The Minority Game (MG) is a simple, generalized framework, belonging to the Game Theory field, which represents the collective behaviour of agents in an idealized situation where they have to compete through adaptation for some finite resource. it generalizes the study of how many individuals may reach a collective solution to a problem under adaptation of each one’s expectations about the future. It is assumed that an odd number of players take a decision at each step of the simulation; the agents that take the minority decision win, while the others loose. The Minority Game in its original formulation state that there is no communication among the agents involved in the simulation; the idea in this paper is to introduce in the model a sort of a social network, in order to see how the links among certain agents can change the results of the simulation. A software model is built, in which the user can define the number of the agents involved and the number of links among them; some examples are studied and analyzed in order to find some general rule. Besides, two communication protocols are implemented in the model: the asynchronous one, in which the agents act sequentially. So the first agents which act take their decision, and from then on they reply to the other agents with the new decision taken. The synchronous protocol states that the agents always communicate to the others their original opinion: they broadcast their opinion to all the agents which are linked to them. Finally, after having collected all the opinions of their friends, they reconsider their choice. We examine the difference among the two protocols using the same starting parameters in the simulation.
Minority Game with Communication: an Agent Based Model
REMONDINO, Marco;
2004-01-01
Abstract
The Minority Game (MG) is a simple, generalized framework, belonging to the Game Theory field, which represents the collective behaviour of agents in an idealized situation where they have to compete through adaptation for some finite resource. it generalizes the study of how many individuals may reach a collective solution to a problem under adaptation of each one’s expectations about the future. It is assumed that an odd number of players take a decision at each step of the simulation; the agents that take the minority decision win, while the others loose. The Minority Game in its original formulation state that there is no communication among the agents involved in the simulation; the idea in this paper is to introduce in the model a sort of a social network, in order to see how the links among certain agents can change the results of the simulation. A software model is built, in which the user can define the number of the agents involved and the number of links among them; some examples are studied and analyzed in order to find some general rule. Besides, two communication protocols are implemented in the model: the asynchronous one, in which the agents act sequentially. So the first agents which act take their decision, and from then on they reply to the other agents with the new decision taken. The synchronous protocol states that the agents always communicate to the others their original opinion: they broadcast their opinion to all the agents which are linked to them. Finally, after having collected all the opinions of their friends, they reconsider their choice. We examine the difference among the two protocols using the same starting parameters in the simulation.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
