Equilibrium selection in coordination games with simultaneous play

Summary. We apply the dynamic stochastic framework proposed in recent evolutionary literature to a class of coordination games played simultaneously by the entire population. In these games payoffs, and hence best replies, are determined by a summary statistic of the population strategy profile. We demonstrate that with simultaneous play, the equilibrium selection depends crucially on how best responses to the summary statistic remain piece-wise constant. In fact, all the strict Nash equilibria in the underlying stage game can be made stochastically stable depending on how the best response mapping generates piece-wise constant best responses. Received: February 12, 2001; revised version: October 29, 2001     

Two equivalence results for two-person strict games

A game is strict if for both players, different profiles have different payoffs. Two games are best response equivalent if their best response functions are the same. We prove that a two-person strict game has at most one pure Nash equilibrium if and only if it is best response equivalent to a strictly competitive game, and that it is best response equivalent to an ordinal potential game if and only if it is best response equivalent to a quasi-supermodular game.     

Equilibrium play and best response to (stated) beliefs in normal form games

We report experimental results on a series of ten one-shot two-person 3×3 normal form games with unique equilibrium in pure strategies played by non-economists. In contrast to previous experiments in which game theory predictions fail dramatically, a majority of actions taken coincided with the equilibrium prediction (70.2%) and were best-responses to subjects' stated beliefs (67.2%). In constant-sum games, 78% of actions taken were predicted by the equilibrium model, outperforming simple K-level reasoning models. We discuss how non-trivial game characteristics related to risk aversion, efficiency concerns and social preferences may affect the predictive value of different models in simple normal form games.     

Fictitious play in ‘one-against-all’ multi-player games

Summary. A compound game is an (n + 1) player game based on n two-person subgames. In each of these subgames player 0 plays against one of the other players. Player 0 is regulated, so that he must choose the same strategy in all n subgames. We show that every fictitious play process approaches the set of equilibria in compound games for which all subgames are either zero-sum games, potential games, or games. Received: July 18, 1997; revised version: December 4, 1998     

Interconnected games and international environmental problems

The purpose of this paper is to introduce the concept of interconnected games and to show its relevance for modeling international environmental problems. It is argued that an interconnected game approach to international environmental problems may enhance cooperation and provide an alternative to the use of financial side payments to induce countries to cooperate. Two types of interconnected games are distinguished in this paper, i.e. direct sum games and tensor games. In the former all the constituting isolated games are games in strategic form and in the latter they are repeated games. In both cases the interconnected game can be interpreted as a multiple objective game, but only the setting where a trade-off is made for the vector-payoffs is considered. In addition to the formal definition of these types of interconnected games, some elementary results concerning Nash equilibria of such games are derived.Folmer and v. Mouche: Landbouwuniversiteit Wageningen, Postbus 8130, 6700 EW Wageningen, The Netherlands; Ragland: University of Colorado at Boulder, Boulder, Colorado 80309-0256, U.S.A. An earlier version of this paper was presented at the EAERE conference in Stockholm, June 1991. The authors appreciate comments made by conference participants and journal referees.     

On the limit perfect public equilibrium payoff set in repeated and stochastic games

This paper provides a dual characterization of the existing ones for the limit set of perfect public equilibrium payoffs in a class of finite stochastic games (in particular, repeated games) as the discount factor tends to one. As a first corollary, the folk theorems of Fudenberg et al. (1994), Kandori and Matsushima (1998) and Hörner et al. (2011) obtain. As a second corollary, it is shown that this limit set of payoffs is a convex polytope when attention is restricted to perfect public equilibria in pure strategies. This result fails for mixed strategies, even when attention is restricted to two-player repeated games.     

The evolution of conventions in minimum effort games

The literature on minimum effort game has been concerned with a symmetric game with linear payoff functions. The main aim of the present paper is to study the coordination problem arising in a not necessarily symmetric minimum effort game with two players. The sources of asymmetry can be twofold: the productivity of effort and the distribution of the join output. To select among the Pareto ranked equilibria we use the stochastic stability criterion. We show that, for any configuration parameters, the set of stochastically stable equilibria coincides with the set of potential maximizers. We also show that when the disutility of effort is linear, the Pareto dominant equilibrium is stochastically stable provided that the distributive parameter belongs to a well defined range. When the disutility of effort is nonlinear no distributive arguments can be used to successfully affect the selection process. Lastly we prove that the connection between stochastic stability and maximum potential can fail when more than two agents are considered.     

Iterated strict dominance in general games

We offer a definition of iterated elimination of strictly dominated strategies (IESDS^*) for games with (in)finite players, (non)compact strategy sets, and (dis)continuous payoff functions. IESDS^* is always a well-defined order independent procedure that can be used to solve Nash equilibrium in dominance-solvable games. We characterize IESDS^* by means of a “stability” criterion, and offer a sufficient and necessary epistemic condition for IESDS^*. We show by an example that IESDS^* may generate spurious Nash equilibria in the class of Reny's better-reply secure games. We provide sufficient/necessary conditions under which IESDS^* preserves the set of Nash equilibria.     

Scale functions in equilibrium selection games

We present here an evolutionary game model, and address the issue of equilibrium selection working with the scale function of a diffusion process describing the dynamics of population processes with mutation modeled as white noise. This model is the same as the one in Foster and Young (1990) but with a different interpretation at the boundaries and with different mutation modelings. First, we justifiably assume that the boundaries of the solution of the stochastic differential equation are absorbing so that the first boundary of the interval [0,1] hit will determine the equilibrium selected. Then, working with the scale function, we obtain for 2×2 symmetric games and different mutation parameters, some new and interesting equilibrium selection results. The aim of this article is to describe another method of approach in evolutionary games with mutation which we believe will prove to be very useful in studying more general normal form games and different mutation modelings.     

Playing safe in coordination games:: the roles of risk dominance, payoff dominance, and history of play

This paper presents results from a series of experiments designed to test the impact on subject behavior of changes in the risk dominance and payoff dominance characteristics of two player coordination games. The main finding is that changes in risk dominance significantly affect play of the subjects, whereas changes in the level of payoff dominance do not. Observed history of play also has an important influence on subject behavior, both when subjects are randomly rematched after each game and when they remain matched with the same individual for a sequence of games.     

Markov Perfect Equilibrium: I. Observable Actions

We define Markov strategy and Markov perfect equilibrium (MPE) for games with observable actions. Informally, a Markov strategy depends only on payoff-relevant past events. More precisely, it is measurable with respect to the coarsest partition of histories for which, if all other players use measurable strategies, each player's decision-problem is also measurable. For many games, this definition is equivalent to a simple affine invariance condition. We also show that an MPE is generically robust: if payoffs of a generic game are perturbed, there exists an almost Markovian equilibrium in the perturbed game near the initial MPE. Journal of Economic Literature Classification Numbers: C72, C73.     

Equilibrium selection through incomplete information in coordination games: an experimental study

We perform an experiment on a pure coordination game with uncertainty about the payoffs. Our game is closely related to models that have been used in many macroeconomic and financial applications to solve problems of equilibrium indeterminacy. In our experiment, each subject receives a noisy signal about the true payoffs. This game (inspired by the “global” games of Carlsson and van Damme, Econometrica, 61, 989–1018, 1993) has a unique strategy profile that survives the iterative deletion of strictly dominated strategies (thus a unique Nash equilibrium). The equilibrium outcome coincides, on average, with the risk-dominant equilibrium outcome of the underlying coordination game. In the baseline game, the behavior of the subjects converges to the theoretical prediction after enough experience has been gained. The data (and the comments) suggest that this behavior can be explained by learning. To test this hypothesis, we use a different game with incomplete information, related to a complete information game where learning and prior experiments suggest a different behavior. Indeed, in the second treatment, the behavior did not converge to equilibrium within 50 periods in some of the sessions. We also run both games under complete information. The results are sufficiently similar between complete and incomplete information to suggest that risk-dominance is also an important part of the explanation.      

协调博弈均衡随机稳定性的研究现状与展望

协调博弈均衡选择不仅依赖于博弈支付,更重要的是依赖于参与人之间复杂的行为预期过程,它是继囚徒困境博弈模型之后又一被广泛研究的博弈类型,本文基于随机稳定性思想,结合已有的研究文献,基于实验经济学的研究结果阐述了除博弈支付之外还有许多其他影响均衡稳定性的因素,在此基础上,进一步论述了相关理论模型的研究结论及其存在的缺陷,因此,要更现实地研究协调博弈均衡的稳定性,就需要加入除博弈支付之外的其他因素,最后,文章结合实验研究及理论模型提出了协调博弈均衡稳定性研究的思路与方向,为学者们进一步研究提供了新的视角。     

Recommended play and correlated equilibria: an experimental study

This study reports a laboratory experiment wherein subjects play a hawk–dove game. We try to implement a correlated equilibrium with payoffs outside the convex hull of Nash equilibrium payoffs by privately recommending play. We find that subjects are reluctant to follow certain recommendations. We are able to implement this correlated equilibrium, however, when subjects play against robots that always follow recommendations, including in a control treatment in which human subjects receive the robot “earnings.” This indicates that the lack of mutual knowledge of conjectures, rather than social preferences, explains subjects’ failure to play the suggested correlated equilibrium when facing other human players. We are grateful for financial support provided by the Purdue University Faculty Scholar program and the Asociación Méxicana de Cultura, as well as for the valuable research assistance provided by Shakun Datta and Marikah Mancini. We received helpful comments from Shurojit Chatterji, David Cooper, Arthur Schram, Ricard Torres, an anonymous referee, and from conference and seminar participants at Royal Holloway, the University of Amsterdam, Purdue University, the Economic Science Association and the Society for the Advancement of Economic Theory.     

Transient and asymptotic dynamics of reinforcement learning in games

Reinforcement learners tend to repeat actions that led to satisfactory outcomes in the past, and avoid choices that resulted in unsatisfactory experiences. This behavior is one of the most widespread adaptation mechanisms in nature. In this paper we fully characterize the dynamics of one of the best known stochastic models of reinforcement learning [Bush, R., Mosteller, F., 1955. Stochastic Models of Learning. Wiley & Sons, New York] for 2-player 2-strategy games. We also provide some extensions for more general games and for a wider class of learning algorithms. Specifically, it is shown that the transient dynamics of Bush and Mosteller's model can be substantially different from its asymptotic behavior. It is also demonstrated that in general—and in sharp contrast to other reinforcement learning models in the literature—the asymptotic dynamics of Bush and Mosteller's model cannot be approximated using the continuous time limit version of its expected motion.     

On eliciting beliefs in strategic games

Several recent studies in experimental economics have tried to measure beliefs of subjects engaged in strategic games with other subjects. Using data from one such study we conduct an experiment where our experienced subjects observe early rounds of strategy choices from that study and are given monetary incentives to report forecasts of choices in later rounds. We elicit beliefs using three different scoring rules: linear, logarithmic, and quadratic. We compare forecasts across the scoring rules and compare the forecasts of our trained observers to forecasts of the actual players in the original experiment. We find significant differences across scoring rules. The improper linear scoring rule produces forecasts closer to 0 and 1 than the proper rules, and these forecasts are poorly calibrated. The two proper scoring rules induce significantly different distributions of forecasts. We find that forecasts by observers under both proper scoring rules are significantly different from the forecasts of the actual players, in terms of accuracy, calibration, and the distribution of forecasts. We also find evidence for belief convergence among the observers.     

Dissertation abstract: Voting in the laboratory

Almost every week national elections are held somewhere in the world. Many more elections take place at federal and local levels of government. Surely, these are important events to many of us. This thesis aims at providing a better understanding of why and how people vote in elections. Three original modifications of Palfrey and Rosenthal’s (1983) participation game are used to study voter turnout theoretically and experimentally.¹ In the basic game, each voter supports (i.e., prefers) one of two exogenous candidates and privately decides between voting at a cost and abstaining (without costs). The candidate who receives more votes wins the election (ties are broken randomly) and each supporter of this candidate receives an equal reward, independent of whether or not she voted. The first study (published in the American Political Science Review 100, pp. 235–248) analyzes the effects of social embeddedness on turnout, assuming that voters may be influenced by observing the decisions of other voters around them (e.g., a family or working place). Our experimental results show that the social context matters: this information increases turnout by more than 50%. The increase is greater when neighbors support the same candidate rather than when they support opponents. The second study investigates the effects of public opinion polls on voter turnout and welfare. Poll releases resolve uncertainty about the level of support for each candidate caused by `floating’ voters, whose preferences change across elections. This information increases turnout in the laboratory by 28–34%, depending on the fraction of floating voters in the electorate. If polls indicate equal levels of support for both candidates—in which case aggregate benefits for society are not affected by the outcome—welfare decreases substantially due to costs from excessive turnout. In the final study, elections are preceded by the competition between two candidates: they simultaneously announce binding policy offers in which some voters can be favored at the expense of others through inclusion and exclusion in budget expenditure (Myerson 1993).² We observe that policy offers include 33% more voters—yielding a smaller budget share for each—when voting is compulsory rather than voluntary. Moreover, we find evidence of political bonds between voters and long-lived parties. Overall, in all three experiments many subjects strongly react to economic incentives (i.e., benefits, costs, and informational clues), often in line with what is observed outside of the laboratory. JEL Classification C72, C92, D72 Dissertation Committee: Arthur Schram, University of Amsterdam (advisor) Axel Ockenfels, University of Cologne Thomas Palfrey, California Institute of Technology Cees van der Eijk, University of Nottingham Frans van Winden, University of Amsterdam ¹Palfrey, T.R., & Rosenthal, H. (1983). A strategic calculus of voting. Public Choice, 41, 7–53. ²Myerson, R.B. (1993). Incentives to cultivate favored minorities under alternative electoral systems. American Political Science Review, 87, 856–869.     

Behavior in one-shot traveler’s dilemma games: model and experiments with advice

We report results of one-shot traveler’s dilemma game experiments to test the predictions of a model of introspection. The model describes a noisy out-of-equilibrium process by which players reach a decision of what to do in one-shot games. To test the robustness of the model and to compare it to other models of introspection without noise, we introduce non-binding advice. Advice has the effect of coordinating all players’ beliefs onto a common strategy. Experimentally, advice is implemented by asking subjects who participated in a repeated traveler’s dilemma game to recommend an action to subjects playing one-shot games with identical parameters. In contrast to observations, models based on best-response dynamics would predict lower claims than the advised. We show that our model’s predictions with and without advice are consistent with the data.      

政治理论教学中游戏式教学探析

在新世纪,政治理论教学作为院校思想政治建设的重要环节,缺乏吸引力,教学效果不佳。游戏式教学法能够营造出良好的教学氛围,激发学生强烈的求知欲和兴趣。在政治理论课上采用游戏教学法,可以提高政治理论课的教学质量,增强教学效果。     

Learning to communicate in cheap-talk games

We study learning in communication games. Our main finding is that a simple forward-looking learning rule leads to communication in a large class of games. This class is characterized by a partial-common-interest condition. In contrast, we show that a variety of purely backward looking dynamics may fail to guarantee communication. Memory is a partial substitute for looking forward: With long memory, backward-looking learning leads to communication in a class of games with perfect incentive alignment.