soriano

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a group of friends deciding how to split a pot of money they earned together. Some friends did a lot of work, some did a little, and some didn't do anything at all. How do you decide who gets what?

This paper explores two main ways of thinking about this problem, using a branch of math called "cooperative game theory." The authors introduce a new rule to see how it changes the outcome.

The Two Extreme Philosophies

First, let's look at the two famous ways people usually solve this:

The "Hard Worker" Rule (The Shapley Value): This approach says, "You get paid exactly for what you contributed." If you did nothing, you get nothing. If you helped the group make \ $100, you get a share based on how much you added to that \$ 100. It's very strict and fair in a "meritocratic" sense.
The "Team Spirit" Rule (Equal Division): This approach says, "We are all in this together." It doesn't matter if you did nothing or everything; everyone gets an equal slice of the pie. It's purely about equality.

The New Rule: "Null Player Neutrality"

The authors noticed that the "Hard Worker" rule has a specific requirement: if you contribute nothing, you must get zero. They asked, "What if we relax that rule just a little bit?"

They created a new rule called Null Player Neutrality. Here is the analogy:

Imagine you are adding a "ghost" friend to your group. This ghost friend does absolutely nothing (they are a "null player").

The Old Strict Rule: If you add this ghost, they get $0, and it doesn't matter what other games or scenarios you imagine them playing.
The New Neutral Rule: The authors say, "Okay, let's say we add this ghost friend to a few different scenarios. As long as the total amount of money the whole group has at the end is the same in all those scenarios, the ghost friend should get the same amount of money in all of them."

The Catch: The rule doesn't say the ghost gets $0. It just says their payment shouldn't depend on which specific "do-nothing" scenario we are looking at, as long as the total pot size is fixed.

The Big Discovery

When the authors applied this new, slightly softer rule to the math, they found something surprising.

Before: If you demanded the "Hard Worker" rule (Zero for non-contributors), the only solution was the Shapley Value (strictly based on contribution).
After: By switching to "Null Player Neutrality," the solution isn't just one single point anymore. It opens up a whole spectrum of possibilities.

Think of it like a dimmer switch on a light.

Shapley Value is the light turned all the way up (100% contribution).
Equal Division is the light turned all the way down (0% contribution, 100% equality).
The New Family: The authors found that any setting on that dimmer switch is now mathematically valid. You can have a solution that is 50% contribution and 50% equality, or even 120% equality (giving the non-workers more than their share to compensate for something else).

They proved that if you keep the basic rules of fairness (everyone gets paid, symmetric players get the same, and the math works smoothly), the only solutions that fit are any mix of the "Hard Worker" rule and the "Team Spirit" rule.

The Twist: The "Ruiner" Player

The paper also looked at the opposite of a "do-nothing" player. They looked at a "Nullifying Player"—someone who, if they join a group, causes the whole group's value to drop to zero (like a toxic team member who ruins the project).

They tried to apply a similar "neutrality" rule to these "Ruiners."

The Result: The math broke differently. Instead of opening up a spectrum of solutions, this rule forced the answer to be only the "Team Spirit" rule (Equal Division).

Why the difference?

For the "Do-Nothing" player, the new rule was flexible enough to allow many different ways to split the money.
For the "Ruiner" player, the rule was so strict that the only way to satisfy it was to ignore individual contributions entirely and just split the pot equally.

Summary

The paper is essentially a mathematical experiment in fairness.

They took a strict rule about "people who do nothing get nothing."
They softened it to "people who do nothing get the same amount, regardless of the specific scenario, as long as the total pot is the same."
The Result: This small change allows for a massive variety of fair solutions, ranging from strict meritocracy to pure equality, and everything in between.
The Contrast: When they tried this with "people who ruin the game," the math forced them back to a single, strict solution: pure equality.

The authors conclude that the strict "do-nothing gets zero" rule is what forces us to be purely meritocratic. Once you soften that specific requirement, you unlock a whole world of mixed solutions that balance merit and equality in any proportion you choose.

Technical Summary: Null Player Neutrality in TU-Games

Problem Statement
In cooperative game theory with transferable utility (TU-games), the allocation of value among players is typically governed by structural axioms such as efficiency, linearity, and symmetry. Two polar solutions dominate the literature: the Shapley value, which embodies a marginalist perspective by rewarding players based on their expected marginal contributions, and the equal division solution, which allocates the grand coalition's worth uniformly regardless of individual productivity.

The Shapley value is classically characterized by the Null Player Property, which mandates that any player contributing nothing to every coalition receives a payoff of zero. While mathematically elegant, this property embeds normative assumptions that may conflict with egalitarian or needs-based principles of distributive justice (e.g., Rawlsian fairness), as it excludes non-productive agents from any share of collective wealth. Conversely, the equal division solution satisfies a different set of intuitions but fails the null player property.

The paper addresses the gap between these extremes by introducing a new axiom, Null Player Neutrality, to systematically analyze how weakening the strict requirements of the null player property affects the set of admissible solutions.

Methodology
The authors employ an axiomatic approach within the framework of TU-games. The methodology involves:

Defining New Axioms: The paper introduces Null Player Neutrality and its dual, Nullifying Player Neutrality.
- Null Player Neutrality: For any game $v$ and two augmenting games $w$ and $u$ where player $i$ is null in both, if the grand coalition values are equal ( $w(N) = u(N)$ ), then the payoff to player $i$ in the augmented games must be identical ( $\phi_i(v+w) = \phi_i(v+u)$ ). This weakens the classical Coalitional Strategic Equivalence, which requires the payoff to remain unchanged regardless of the augmenting game's total value.
- Nullifying Player Neutrality: A parallel axiom for "nullifying players" (those who reduce any coalition's value to zero).
Logical Analysis: The authors establish logical relationships between the new axioms and existing properties (e.g., the Null Player Property, Coalitional Strategic Equivalence, and the "Null Player in a Productive Environment" property).
Characterization Theorems: Using the unanimity game basis and linearity, the paper derives necessary and sufficient conditions for specific solution families.

Key Contributions and Results

Characterization of Linear Combinations (Theorem 2):
The primary result demonstrates that Efficiency, Linearity, Symmetry, and Null Player Neutrality uniquely characterize the family of solutions defined by:
$\phi = \alpha \phi^{ED} + (1-\alpha) \phi^{Sh}$
where $\alpha \in \mathbb{R}$ .
- This family strictly extends the known $\alpha$ -egalitarian Shapley values (where $\alpha \in [0, 1]$ ) to arbitrary real numbers.
- The result generalizes Shapley's classical characterization: replacing the Null Player Property with the weaker Null Player Neutrality expands the solution set from a single point ( $\alpha=0$ , the Shapley value) to a one-parameter family.
- The paper clarifies that the Null Player Property is the specific constraint that forces $\alpha = 0$ .
Comparison with Existing Literature:
- vs. Weak Monotonicity: The paper contrasts its result with van den Brink et al. (2013), who characterized the $\alpha \in [0, 1]$ subfamily using Weak Monotonicity. The authors show that Null Player Neutrality is logically independent of Weak Monotonicity and allows for $\alpha < 0$ and $\alpha > 1$ , thus strictly enlarging the admissible set.
- vs. Productive Environment Property: The paper compares with Casajus and Huettner (2013), who used the "Null Player in a Productive Environment" property to characterize the subfamily with $\alpha \geq 0$ . The authors prove that under linearity, the productive environment property implies Null Player Neutrality, but the converse does not hold, explaining the broader scope of Theorem 2.
Dual Analysis of Nullifying Players (Theorem 4):
The paper introduces a parallel analysis for nullifying players.
- Unlike the null player case, Nullifying Player Neutrality is shown to be equivalent to Coalitional Standard Equivalence (Proposition 2).
- Consequently, Efficiency, Symmetry, and Nullifying Player Neutrality uniquely characterize the Equal Division Solution ( $\phi \equiv \phi^{ED}$ ).
- Notably, this characterization does not require the Linearity axiom, unlike the null player case. This creates a sharp dichotomy: Null Player Neutrality permits a continuum of solutions, while Nullifying Player Neutrality collapses the set to a single solution.

Significance and Claims
The paper claims that the introduction of Null Player Neutrality offers a "natural middle ground" that relaxes the normative strictness of the Null Player Property without discarding the structural discipline of efficiency and symmetry.

Normative Implication: The relaxation allows for solutions that can compensate null players (when $\alpha > 1$ ) or penalize the equal division component relative to the Shapley value (when $\alpha < 0$ ), providing a more flexible axiomatic foundation for studying solidarity and redistribution.
Structural Insight: The work highlights a fundamental asymmetry in cooperative game theory: the treatment of null players (who add no value) allows for a spectrum of allocations, whereas the treatment of nullifying players (who destroy value) rigidly enforces equal division.
Theoretical Unification: By extending the $\alpha$ -egalitarian Shapley values to $\alpha \in \mathbb{R}$ , the paper provides a unified framework that encompasses both marginalist and egalitarian extremes and the linear combinations between and beyond them.

The authors maintain a modest stance, noting that their characterization relies on the specific set of axioms chosen and that the independence of these axioms (demonstrated in Remarks 3 and 6) confirms the necessity of each condition for the derived results.