The biased interaction game: Its dynamics and application in modelling social systems
Phil Mercy, Martin Neil
Abstract
The biased interaction game described the operation of systems rooted in boundedly rational interactions under conditions of scarcity. The game explored the influence of bias and demonstrated how hierarchy and inequality are emergent system properties when sources of bias, such as power and scarcity, affect the outcome of interactions in an environment. Bias also impacts the likelihood of the emergence of cooperation. This paper serves as a companion piece to the paper introducing the biased interaction game. It investigates the general applicability of the game and demonstrates how the consideration of bias can modify and improve upon prior systems thinking. In particular, it shows how social systems can be successfully modelled using the biased interaction game and confirms its suitability for modelling extreme examples such as hyper-capitalism and social egalitarianism. It also reveals how biased systems can demonstrate non-linear behaviour, where long periods of system stability are punctuated by short bursts of rapid hierarchical transitions, mimicking real-world observations of social mobility. The paper concludes with a simplified real-world application, modelling the merits of two competing wealth redistribution philosophies: social welfare and a universal basic income
AI Impact Assessments
(3 models)Scientific Impact Assessment: "The Biased Interaction Game: Its Dynamics and Application in Modelling Social Systems"
1. Core Contribution
This paper serves as a companion to an earlier publication (Mercy & Neil, 2025) introducing the "biased interaction game" — a game-theoretic framework incorporating two key factors: agent incumbent value (accumulated wealth/power) and environmental scarcity. The core contribution of *this* paper is threefold: (a) testing edge-case behavior of the model (utopian equality, hyper-capitalism, social egalitarianism), (b) demonstrating emergent non-linear social mobility dynamics including "mobility cascades," and (c) applying the framework to compare social welfare versus universal basic income (UBI) redistribution policies.
The conceptual novelty lies in showing that a minimal set of boundedly rational interaction rules, combined with scarcity and incumbent value bias, can produce emergent hierarchy, social mobility with punctuated equilibrium dynamics, and differential responses to redistribution policies. The UBI vs. social welfare comparison is the most practically oriented contribution, suggesting that UBI preserves hierarchical structure while reducing inequality, whereas social welfare disrupts middle-band structure at lower tax rates.
2. Methodological Rigor
The methodological approach has several significant weaknesses:
Lack of statistical rigor: The paper explicitly states it "avoids an exhaustive statistical analysis of the model." For a simulation-based study, this is a critical limitation. Results are presented from single experimental runs (N=50 agents in most cases) without confidence intervals, sensitivity analyses, or multiple replications with different random seeds. The "RAND50" starting distribution is generated once and reused — while this aids reproducibility, it raises questions about generalizability of findings.
Small population sizes: Most experiments use N=50 agents, which is extremely small for drawing conclusions about social systems. While the appendix presents larger populations (up to 750), these experiments use only 2,000 iterations, acknowledged as insufficient for convergence.
Claims exceed evidence: The paper claims non-linear/chaotic behavior based on visual inspection of transition density plots (Figures 4-5) without formal tests for chaos (e.g., Lyapunov exponents, correlation dimension analysis). Citing a "complex non-linear systems commentator and entrepreneur" (Kieran Kelly) rather than established complexity science literature weakens the theoretical grounding. The distinction between stochasticity and deterministic chaos is not addressed.
Wealth redistribution model simplifications: The taxation model is highly simplified — flat tax on total wealth per iteration, binary redistribution schemes, no behavioral responses to taxation. The paper acknowledges these simplifications but still draws policy-relevant conclusions.
No validation against empirical data: While the paper mentions that S=0.3 produces Gini coefficients similar to the US and UK, no systematic calibration or validation against real-world wealth distributions, social mobility data, or redistribution outcomes is performed.
3. Potential Impact
The potential impact is moderate but constrained by the limitations noted above. The framework's ability to produce emergent hierarchy from simple rules is not itself novel — this has been demonstrated in numerous agent-based models (e.g., Sugarscape by Epstein & Axtell, 1996; Bonabeau's hierarchy models). The specific contribution of incorporating "bias" through incumbent value and scarcity is interesting but would benefit from more rigorous comparison with existing models.
The UBI vs. social welfare comparison could attract attention from policy researchers, though the simplifications are substantial enough that policy recommendations would be premature. The mobility cascade concept is potentially interesting for complexity science, but requires formal characterization.
4. Timeliness & Relevance
The paper addresses relevant topics — wealth inequality, social mobility, and redistribution policy are active areas of research and public discourse. UBI experiments are ongoing globally (Finland, Kenya, various U.S. pilots), making computational models of UBI effects timely. However, there is a substantial existing literature on agent-based models of wealth distribution and inequality (e.g., Bouchaud & Mézard, 2000; Chatterjee et al., 2004; the "yard sale" model of Boghosian, 2019) that is not discussed or compared against. This omission is a significant gap — the paper does not position itself relative to established computational economics or econophysics literature on wealth dynamics.
5. Strengths & Limitations
Strengths:
Limitations:
Overall Assessment
This paper presents an exploratory investigation of a novel game-theoretic framework with some interesting emergent properties. The mobility cascade phenomenon and the differential effects of redistribution schemes are genuinely intriguing results. However, the work suffers from insufficient statistical rigor, inadequate engagement with prior literature on computational models of wealth dynamics, absence of empirical validation, and a tendency to over-interpret simulation results. The paper reads more as a technical report exploring a model's behavior than as a rigorous scientific contribution. With substantial strengthening — formal statistical analysis, comparison with established models, empirical calibration, and engagement with the broader literature — this could become a meaningful contribution to computational social science.
Generated Mar 10, 2026
Comparison History (18)
Paper 1 likely has higher impact due to broader scope and generalizable novelty: it proposes and applies a new game-theoretic/complex-systems framework for biased interactions under scarcity, yielding emergent hierarchy, inequality, cooperation dynamics, and punctuated transitions—phenomena relevant across social science, economics, and systems modeling. Its applications (e.g., hyper-capitalism vs egalitarianism, welfare vs UBI) are timely and policy-relevant. Paper 2 is methodologically solid and empirically informed, but its core contribution is more domain-specific (sticker-trading norms), limiting breadth and cross-field uptake.
Paper 1 introduces a novel game-theoretic framework with broad interdisciplinary applications spanning social systems, economics, inequality, and policy analysis. It models emergent hierarchy, cooperation, non-linear dynamics, and compares wealth redistribution philosophies (welfare vs. UBI), offering significant real-world relevance. Paper 2, while solving an interesting algorithmic problem in fair division with improved efficiency, addresses a more narrowly scoped technical contribution. Paper 1's breadth of impact across sociology, economics, complexity science, and policy makes it likely to generate broader scientific interest and citations.
Paper 2 has higher potential impact due to greater conceptual novelty (a bias-aware interaction game with emergent inequality/cooperation and punctuated transitions), broader cross-field applicability (complex systems, economics, political science, sociology, policy), and clear real-world relevance via redistribution-policy modeling. Paper 1 is a scientometric review using established tools (CiteSpace, WoS) and is mainly descriptive/meta-analytic; it can be useful for mapping a field but is less likely to generate new theory or generalizable mechanisms. Overall, Paper 2 is more innovative and transferable.
Paper 1 addresses broad, highly relevant socio-economic issues such as inequality, social mobility, and wealth redistribution (UBI vs. welfare), offering significant potential for real-world application and cross-disciplinary impact in economics and sociology. In contrast, Paper 2, while methodologically rigorous, focuses on strategies for a specific traditional card game, heavily limiting its broader scientific relevance and practical impact.
Paper 2 has higher likely scientific impact due to stronger methodological rigor (clear probabilistic model, analytic results with asymptotics, and simulation validation) and tighter novelty (optimal strategic partitioning with a precise 1/5 limit). Its contributions connect to active fields (computational social choice, electoral design, mechanism manipulability) with broad applicability to political science, economics, and algorithms. Paper 1 is timely and potentially useful for social modeling, but the abstract suggests a more conceptual/systems-theory contribution with less clearly defined formalism or falsifiable/transferable results, which typically limits scientific uptake and cross-field reuse.
Paper 1 presents a novel game-theoretic framework with broad interdisciplinary applications spanning social sciences, economics, and complex systems. It addresses fundamental questions about hierarchy, inequality, and cooperation emergence, with practical policy applications (welfare vs. UBI). Its modeling of non-linear social dynamics and real-world validation demonstrates both theoretical depth and practical relevance. Paper 2, while technically sound, represents an incremental generalization of an existing algorithm to multiplayer settings, with narrower impact limited primarily to game AI/search algorithm research.
Paper 2 demonstrates higher potential scientific impact due to its broad interdisciplinary applications and high relevance to pressing real-world issues. While Paper 1 offers rigorous computational contributions to a specific subfield of algorithmic game theory, Paper 2 tackles widely relevant socio-economic phenomena like wealth redistribution, UBI, inequality, and social mobility. This broader scope and direct applicability to sociology, economics, and complex systems are likely to attract a wider readership and generate greater impact across multiple disciplines.
Paper 2 exhibits higher potential scientific impact due to its broader interdisciplinary reach, bridging game theory, complex systems, and sociology. While Paper 1 offers rigorous analysis within the specific niche of social choice theory, Paper 2 provides a generalized framework for understanding emergent properties like inequality, cooperation, and non-linear transitions. Its capacity to mathematically model diverse macro-level social and economic structures gives it wider applicability and methodological relevance to current complex systems research across multiple scientific domains.
Paper 2 presents a technically rigorous and novel framework (DID with DEBs) that bridges deep learning and game theory for automated incentive design. It demonstrates broader applicability across economics and computer science, offers stronger theoretical grounding (convex equilibrium selection, equivariant architectures), and addresses a timely need in multi-agent AI systems. Paper 1, while exploring interesting emergent phenomena, lacks statistical rigor, empirical validation, engagement with prior literature, and presents preliminary results from a simplified model with small populations.
Paper 2 has higher impact potential due to stronger methodological rigor (formal spectral/random-walk analysis, provable consensus and stability conditions, and empirical validation), clearer positioning within established theory, and broader applicability to multiplex-network phenomena relevant across network science, control, and social computing. Paper 1 is conceptually interesting and timely for inequality/redistribution debates, but its limited statistical robustness, small simulations, weak engagement with prior ABM/econophysics literature, and overextended claims without validation substantially reduce likely scientific uptake.