We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cond-mat

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Condensed Matter > Statistical Mechanics

Title: Nash equilibrium mapping vs Hamiltonian dynamics vs Darwinian evolution for some social dilemma games in the thermodynamic limit

Abstract: How cooperation evolves and manifests itself in the thermodynamic or infinite player limit of social dilemma games is a matter of intense speculation. Various analytical methods have been proposed to analyse the thermodynamic limit of social dilemmas. In a previous work [Chaos Solitons and fractals 135, 109762(2020)] involving one among us, two of those methods, Hamiltonian Dynamics(HD) and Nash equilibrium(NE) mapping were compared. The inconsistency and incorrectness of HD approach vis-a-vis NE mapping was brought to light. In this work we compare a third analytical method, i.e, Darwinian evolution(DE) with NE mapping and a numerical agent based approach. For completeness, we give results for HD approach as well. In contrast to HD which involves maximisation of payoffs of all individuals, in DE, payoff of a single player is maximised with respect to its nearest neighbour. While, HD utterly fails as compared to NE mapping, DE method gives a false positive for game magnetisation -- the net difference between the fraction of cooperators and defectors -- when payoffs obey the condition a+d=b+c, wherein a, d represent the diagonal elements and b, c the off diagonal elements in symmetric social dilemma games. When either a+d =/= b+c or, when one looks at average payoff per player, DE method fails much like the HD approach. NE mapping and numerical agent based method on the other hand agree really well for both game magnetisation as well as average payoff per player for the social dilemmas in question, i.e., Hawk-Dove game and Public goods game. This paper thus bring to light the inconsistency of the DE method vis-a-vis both NE mapping as well as a numerical agent based approach.
Comments: 15 pages, 4 figures, 2 tables
Subjects: Statistical Mechanics (cond-mat.stat-mech); Theoretical Economics (econ.TH); Physics and Society (physics.soc-ph); Populations and Evolution (q-bio.PE)
Cite as: arXiv:2103.00295 [cond-mat.stat-mech]
  (or arXiv:2103.00295v1 [cond-mat.stat-mech] for this version)

Submission history

From: Colin Benjamin [view email]
[v1] Sat, 27 Feb 2021 19:13:49 GMT (1445kb,D)

Link back to: arXiv, form interface, contact.