http://www.bibsonomy.org/burst/user/bertil.hatt/SocialBibSonomy BuRST Feed for /user/bertil.hatt/Social2008-09-06T20:19:46+02:00March 2008http://www.bibsonomy.org/bibtex/25c3c867d2ba22cc0d841a2f3f43757a4/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00and Social Holes Networks Structral <span style="color:#555555;">Sanjeev <a href="http://www.bibsonomy.org/author/Goyal">Goyal</a> und Fernando <a href="http://www.bibsonomy.org/author/Vega-Redondo">Vega-Redondo</a> </span>(<em>2004</em>) <em>working paper . </em>Thu Mar 13 16:33:57 CET 2008working paperStructural Holes in Social Networks2004and Social Holes Networks Structral We consider a setting where every pair of players that interact (e.g. exchange goods or information) create a surplus. An interaction can take place only if the players involved have a connection. If the connection is direct the two players split the surplus equally while if it is indirect then intermediate players also get a share of the surplus. Thus individuals form link with others to create surplus, to gain intermediation rents and ti circumvent others who are trying to becom inermediary. Our principal result is that stratgic link formation in such a setting leads to the star network. In a star a single agent acts as an intermediary for all transactions and there is significant payoff inequality across ex-ante identical players.March 2008http://www.bibsonomy.org/bibtex/2fbc1c9a0012e5909fe7f37c56497f33f/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00and Networks Static Social q-component Model <span style="color:#555555;">D-H. <a href="http://www.bibsonomy.org/author/Kim">Kim</a> und B <a href="http://www.bibsonomy.org/author/Kahng">Kahng</a> und D. <a href="http://www.bibsonomy.org/author/Kim">Kim</a> </span>(<em>2005</em>) <em>arXiv.org . </em>Thu Mar 13 16:33:57 CET 2008arXiv.orgThe {\$}q{\$}-component Static Model: Modeling Social Networks2005and Networks Static Social q-component Model We generalize the static model by assigning a q-component weight on each vertex. We first choose a component ({\$}/micro{\$}) among the {\$}q{\$} components at random and a pair of vertices is linked with a color {\$}/micro{\$} according to their weights of the component as a static model. A {\$}(1-f){\$} fraction of the entire edges is connected following this way. The remaining fraction {\$}f{\$} is added with the {\$}(q+1){\$}-th color as in the static model but using the maximum weights among the {\$}q{\$} components each individual has. This model is motivated by social networks. It exhibits similar topological features to real social networks in that: (i) the degree distribution as a highly skewed form, (ii) the diameter is as small and (iii) the assortativity coefficient {\$}r{\$} is as positive and large as those in real social networks with {\$}r{\$} reaching a maximum around {\$}f /approx 0.2{\$}.March 2008http://www.bibsonomy.org/bibtex/26abd95a03c65bf072d08794c0d52dce4/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Innovation Influence Structure Diffusion Learning Networks Law Social Power and Curve <span style="color:#555555;">Alexandre <a href="http://www.bibsonomy.org/author/Steyer">Steyer</a> und Jean-Beno^it <a href="http://www.bibsonomy.org/author/Zimmerman">Zimmerman</a> </span><em>Math\'ematiques \\&amp; Sc. Humaines</em><em>42(168):43--57</em>(<em>2004</em>)Thu Mar 13 16:33:57 CET 2008Math{\'e}matiques \{\&} Sc. Humaines16843--57Influence sociale et diffusion de l'innovation422004Innovation Influence Structure Diffusion Learning Networks Law Social Power and Curve The notion of diffusion holds a central place in any social system, because it is at the heart of individuals behavior or representation phasing, hence of the coordination of their actions. The idea at the origin of the notion of diffusion is that inter-individual interactions are the driving forces of the evolution of individuals' behaviours, beliefs and representations. Our approach in this paper is based on social influence networks. Agents are embedded in network structures where the influence advance depends on cumulative effects. First we draw the foundations of a diffusion model based on social influence networks. Then we study the way of propagation of influence trough ``avalanches'' giving a central importance to the network topology. We consider the noise produced by those avalanches as a characteristic of the social structure that can contribute, by learning effect, to transform the network structure, hence the dynamics of the diffusion. We then explain why peculiar ``critical'' diffusion curves do emerge characterized by a power law instead of the exponential form of traditional diffusion curves.March 2008http://www.bibsonomy.org/bibtex/21c2d1d0e3ffcf36ffb0d1e8054d9ed52/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Analysis and Modeling Social Nerwork Mathematical <span style="color:#555555;">Alessandro <a href="http://www.bibsonomy.org/author/Lomi">Lomi</a> und Philippa E. <a href="http://www.bibsonomy.org/author/Pattison">Pattison</a> </span><em>CMOT</em>(<em>2004</em>)Thu Mar 13 16:33:57 CET 2008CMOT5--15Introduction to the CMOT Special Issue on Mathematical Representations and Models for the Analysis of Social Networks within and between Organization102004Analysis and Modeling Social Nerwork Mathematical March 2008http://www.bibsonomy.org/bibtex/2568a84f64fffc4f735ac6a6cd799136c/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Diffusion Studies Networks Conferencing Technology Social and Adoption CSCW Data <span style="color:#555555;">Gloria <a href="http://www.bibsonomy.org/author/Mark">Mark</a> und Steven <a href="http://www.bibsonomy.org/author/Poltrock">Poltrock</a> und Danyel <a href="http://www.bibsonomy.org/author/Fisher">Fisher</a> </span><em>Communication Dimensions</em>(<em>2001</em>)Thu Mar 13 16:33:57 CET 2008Communication Dimensions53--70Remote Communication and Technology Diffusion2001Diffusion Studies Networks Conferencing Technology Social and Adoption CSCW Data The rise of virtual collocation in distributed organizations may lead to new patterns of technology adoption. Whereas previous studies of technology diffusion (e.g. Rogers 1995) point the role of mass media and interpersonal communication in adoption, we find that collaborating partners who rarely see one another are important contributors to the diffusion of virtual collaboration technologies. We studied a large distributed organization to discover how and why a data conferencing technology was disseminated rapidly in a relatively short-time. We interpret our results to show that the technology has on an on-demand basis rather than through formal channels that the company had established. We connect this growth in usage with Rogers (1995) theories about the diffusion of innovations and the social network in place at this organization.March 2008http://www.bibsonomy.org/bibtex/2812063b822ecf8e06a62e9550fe5b139/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Heterogeneity Information Social and Model Networks Diffusion <span style="color:#555555;">Vincent <a href="http://www.bibsonomy.org/author/Buskens">Buskens</a> und Kazuo <a href="http://www.bibsonomy.org/author/Yamaguchi">Yamaguchi</a> </span><em>Soc. Methodology</em>(<em>1999</em>)Thu Mar 13 16:33:57 CET 2008Soc. Methodology281--325A New Model for Information Diffusion in Heterogeneous Social Networks291999Heterogeneity Information Social and Model Networks Diffusion This paper discusses a new model for the diffusion of information through heterogeneous social networks. In the earlier models, when information was given by one actor to another the sender did not retain the information. The new model is an improvement over the earlier ones because it allows a sender of information to retain that information after telling it to somebody else. Consequently, the new model allows more actors to have information during the information diffusion process. The model provides predictions of diffusion times in a given network at the global, dyadic, and individual levels. This leads to straightforward generalizations of network measures, such as closeness centrality and betweenness centrality, for research problems that focus on the efficiency of information transfer in a network. We analyze in detail how information diffusion times and centrality measures depend on a series of network measures, such as degrees, bridges, etc. One important finding is that predictions about the time actors need to spread information in the network differ considerably between the new and old models, while the predictions about the time needed to receive information hardly differ. Finally, some cautionary remarks are made about using the model in empirical research.March 2008http://www.bibsonomy.org/bibtex/26ed7fa389b55a6645bff2248c3d1a4e3/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Opinion Social Formation Dynamics and <span style="color:#555555;">M. F. <a href="http://www.bibsonomy.org/author/Laguna">Laguna</a> und Guillermo <a href="http://www.bibsonomy.org/author/Abramson">Abramson</a> und Dami'an <a href="http://www.bibsonomy.org/author/Zanette">Zanette</a> </span><em>Physica A</em><em>329(3-4):459--472</em>(<em>2003</em>)Thu Mar 13 16:33:57 CET 2008Physica A3-4459--472Vector Opinion Dynamics in a model for social influence3292003Opinion Social Formation Dynamics and We present numerical simulations of a model of social influence, where the opinion of each agent is represented by a binary vector. Agents adjust their opinion as a result of random encounters, whenever the difference between opinions is below a given threshold. Evolution leads to a steady state, which highly depends on the threshold and a convergence parameter of the model. We analyze the transition between clustered and homogenous steady states. Results of the cases of complete mixing and small-world networks are compared.March 2008http://www.bibsonomy.org/bibtex/217d85547dfcb5cb1ca617b56e829051e/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Networks Neighbourhood and Social <span style="color:#555555;">Patricia <a href="http://www.bibsonomy.org/author/Pattison">Pattison</a> und Garry L <a href="http://www.bibsonomy.org/author/Robins">Robins</a> </span>(<em>2001</em>)Thu Mar 13 16:33:57 CET 2008Neighbourhood-based models for social networks2001Networks Neighbourhood and Social We argue that social networks can be modeled as the outcome of processes that occur in overlapping local regions of the network, termed _local social neighborhoods_. Each neighborhood is conceived as a possible site of interaction and corresponds to a subset of possible network ties. In this paper, we discuss hypotheses about the form of these neighborhoods, and we present two new and theoretically plausible ways in which neighborhood-based models for networks can be constructed. In the first, we introduce the notion of a _setting structure_, a directly hypothesized (or observed) set of exogenous constraints on possible neighborhood forms. In the second, we propose higher-order neighborhoods that are generated, in part, by the outcome of interactive network process themselves. Applications of both approaches to model construction are presented, and the developments are considered within a general conceptual framework of locale for social networks. We show how assumptions about neighborhoods can be cast within a hierarchy of increasingly complex models; these models represent a progressively greater capacity for network processes to ``reach'' across a network through long cycles or semi-paths. We argue that this class of models holds new promise for the development of empirically plausible models for networks and networks-based processes.March 2008http://www.bibsonomy.org/bibtex/2a9ec6de12243ddc321a2a7325525a055/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Markov Small Networks World Model Social and Chains <span style="color:#555555;">Peter <a href="http://www.bibsonomy.org/author/Killworth">Killworth</a> und Christopher <a href="http://www.bibsonomy.org/author/McCarthy">McCarthy</a> und H. <a href="http://www.bibsonomy.org/author/Russell Bernard">Russell Bernard</a> und Mark <a href="http://www.bibsonomy.org/author/House">House</a> </span><em>SocNet</em><em>28(1):86--96</em>(<em>2006</em>)Thu Mar 13 16:33:57 CET 2008SocNet186--96The Accuracy of Small World Chains in Social Networks282006Markov Small Networks World Model Social and Chains We analyse 10,920 shortest path connections between 105 members of an interviewin bureau, together with the equivalent conceptual, or `small world' routes, which use individuals' selections of intermediaries. This permits the first study of the impact of accuracy within small world chains. The mean small world par length (3.23) is 40% longer than the mean of the actual shortest paths (2.30), showing that mistakes are prevalent. A Markov model with a probability of simply guessing an intermediary of 0.52 gives an excellent fit to the observations, suggesting that people make the wrong small world chance more than half the time.March 2008http://www.bibsonomy.org/bibtex/234eaa0ed0ba6b03a6a4fc735950256ac/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Policy, Morphogenesis Network, Morphogenesis, Social Technological Influence Innovation, Hopf Network and Innovation Bifurcation <span style="color:#555555;">Fr'ed'eric <a href="http://www.bibsonomy.org/author/Dero\u{\i}an">Deroian</a> </span>(<em>2000</em>) <em>Grequam . </em>Thu Mar 13 16:33:57 CET 2008GrequamMorphogenesis of Social Network and Coexistence of Technologies2000Policy, Morphogenesis Network, Morphogenesis, Social Technological Influence Innovation, Hopf Network and Innovation Bifurcation March 2008http://www.bibsonomy.org/bibtex/2a50644674b22dd502391ff4098104371/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Learning Social externality Adoption Diffusion and Network Criticality influence <span style="color:#555555;">Franck <a href="http://www.bibsonomy.org/author/Plourabou{\&#039;e}">Plourabou'e</a> und Alexandre <a href="http://www.bibsonomy.org/author/Steyer">Steyer</a> und Jean-Beno^it <a href="http://www.bibsonomy.org/author/Zimmerman">Zimmerman</a> </span>(<em>1996</em>) <em>working paper . </em>Thu Mar 13 16:33:57 CET 2008working paperLearning Induced Criticality in Consumers' Adoption Pattern: A Neural Network Approach Learning Induced Criticality in Consumers' Adoption Pattern: A Neural Network Approach Learning Induced Criticality in Consumers' Adoption Pattern: A Neural Network Approach1996Learning Social externality Adoption Diffusion and Network Criticality influence The aim of this paper is to lay the foundations of a social influence based approach for the diffusion of an innovation or a technological standard. A model built on the principles of a neural network is proposed and a learning procedure is set up, making the network formation endogenous, the strength of connections among agents being determined by their shared histories. Referring to the concept of criticality developed by physicists, it shall be shown that learning, in a social structure, can lead the network to a critical state, called `learning induced criticality', where some agents are able to exert a macroscopic influence over the network. The distribution of influence spheres' size follows a Pareto law. This approach shows an interesting similarity with that of the social coherence in sociology, whereby individuals within a social structure are led to share a close assessment of a given innovation.March 2008http://www.bibsonomy.org/bibtex/2dd6d5bcff2977888635ea8f584389d3b/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Links Networks Law Evolution Power Social Diffusion Structure Learning and Influence <span style="color:#555555;">Alexandre <a href="http://www.bibsonomy.org/author/Steyer">Steyer</a> und Jean-Beno^it <a href="http://www.bibsonomy.org/author/Zimmerman">Zimmerman</a> </span>(<em>2005</em>) <em>working paper Gr'equam . </em>Thu Mar 13 16:33:57 CET 2008working paper Gr{\'e}quamSocial Networks and Diffusion: Avalanches and Links Evolution2005Links Networks Law Evolution Power Social Diffusion Structure Learning and Influence The concept of diffusion is central to every social system, because it underpins the coherence of individuals' behaviour and representations, and hence the coordination of their actions. The idea at the origin of the concept of diffusion is that inter-individual interactions are the driving force behind the evolution of individual's behaviour, beliefs and representations. Our approach in this paper in this paper is based on social influence networks. Agents are embedded in social networks where the advance of influence depends on the propagation of ``avalanches'', giving central importance to the network structure. We consider the noise produced by those avalanches as the characteristic of the social structure that can contribute, through the evolution of links, to transforming the network structure, and hence the dynamics of the diffusion. We then explain why peculiar ``critical'' diffusion dynamics emerge, characterised by a power law distribution, instead of the exponential shape of traditional diffusion curves.March 2008http://www.bibsonomy.org/bibtex/242420b959ce9133460c70c6fc486b409/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Medical Innovation contagion Social and Marketing <span style="color:#555555;">Christophe Van den <a href="http://www.bibsonomy.org/author/Butte">Butte</a> und Gary <a href="http://www.bibsonomy.org/author/Lilien">Lilien</a> </span><em>AJS</em><em>106(5):1409--1435</em>(<em>2001</em>)Thu Mar 13 16:33:57 CET 2008AJS51409--1435Medical Innovation Revisited: Social Contagion versus Marketing Effort1062001Medical Innovation contagion Social and Marketing This article shows that Medical Innovation---the landmark study by Coleman, Katz and Menzel---and several subsequent studies analyzing the diffusion of the drug tetracycline have confounded social contagion with marketing effect. The article describes the medical community's understanding of tetracycline and how the drug was marketed. This situational analysis finds no reasons to expect social contagion: instead, aggressive marketing efforts may have played an important role. The Medical Innovation data set is reanalyzed and supplemented with newly collected advertising data. When marketing efforts are controlled for, contagion effects disappear. The article underscores the importance of controlling for potential confounds when studying the role of social contagion in innovation diffusion.March 2008http://www.bibsonomy.org/bibtex/25108e18b83eb194d91eddc482d54fdcc/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Diversity Unanimity and Structure Networks Social <span style="color:#555555;">Fr'ed'eric <a href="http://www.bibsonomy.org/author/Dero\u{\i}an">Deroian</a> </span><em>Annales d'\'Eco. \\&amp; Stat.</em>(<em>2004</em>)Thu Mar 13 16:33:57 CET 2008Annales d'{\'E}co. \{\&} Stat.209--224Comportements collectifs dans les r{\'e}seaux d'influence sociale742004Diversity Unanimity and Structure Networks Social Social influence plays a significant role in lots of economic situations (Becker, 1991, about restaurants, Valente, 1995, on diffusion of innovations, or Topol, 1991, for the functioning of financial markets). Yet, this externality is not always unilateral. Does social influence always impact individual behaviors in the sense of social cohesion or can it entail antagonist phenomena? On the basis of a simple threshold model, we suggest that positive retroactions induced by social influence can originate both unanimity and diversity in individual behaviors, and we rely this to the structure of the influence network. Particularly, we develop a simple model with non strategic and interacting agents subjected to cumulation in social influence. We determine first a necessary and sufficient condition to establish autonomous behaviors. When relaxing the condition, we find in simple cases conditions for consensus in behaviors, based on a mechanism of critical mass like in global interaction models. But studying direct and differentiated interactions is interesting to the extent that structural effects are not reducible to global dynamics. These structural effects are crucial in some circumstances. When behaviors are strongly interdependent, taking account of both social structure and initial opinions is necessary to determine the stationary state of the system. The spectrum of the social influence matrix contains these structural effects. We establish a simple condition related to the eigenvalue of smallest modulus, which entails strong interdependent opinions. Then we take the example of a social structure containing strongly connected subgroups with weak interconnections, and we show that the whole spectrum is required in order to foresee between unanimity and diversity in the long run.March 2008http://www.bibsonomy.org/bibtex/28e4e51ef83d0dfbaafdf3becaa4c16d9/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00diffusion networks Learning Intermediarity Social and Critical State Innovation <span style="color:#555555;">Fr'ed'eric <a href="http://www.bibsonomy.org/author/Dero\u{\i}an">Deroian</a> und Alexandre <a href="http://www.bibsonomy.org/author/Steyer">Steyer</a> und Jean-Beno^it <a href="http://www.bibsonomy.org/author/Zimmerman">Zimmerman</a> </span>(<em>2003</em>) <em>working paper . </em>Thu Mar 13 16:33:57 CET 2008working paperApprentissage social et diffusion de l'innovation : r{\'e}seaux critiques et interm{\'e}diarit{\'e}2003diffusion networks Learning Intermediarity Social and Critical State Innovation Le r{\^o}le des r{\'e}seaux sociaux dans la diffusion de l'information demeure une question strat{\'e}gique. Dans des travaux ant{\'e}rieurs, nous avons introduit un apprentissage relationnel, de type hebbien, qui conduit {\`a} un {\'e}tat critique dans lequel certains agents aqui{\`e}rent des positions, purement structurelles, de leaders d'opinion. Dans cet article, nous montrons que l'auto-organisation d'un r{\'e}seau d'influence, par effet de l'apprentissage social, ne constitue pas un ph{\'e}nom{\`e}ne monotone, mais aussi bien du point de vue des caract{\'e}ristiques structurelles du r{\'e}seau que de celui de ses performances en diffusion. Ceci n{\'e}cessite, pour {\^e}tre analys{\'e}, de recourir {\`a} la notion d'interm{\'e}diarit{\'e} qui est inh{\'e}rente au concept de r{\'e}seau. Une analyse relative au r{\^o}le des liens faibles dans les diff{\'e}rents r{\'e}gimes de diffusion devrait alors permettre d'offrir un {\'e}clairage nouveau sur cette dynamique d'{\'e}volution.March 2008http://www.bibsonomy.org/bibtex/29e01335583e20bd5aba58cacb98404e0/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00games, emergence cooperation, networks agents., evolutionary endogenous social of selection interactions, heterogeneous <span style="color:#555555;">David <a href="http://www.bibsonomy.org/author/Chavalarias">Chavalarias</a> </span>(<em>2005</em>) <em>arXiv.org . </em>Thu Mar 13 16:33:57 CET 2008arXiv.orgModeling Endogenous Social Networks: The Example of Emergence and Stability of Cooperation without Refusal2005games, emergence cooperation, networks agents., evolutionary endogenous social of selection interactions, heterogeneous March 2008http://www.bibsonomy.org/bibtex/20678507dcac033f699f19d29409953ce/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Bifurcation Diffusion Learning Network Interaction Social and Slow <span style="color:#555555;">Fr'ed'eric <a href="http://www.bibsonomy.org/author/Dero\u{\i}an">Deroian</a> </span><em>Research Policy</em><em>1331(1):1--12</em>(<em>2001</em>)Thu Mar 13 16:33:57 CET 2008Research Policy11--12Formation of social networks and diffusion of innovations13312001Bifurcation Diffusion Learning Network Interaction Social and Slow Some innovations need delay to diffuse, others often fail. The formation of social networks is a possible ecplanation. Considering a population of potential adopters of a technology, we set-up a model composed of inteacting agents. Interaction is conceived as ofluence effects and the network of interpersonal influences is learning step-by-step. The gradual foramtion of the social network leads, after a period of latency, to a collective evaluation of the innovation.March 2008http://www.bibsonomy.org/bibtex/21f8f1282da92501aa58cda919b9b28d5/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Community Biological Social Structure Networks and <span style="color:#555555;">Michelle <a href="http://www.bibsonomy.org/author/Girvan">Girvan</a> und Mark E. <a href="http://www.bibsonomy.org/author/J Newman">J Newman</a> </span>(<em>2002</em>)Thu Mar 13 16:33:57 CET 2008PNASCommunity structure in social and biological networks2002Community Biological Social Structure Networks and A number of recent studies have focused on the statistical properties of networked systems such as social networks and the Worldwide Web. Researchers have concentrated particularly on a few properties that seem to be common to many networks: the small-world property, power-law degree distributions, and network transitivity. In this article, we highlight another property that is found in many networks, the property of community structure, in which network nodes are joined together in tightly knit groups, between which there are only looser connections. We propose a method for detecting such communities, built around the idea of sing centrality indices to find community structure is already known and find that the the method detects this known structure with high sensitivity and reliability. We also apply the method to two networks whose community structure is not well known---a collaoration network and a food web---and find it detects significant and informative community divisions in both cases.March 2008http://www.bibsonomy.org/bibtex/2f188c8558e095a05d233bfdfd49bc153/bertil.hattbertil.hatt2008-03-13T16:33:57+01:00Diffusion and Social Networks <span style="color:#555555;">H. <a href="http://www.bibsonomy.org/author/Peyton Young">Peyton Young</a> </span>(<em>1999</em>) <em>Center on Social and Economic Dynamics . </em>Thu Mar 13 16:33:57 CET 2008Center on Social and Economic DynamicsDiffusion in Social Networks1999Diffusion and Social Networks We consider processes in which norms of mehavior are transmitted through social or geographic networks. Agents adopt behaviors based on a combination of their inherent payoff anf their local popularity (the number of neighbors who have adopted them) subject to some random error. Extending work of Blume (1993, 1995), Ellison (1993) and Morris (1997), we characerize the long-run dynamics of such processes in terms of the geometry of of the network, but without placing a priority restriction on the network structure. We show first that the relative likelihood of different states can be described in terms of a potential function that is inversely related to the length of the boundary between regions where norms shere norms of behaviour differ. As in a variety of other evolutionnary models, the most likely state is the one in which everyone is coordinated on the risk-dominant equilibrium. We then show that, if agents interact in sufficiently small, close-knit groups, the expected waiting time until almost everyone is playing the risk-dominant equilbrium is bounded above independently of the number of agents and independently of the initial state. Simulation results indicate that convergence is surprisingly rapid, even in very large networks, provided they are close-knit.