Fair Game: Formal and Applied Perspectives

• Fair Game is a collection of domain-specific constructions that formalize fairness via symmetry, proportional allocation, or recurring constraints across diverse fields.
• It encompasses models in probability (coin-toss games), bargaining (Ultimatum Game dynamics), and machine learning (cooperative allocation and dynamic debiasing) to ensure equitable outcomes.
• It also informs methodologies in formal verification, autonomous systems, and redistricting, illustrating how fairness can be operationalized through precise mathematical frameworks.

Searching arXiv for the specified "^^^^^^^^6Fair Game6^^^^^^^^" papers to ground the article in current records. arxiv_search(^^^^^^^^6query6 Game6^^^^^^^^", max_results=^^^^6Fair Game6Fair Game6^^^^) Searching arXiv for exact records relevant to "^^^^^^^^6Fair Game6^^^^^^^^". {"^^^^^^^^6query6 Game6^^^^^^^^\" OR ti:\6^^^^^^^^" Game6^^^^^^^^\"", "max_results": ^^^^6Fair Game6Fair Game6^^^^} “Fair game” is not a single formal notion in contemporary research; it is a family of domain-specific constructions in which fairness is encoded as symmetry, admissibility, proportionality, or corrective control. In combinatorial probability, fairness can mean exact equality of winning probabilities produced by a score-exchanging involution (&&&^^^^6Fair Game6^^^^&&&). In bargaining and evolutionary models, it can refer to the emergence or promotion of PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^ behavior in the Ultimatum Game under occasional fair actions or targeted intervention (&&&^^^^6Fair Game6^^^^&&&, &&&^^^^6Fair Game6^^^^&&&). In machine learning, it can denote either a modified cooperative game with fair-and-stable payoff allocation among data providers (&&&^^^^6query6^^^^&&&) or a closed-loop auditor–debiaser architecture that adapts fairness objectives over time (&&&^^^^6all:\6^^^^&&&). In formal methods, it denotes graph games in which one or both players are constrained by strong transition-fairness conditions (&&&^^^^6 OR ti:\6^^^^&&&, Anand et al., 28 Jan 2025).

Domain	Formal object	Operational meaning of fairness
Probability and combinatorics	Coin-toss word games; dice-sum distributions	Symmetry of win laws or indistinguishability from a fair distribution
Evolutionary and bargaining theory	Ultimatum-game populations	Fair offers, fair responses, or low-cost intervention toward equitable outcomes
Multi-party ML and Fair ML	Cooperative allocation; RL auditor–debiaser loop	Proportional reward allocation or dynamic bias control
Formal verification and control	PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^-regular, mean-payoff, energy games	Obligatory recurrence of designated transitions
Strategic and quantum game theory	Bayesian games; stochastic Prisoner’s Dilemma	Equalized payoffs, fair equilibria, or fairness-enforcing strategies
Applied systems	Autonomous racing; redistricting	Sportsmanship constraints or near-proportional procedural guarantees

^^^^6Fair Game6^^^^. Probabilistic symmetry and exact fairness

A mathematically sharp use of “fair game” appears in Litt’s coin-tossing game as analyzed by Basdevant, Hénard, Maurel-Ségala, and Singh. A fair coin is flipped PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^ times, two words PRESERVED_PLACEHOLDER_^^^^6query6^^^^ of equal length are fixed, and overlapping occurrences are counted. The paper defines the overlap set PRESERVED_PLACEHOLDER_^^^^6all:\6^^^^, the correlation number PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6^^^^, and the generating polynomial $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$. Its main theorem states that if $C(A,A)=C(B,B)$, then for every $n\ge 1$ the random vectors $(N_A(X_n),N_B(X_n))$ and PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^ have the same law, hence PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^ (&&&^^^^6Fair Game6^^^^&&&).

The result is notable because the relevant invariant is the auto-correlation structure alone. The paper gives PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^ and PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6^^^^ as the smallest nontrivial example: both have PRESERVED_PLACEHOLDER_^^^^6Fair Game6all:\6^^^^, even though their inter-correlations differ, and the game is exactly fair for every PRESERVED_PLACEHOLDER_^^^^6Fair Game6 OR ti:\6^^^^ (&&&^^^^6Fair Game6^^^^&&&). The proof is bijective rather than asymptotic: an explicit involution PRESERVED_PLACEHOLDER_^^^^6Fair Game66^^^^ on length-PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^7 bit-strings exchanges the Alice and Bob scores.

A related but distinct notion of fairness appears in the loaded-dice analysis of craps. There the question is whether two non-uniform six-sided dice can reproduce exactly the sum distribution of two fair dice. Writing PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^8, PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^9, and PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^, the condition becomes

PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^

where PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^. The analysis finds PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6^^^^ complex solutions counted with multiplicity, PRESERVED_PLACEHOLDER_^^^^6Fair Game6all:\6^^^^ distinct unordered pairs, and exactly one solution with all PRESERVED_PLACEHOLDER_^^^^6Fair Game6 OR ti:\6^^^^: the trivial fair pair PRESERVED_PLACEHOLDER_^^^^6Fair Game66^^^^ (Morrison et al., 2013). In this setting, a “fair game” of craps is not generated by hidden loading; fairness is identifiable from the full feasible nonnegative factorization structure.

^^^^6Fair Game6^^^^. Fairness in bargaining populations and intervention design

In evolutionary Ultimatum-Game models, fairness is often operationalized by offer and acceptance levels rather than by symmetry of win probabilities. In the “good Samaritan” model, each player carries a strategy PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^7, with PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^8 the fraction offered as proposer and PRESERVED_PLACEHOLDER_^^^^6Fair Game6^^^^9 the minimal accepted offer as responder. In each round, with probability PRESERVED_PLACEHOLDER_^^^^6query6Fair Game6^^^^, a player temporarily abandons PRESERVED_PLACEHOLDER_^^^^6query6Fair Game6^^^^ and uses the fixed fair strategy PRESERVED_PLACEHOLDER_^^^^6query6Fair Game6^^^^ with PRESERVED_PLACEHOLDER_^^^^6query6query6^^^^. A mean-field replicator-style ODE is derived: PRESERVED_PLACEHOLDER_^^^^6query6all:\6^^^^ with a saddle-node bifurcation at

PRESERVED_PLACEHOLDER_^^^^6query6 OR ti:\6^^^^

For PRESERVED_PLACEHOLDER_^^^^6query66^^^^ there are two attractors, while for PRESERVED_PLACEHOLDER_^^^^6query6^^^^7 only full fairness remains (&&&^^^^6Fair Game6^^^^&&&).

The numerical phenomenology depends strongly on network structure. In well-mixed populations, fairness exhibits a first-order transition and hysteresis: for PRESERVED_PLACEHOLDER_^^^^6query6^^^^8, the population remains in a low-fairness state with PRESERVED_PLACEHOLDER_^^^^6query6^^^^9; between roughly PRESERVED_PLACEHOLDER_^^^^6all:\6Fair Game6^^^^ and PRESERVED_PLACEHOLDER_^^^^6all:\6Fair Game6^^^^, runs are bistable; and small fair frequencies around PRESERVED_PLACEHOLDER_^^^^6all:\6Fair Game6^^^^ are sufficient to tip a homogeneous society into universal PRESERVED_PLACEHOLDER_^^^^6all:\6query6^^^^ sharing (&&&^^^^6Fair Game6^^^^&&&). On Barabási–Albert graphs the transition becomes continuous, and even a single highest-degree node acting as a good Samaritan can drive the population to full fairness once PRESERVED_PLACEHOLDER_^^^^6all:\6all:\6^^^^ (&&&^^^^6Fair Game6^^^^&&&).

A more interventionist formulation appears in the spatial Ultimatum Game with an external investor. Players occupy a PRESERVED_PLACEHOLDER_^^^^6all:\6 OR ti:\6^^^^-D lattice, use one of four discrete strategies PRESERVED_PLACEHOLDER_^^^^6all:\66^^^^, update by the Fermi rule with noise PRESERVED_PLACEHOLDER_^^^^6all:\6^^^^7, and may mutate with probability PRESERVED_PLACEHOLDER_^^^^6all:\6^^^^8. The investor observes either global frequencies PRESERVED_PLACEHOLDER_^^^^6all:\6^^^^9 or local neighborhood frequencies PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6Fair Game6^^^^, gives supplementary endowments PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6Fair Game6^^^^ or PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6Fair Game6^^^^, and incurs long-run cost

PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6query6^^^^

The main design finding is asymmetrical: globally, the cheapest scheme targets only PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6all:\6^^^^-players with PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6 OR ti:\6^^^^; locally, cost efficiency improves when support is restricted to neighborhoods with PRESERVED_PLACEHOLDER_^^^^6 OR ti:\66^^^^. For PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6^^^^7, achieving PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6^^^^8 is summarized as PRESERVED_PLACEHOLDER_^^^^6 OR ti:\6^^^^9 for the global $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$^^^^6Fair Game6^^^^-only rule and $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$^^^^6Fair Game6^^^^ for the local rare-proposer rule (&&&^^^^6Fair Game6^^^^&&&).

^^^^6query6^^^^. Fairness as stability, proportionality, and strategic advantage

In multi-party machine learning, “fair game” is defined through a modified cooperative-game model for non-rivalrous goods. A coalition-value map $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$^^^^6Fair Game6^^^^ is assumed monotonic, and an outcome is a payoff vector $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$^^^^6query6^^^^. Stability is modified from the classical core: $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$^^^^6all:\6^^^^ is stable iff for every coalition $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$^^^^6 OR ti:\6^^^^, there exists $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$6 with $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$7. Fairness is proportionality to a baseline contribution vector $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$8, taken to be the Shapley value, so that $P_{A,B}(z)=\sum_{k\in \mathrm{Corr}_k(A,B)} z^k$9 for some $C(A,A)=C(B,B)$^^^^6Fair Game6^^^^. Proposition ^^^^6 OR ti:\6^^^^ reduces stability checking to the $C(A,A)=C(B,B)$^^^^6Fair Game6^^^^ prefix inequalities $C(A,A)=C(B,B)$^^^^6Fair Game6^^^^, and Theorem 6 gives the unique optimal fair-and-stable allocation when it exists: $C(A,A)=C(B,B)$^^^^6query6^^^^ provided the corresponding prefix constraints hold (&&&^^^^6query6^^^^&&&). In experiments, Fisher-information valuation satisfies these constraints, whereas mutual-information valuation does not, forcing a stable but non-proportional compromise (&&&^^^^6query6^^^^&&&).

A different sense of fairness appears in Bayesian games with classical, quantum, and no-signaling advice. In the two-player family $C(A,A)=C(B,B)$^^^^6all:\6^^^^, a fair equilibrium has $C(A,A)=C(B,B)$^^^^6 OR ti:\6^^^^, while an unfair equilibrium has $C(A,A)=C(B,B)$6. The analysis shows that nonlocal correlations can outperform not only fair classical equilibria but also unfair ones. When the fair-equilibrium benchmark is $C(A,A)=C(B,B)$7, the outperforming condition reduces to $C(A,A)=C(B,B)$8, where $C(A,A)=C(B,B)$9 is the CHSH expression. For unfair equilibria, the advantageous region is characterized by explicit linear inequalities in $n\ge 1$^^^^6Fair Game6^^^^, $n\ge 1$^^^^6Fair Game6^^^^, and $n\ge 1$^^^^6Fair Game6^^^^ (&&&^^^^6Fair Game66^^^^&&&). This directly refutes the misconception that nonlocal advantage is relevant only against fair classical baselines.

Fairness in payoff control becomes more restrictive in stochastic repeated games. In the periodic Prisoner’s Dilemma, a fair zero-determinant strategy for player $n\ge 1$^^^^6query6^^^^ enforces $n\ge 1$^^^^6all:\6^^^^. Nakamura and Ueda derive necessary and sufficient existence conditions in terms of $n\ge 1$^^^^6 OR ti:\6^^^^, and show that fair ZD strategies do not necessarily exist in the periodic game, unlike in the standard repeated Prisoner’s Dilemma (&&&^^^^6Fair Game6^^^^7&&&). Tit-for-Tat is fair only in the special case

$n\ge 1$6

so equal-payoff enforcement is no longer generic once state dynamics are introduced (&&&^^^^6Fair Game6^^^^7&&&).

^^^^6all:\6^^^^. Dynamic fairness in machine learning and sequential decision-making

In Fair ML, “^^^^^^^^6Fair Game6^^^^^^^^” names a closed-loop framework in which an auditor and a debiaser surround an ML model and adapt over time. The joint auditor–debiaser is modeled as an RL agent interacting with an environment $n\ge 1$7. The state is $n\ge 1$8, where $n\ge 1$9 are current model parameters and $(N_A(X_n),N_B(X_n))$^^^^6Fair Game6^^^^ is the vector of last-observed fairness metrics. Actions include threshold choices, reweighting factors, or Lagrange multipliers, while the reward penalizes both predictive loss and fairness violation: $(N_A(X_n),N_B(X_n))$^^^^6Fair Game6^^^^ The auditor estimates metrics such as Statistical Parity, Demographic Parity ratio, Equalized Odds gap, and Predictive Value Parity gap, and is assumed “anytime-accurate PAC” (&&&^^^^6all:\6^^^^&&&).

The theoretical guarantee combines auditing accuracy and RL regret. Under an anytime-PAC auditor and an $(N_A(X_n),N_B(X_n))$^^^^6Fair Game6^^^^-regret debiaser, the time-averaged bias satisfies

$(N_A(X_n),N_B(X_n))$^^^^6query6^^^^

with probability at least $(N_A(X_n),N_B(X_n))$^^^^6all:\6^^^^ (&&&^^^^6all:\6^^^^&&&). On UCI-Adult, COMPAS, and German Credit, the reported excerpt gives ^^^^^^^^6Fair Game6^^^^^^^^ an accuracy of $(N_A(X_n),N_B(X_n))$^^^^6 OR ti:\6^^^^, with SP gap $(N_A(X_n),N_B(X_n))$6, EO gap $(N_A(X_n),N_B(X_n))$7, and PVP gap $(N_A(X_n),N_B(X_n))$8; over a $(N_A(X_n),N_B(X_n))$9-step horizon it reduces SP and EO gaps by about PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^ beyond static-RL at an accuracy cost of about PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^ (&&&^^^^6all:\6^^^^&&&).

A related but distinct sequential formalism replaces utilitarian welfare by Proportional Fairness in mixed-motive Markov games. The fair altruistic utility is

PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^

and in two-player social dilemmas cooperation at PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6query6^^^^ is stabilized when

PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6all:\6^^^^

for Prisoner’s Dilemma and Chicken (&&&^^^^6Fair Game6Fair Game6^^^^&&&). The sequential extension defines a Fair Markov Game and derives Fair MAA^^^^6Fair Game6^^^^C and Fair MAPPO. In CleanUp, under PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6 OR ti:\6^^^^, PF-MAPPO achieves approximately PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game66^^^^ apples with Gini PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^7, whereas utilitarian MAPPO achieves approximately PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^8 apples with Gini PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^9 (&&&^^^^6Fair Game6Fair Game6^^^^&&&). Here fairness is not merely a constraint; it is embedded directly into the objective.

^^^^6 OR ti:\6^^^^. Fairness constraints in graph games and quantitative verification

In formal verification, fair games are games on finite graphs whose plays must satisfy fairness obligations on designated moves. In fair PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^-regular games, a game graph PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^ is augmented by a set of fair edges PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^. A play is PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6query6^^^^-fair if, whenever a source vertex of a fair edge owned by player PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6all:\6^^^^ is visited infinitely often, that fair edge is also taken infinitely often. Two winning conditions PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6 OR ti:\6^^^^ and PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game66^^^^ determine the winner of mutually fair and mutually unfair plays; if exactly one player is fair, the fairly playing player wins (&&&^^^^6 OR ti:\6^^^^&&&). These fair PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^7 games are determined, and the parity/parity case admits both a polynomial reduction to ordinary parity games and a direct symbolic PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^8-calculus fixpoint algorithm (&&&^^^^6 OR ti:\6^^^^&&&).

The quantitative extension replaces parity acceptance with mean-payoff or energy objectives while keeping strong transition fairness. For a set PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^9 of fair edges,

PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^

When fairness is imposed on player PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^, the winning condition becomes PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6Fair Game6^^^^ or PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6query6^^^^; when it is imposed on player PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6all:\6^^^^, player PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6 OR ti:\6^^^^ wins any unfair play by player PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game66^^^^ (Anand et al., 28 Jan 2025). The main algorithmic result is gadget-based reduction to ordinary mean-payoff or energy games. Both PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^7-fair and PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^8-fair mean-payoff games are determined, with pseudo-polynomial complexity PRESERVED_PLACEHOLDER_^^^^6Fair Game6Fair Game6^^^^9 for optimal-value computation or PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6Fair Game6^^^^ for threshold solving (Anand et al., 28 Jan 2025). By contrast, PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6Fair Game6^^^^-fair energy games are not determined, although both winning regions remain computable (Anand et al., 28 Jan 2025).

These results make clear that, in verification, fairness is not primarily distributive. It is a liveness-style recurrence obligation on transitions, and its main technical role is to constrain admissible infinite behavior.

6. Procedural and applied notions of fair play

In autonomous racing, fair play is formalized as sportsmanship constraints. Huang et al. define a blocking predicate PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6Fair Game6^^^^, then impose a One-Motion Rule forbidding double-blocking over a four-step window and an Enough-Space Rule forbidding a later block when a faster attacker is already near the track boundary. High-level intentions are selected by a Stackelberg game solved by MCTS, while low-level trajectories are computed as a Generalized Nash Equilibrium Problem with shared dynamics, collision-avoidance, and velocity constraints (&&&^^^^6Fair Game6^^^^9&&&). In simulations with planning horizon PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6query6^^^^ and PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6all:\6^^^^, the “both aware” setting yields PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6 OR ti:\6^^^^ and zero violation, whereas an unaware defender blocks aggressively, produces high violation rates, and prevents successful overtaking (&&&^^^^6Fair Game6^^^^9&&&). Fairness here is a codified behavioral rule set, not an ex post outcome metric.

Procedural fairness also appears in redistricting. In Redistricting Ghost, two parties alternately place voters into districts of size PRESERVED_PLACEHOLDER_^^^^6Fair Game6query66^^^^, with the minority moving first. If the state has PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6^^^^7 districts, PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6^^^^8 voters, and PRESERVED_PLACEHOLDER_^^^^6Fair Game6query6^^^^9 minority voters, the proportional share is defined as

PRESERVED_PLACEHOLDER_^^^^6Fair Game6all:\6Fair Game6^^^^

The central guarantee is that the minority can always secure at least PRESERVED_PLACEHOLDER_^^^^6Fair Game6all:\6Fair Game6^^^^ districts, while the majority has a complementary cracking upper bound expressed through

PRESERVED_PLACEHOLDER_^^^^6Fair Game6all:\6Fair Game6^^^^

(&&&^^^^6query6Fair Game6^^^^&&&). In this setting, “fair game” refers to a protocol-level near-proportional guarantee under perfect information and no geographic constraints.

Finally, the fairness of AI-versus-human game benchmarks has been treated as a taxonomy rather than a theorem. The relevant dimensions are input, output, compute, knowledge, experience, psychology, and common sense. On this view, a contest is fair only when no non-game-relevant circumstance advantages either side (&&&^^^^6query6Fair Game6^^^^&&&). This perspective is useful because it exposes a common ambiguity across the literature: “fair” may refer to equal payoffs, symmetric laws, rule compliance, equitable allocation, or parity of competitive conditions, and these are not interchangeable notions (&&&^^^^6query6Fair Game6^^^^&&&).

In that sense, the modern literature on fair games is best understood as a collection of formally precise but heterogeneous frameworks. Their common feature is not a universal fairness axiom, but the conversion of an intuitive fairness demand into a mathematically checkable object: an involution, a bifurcation threshold, a Shapley-proportional allocation, an auditor feedback loop, a Streett-like recurrence condition, or an explicit sportsmanship rule.