Inter-role reciprocity in evolutionary trust game on square lattices

Chaoqian Wang, Wei Zhang, Xinwei Wang, Attila Szolnoki · 2025

Read Inter-role reciprocity in evolutionary trust game on square lattices by Chaoqian Wang, Wei Zhang, Xinwei Wang, Attila Szolnoki (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-06685.txt.

Introduction

Chaoqian Wang, Wei Zhang, Xinwei Wang, and Attila Szolnoki study how trust can grow in a grid game in a 2025 paper. In the trust game, one side stakes funds and the other may repay or keep them. Two roles in one crowd are hard to stage, so the team uses a square grid that swaps roles by square.

Game play runs across roles, while learning runs within each role's own sub-grid. At a mid return rate, givers and honest takers bunch in mixed clusters that shield trust. Too high a return harms takers, and too low a return harms givers.

Catalog record

CREATORChaoqian Wang, Wei Zhang, Xinwei Wang, Attila Szolnoki
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE36 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-06685

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Inter-role reciprocity in evolutionary trust game on square lattices Preprint: AIP/123-QED Chaoqian Wang, Wei Zhang, Xinwei Wang, Attila Szolnoki Abstract Simulating bipartite games, such as the trust game, is not straightforward due to the lack of a natural way to distinguish roles in a single population. The square lattice topology can provide a simple yet elegant solution by alternating trustors and trustees. For even lattice sizes, it creates two disjoint diagonal sub-lattices for strategy learning, while game interactions can take place on the original lattice. This setup ensures a minimal spatial structure that allows interactions across roles and learning within roles. By simulations on this setup, we detect an inter-role spatial reciprocity mechanism, through which trust can emerge. In particular, a moderate return ratio allows investing trustors and trustworthy trustees to form inter-role clusters and thus save trust. If the return is too high, it harms the survival of trustees; if too low, it harms trustors. The proposed simulation framework is also applicable to any bipartite game to uncover potential inter-role spatial mechanisms across various scenarios. Trust behaviors are ubiquitous in human interactions. How does evolution allow trust in social networks? Since the trust game involves two roles, it is not straightforward to study in a single network. Here, we use a diagonally separated square lattice to simulate trust games. We report that trust can emerge on regular networks through an inter-role spatial reciprocity mechanism. Especially, evolution favors trust and trustworthiness most when the trustee returns moderately. Our results, thus, support field observations and reveal that even minimal network structures can support the emergence of trust. I Introduction Figure 1: The original square lattice (game network) and its two diagonal sub-lattices (learning networks). Trustors and trustees are arranged in an alternating pattern, allowing them to play games with the opposite role through the horizontal and vertical directions of the original lattice. The two diagonal sub-lattices are disjoint (if L is even), functioning as two independent strategy learning networks where learning only occurs between agents of the same role. Social evolution for various behaviors, such as cooperation Hamilton 1964 ; Axelrod and Hamilton 1981 , honesty Krakauer and Pagel 1995 ; Takahashi et al. 2012 ; Kumar et al. 2021 , morality Boehm 2012 ; Helbing et al. 2010 , and trust Abbass, Greenwood, and Petraki 2015 ; Chica et al. 2017 ; Kumar, Capraro, and Perc 2020 , reveals how complex social norms and prosocial behaviors can emerge and stabilize in populations through evolutionary mechanisms. In real-world societies, interactions are embedded in specific social network structures, rather than randomly well-mixed populations. This insight has motivated extensive research in spatial evolutionary game theory Nowak and May 1992 , where the network can play a crucial role in the emergence and stability of prosocial behaviors Allen et al. 2017 . The research of spatial evolutionary game theory has achieved substantial progress Perc et al. 2017 , both through computer simulations Szabó and Tőke 1998 ; Perc and Szolnoki 2010 and theoretical analyses Ohtsuki et al. 2006 ; Szabó and Fath 2007 ; Wang, Perc, and Szolnoki 2024 . Among various social dilemmas, the trust game Berg, Dickhaut, and McCabe 1995 has a specific feature because it distinguishes between two asymmetric roles: the trustor and the trustee. To capture the essence of trust relationships, the classic trust game involves a sequential interaction in which the trustor first decides whether and how much to invest, while the trustee subsequently chooses whether and how much to return Cox 2004 ; Johnson and Mislin 2011 . Unlike symmetric games, such as the prisoner’s dilemma, the trust game requires interactions across roles. Understanding the evolutionary dynamics of trust, thus, requires accounting for the role asymmetries, as they shape the conditions for the emergence of trust and trustworthiness Camerer 2011 . However, simulating asymmetric games, such as the trust game, in structured populations is not straightforward since distinct roles cannot be intuitively differentiated in a single network structure. To avoid this difficulty, previous studies mainly adopted two indirect approaches. (1) Roles can flow, as in the three-strategy N -player trust game Abbass, Greenwood, and Petraki 2015 ; Greenwood, Abbass, and Petraki 2016 ; Chica et al. 2017 ; Wang 2024 . This approach assumes that roles are embedded within strategies, allowing agents to change roles during strategy learning. As a result, it effectively reduces to a multi-strategy symmetric game. Within this framework, previous studies have explored the effects of various additional mechanisms on the evolution of trust, such as a margin system Guo et al. 2024 , loss assessment Liu et al. 2024 , different updating rules Chica et al. 2019a , logit dynamics Towers et al. 2017 , conditional investment in repeated interactions Liu and Chen 2022 , diverse investment patterns Shang, Zhang, and Chang 2023 , reputations Hu et al. 2021 ; Li et al. 2022 ; Xia, Hu, and Zhao 2022 ; Feng et al. 2023 , and punishment & reward mechanisms Chica et al. 2019b ; Fang and Chen 2021 ; Sun et al. 2022 ; Liu, Wang, and Li 2024 ; Liu et al. 2025 ; Wang et al. 2025 . (2) An individual plays both roles Burks, Carpenter, and Verhoogen 2003 . In this case, the strategies for each role are combined to define four “composite strategies,” which compete within the population as a multi-strategy symmetric game. Under this framework, an individual’s role remains fixed in strategy learning. Kumar et al. Kumar, Capraro, and Perc 2020 simulated the evolution of trust under this setting in structured populations and found that trust cannot emerge on lattice networks, but can emerge on scale-free networks. Lim and Capraro Lim and Capraro 2021 and Lim and Masuda Lim and Masuda 2024 also provided theoretical analyses of trust dynamics under this approach. These two approaches bypass the challenges of simulating multipartite games, but in doing so, they also compromise some essential characteristics of the bipartite scenarios. For instance, in approach (1), one may question the rationality of the additional assumption—why should a trustee be able to change their social status and become an investor merely because investors achieve higher payoffs, or vice versa? Similarly, in approach (2), it is unnatural that strategies across different roles are bundled. Even if we accept that an individual could simultaneously act as a trustor and a trustee, it remains unclear why strategies for both roles should change in lockstep. Actually, in a heterogeneous or hierarchical society, these roles cannot be mixed—some individuals are always in a trustor position, while others are mostly in a trustee position. To properly distinguish between different roles in structured populations, more sophisticated frameworks, such as interdependent multilayer networks Wang et al. 2015 ; Lugo and Miguel 2015 , may be required. In fact, it is because strict role separation complicates simulations in structured populations that the mentioned two approaches have often been adopted as compromises. Recently, Hauert and Szabó Hauert and Szabó 2025 pointed out that a square lattice can be divided into two sub-lattices by considering nearest- and next-nearest neighbors separately. If the lattice size is even, then these two sub-lattices are disjoint. On this structure, they observed spontaneous symmetry breaking in symmetric games. For

Read the complete text