Opinion formation and cyclic dominance in adaptive networks

The Rock-Paper-Scissors(RPS) game is a paradigmatic model for cyclic dominance in biological systems. Here we consider this game in the social context of competition between opinions in a networked society. In our model, every agent has an opinion which is drawn from the three choices: rock, paper or scissors. In every timestep a link is selected randomly and the game is played between the nodes connected by the link. The loser either adopts the opinion of the winner or rewires the link. These rules define an adaptive network on which the agent’s opinions coevolve with the network topology of social contacts. We show analytically and numerically that nonequilibrium phase transitions occur as a function of the rewiring strength. The transitions separate four distinct phases which differ in the observed dynamics of opinions and topology. In particular, there is one phase where the population settles to an arbitrary consensus opinion. We present a detailed analysis of the corresponding transitions revealing an apparently paradoxial behavior. The system approaches consensus states where they are unstable, whereas other dynamics prevail when the consensus states are stable.

💡 Research Summary

The paper “Opinion formation and cyclic dominance in adaptive networks” investigates how opinions evolve in a social network when the underlying interaction follows the classic Rock‑Paper‑Scissors (RPS) cyclic‑dominance game. Each agent (node) holds one of three possible opinions—rock, paper, or scissors. At every discrete time step a link is chosen uniformly at random, and the two connected agents play an RPS match. The loser has two possible responses: (i) adoption – the loser copies the winner’s opinion, or (ii) rewiring – the loser cuts the existing link and creates a new link to a randomly selected other node. The probability of rewiring is denoted by (p); when (p=0) the dynamics reduce to pure opinion copying, while (p=1) corresponds to maximal network plasticity.

The authors combine analytical techniques (mean‑field equations and a pair‑approximation closure) with extensive agent‑based simulations to explore the joint evolution of opinion fractions (\rho_R,\rho_P,\rho_S) and the network’s structural properties (average degree, degree correlations, clustering). Their analysis reveals that the system exhibits four distinct dynamical phases as a function of the rewiring strength (p):

1. Mixed‑Oscillatory Phase (low (p)) – For small rewiring probabilities the network remains essentially random. Opinion densities undergo sustained oscillations or quasi‑periodic fluctuations reminiscent of the classic RPS predator‑prey cycles. No opinion dominates for long periods, and the topology does not develop any pronounced community structure.

2. Unstable Consensus Phase (intermediate‑low (p)) – As rewiring increases past a first critical value (p_1), the system can temporarily converge to a consensus (all agents sharing the same opinion). However, linear stability analysis shows that these consensus fixed points are unstable: any infinitesimal perturbation—inevitable due to stochastic link selection—drives the system back into the mixed oscillatory regime. This paradoxical behavior—approaching an unstable state—constitutes one of the paper’s central findings.

3. Stable Consensus Phase (intermediate‑high (p)) – Beyond a second threshold (p_2) the consensus states become stable. The rewiring process now preferentially isolates dissenting agents, allowing the majority opinion to lock in. The network self‑organizes into tightly knit clusters of like‑minded nodes (high assortativity), effectively forming an “echo chamber”. Opinion dynamics freeze, and the system remains in the consensus indefinitely unless external noise is introduced.

4. Fragmented Multi‑Opinion Phase (high (p)) – For rewiring probabilities larger than a third critical value (p_3), the network fragments into many small components, each preserving a different opinion. The high plasticity prevents any single opinion from sweeping the whole population; instead, the system exhibits a modular structure with high intra‑community connectivity and sparse inter‑community links. This phase is characterized by persistent opinion diversity and a high modularity index.

The analytical predictions for the critical points (p_1, p_2, p_3) match the simulation results across a wide range of system sizes (from (N=10^4) to (N=10^5)) and average degrees ((\langle k\rangle\approx 6)). Moreover, the authors test alternative rewiring rules—such as assortative rewiring that preferentially connects to nodes sharing the same opinion—and find that the location of the phase boundaries shifts, but the qualitative four‑phase structure persists. This robustness suggests that the observed phenomena are not artifacts of a specific implementation but rather intrinsic to the feedback loop between cyclic dominance and adaptive topology.

From a sociophysical perspective, the model captures several realistic mechanisms of modern online platforms. “Adoption” mirrors opinion change through persuasive interaction, while “rewiring” abstracts actions such as unfriending, unfollowing, or algorithmic recommendation of new contacts. The existence of an unstable consensus regime implies that even when a platform appears to be converging toward a single narrative, the underlying dynamics may be poised to revert to a pluralistic state if the network’s rewiring intensity is insufficient to lock in the consensus. Conversely, excessive rewiring can entrench echo chambers (stable consensus) or, if too aggressive, lead to fragmentation and polarization (fragmented phase).

The paper concludes by outlining several promising extensions: (i) increasing the number of competing opinions to explore higher‑order cyclic dominance, (ii) embedding the dynamics on heterogeneous degree distributions (e.g., scale‑free networks) to assess the role of hubs, and (iii) incorporating external fields such as media influence or targeted advertising. Such extensions would bring the model closer to empirical social systems and could inform the design of interventions that either promote healthy opinion diversity or prevent the emergence of harmful consensus.

In summary, the study provides a rigorous theoretical and computational framework for understanding how cyclic dominance and network adaptivity jointly shape opinion formation. By identifying four qualitatively distinct phases and highlighting the counter‑intuitive “approach to unstable consensus” phenomenon, it offers fresh insights into the non‑equilibrium nature of social dynamics and underscores the importance of considering both opinion update rules and the malleability of social ties when analyzing collective behavior.