We address the problem of learning and implementation on the Internet. When agents play repeated games in distributed … environments like the Internet, they have very limited {\em a priori} information about the other players and the payoff matrix … the essential properties that constitute ``reasonable'' learning behavior in distributed environments. We then study the …

We study the effect of stochastically delayed communication on common knowledge acquisition (common learning). If … messages do not report dispatch times, communication prevents common learning under general conditions even if common knowledge … is acquired without communication. If messages report dispatch times, communication can destroy common learning under …

equilibrium of the game does not organize the individual data, two individual learning models are studied: one following a … predictive value of such learning models is found to be poor. …

Recently there has been much theoretical and experimental work on learning in games. However, learning usually means … learning about the strategic behavior of opponents rather than learning about the game as such. In contrast, here we report on …

learning processes exhibited by players. …

common prior is induced, while for other distributions behavior is similar under both regimes. We also present a learning … show that this learning model accurately accounts for all main features of the data. -- common prior ; signaling … ; experiment ; learning …

certain modes of behavior due to learning by doing. This paper provides discrete-time adjustment processes for strategic games … beliefs supported by observed play in the recent past, in line with much of the literature on learning. These processes … eventually settle down in the minimal prep sets of Voorneveld (2004, 2005). -- adjustment ; learning ; minimal prep sets …

Observational learning is typically examined when agents have precise information about their position in the sequence … individuals and receive a private signal about the state of the world. We show that social learning is robust to position … uncertainty. Under any sampling rule satisfying a stationarity assumption, learning is complete if signal strength is unbounded …

several well-known learning processes in the minority game with an arbitrary odd number of players. Interestingly, different … learning processes provide considerably different predictions …

We study observational learning among agents with coordination motives. On a discrete time line, communities of agents … observations are perfectly uncorrelated, asymptotic learning becomes achievable. Herding can also be eliminated by making private …