How network position shapes strategic communication

What the study found

The authors study a multiplayer coordination game in which privately informed agents communicate through a social network and choose actions while balancing their own preferred outcomes, or “bliss points,” against those of their neighbours.

They report that equilibrium actions are weighted averages of the players’ expected bliss points. The weights depend on a network-importance measure called degree-discounted Bonacich centrality. The authors also characterize circumstances in which players report private information truthfully, overstate it or understate it; incentives to misreport depend critically on the receiver’s network position.

Even with identical preferences, communication need not fully reveal private information. The authors say information transmission can remain coarse in a star network, while becoming almost fully informative in a complete network. When players choose their biases within the model, their strategic responses to neighbours’ biases depend on centrality, and an unbiased outcome may be a Nash equilibrium.

Why it matters

The model highlights a possible distinction between shared preferences and full disclosure: alignment alone does not necessarily guarantee informative communication. It also offers a framework for considering how a network’s structure relates to strategic messages and coordinated choices.

Limits and unknowns

These are theoretical findings reported in a preprint, not evidence that the same patterns occur in a particular real-world network. The abstract does not provide the model’s full assumptions, proofs, precise conditions for its results, or details of how closely its networks represent actual social relationships.