Natural strategic ability under imperfect information

Abstract : Strategies in game theory and multi-agent logics are mathematical objects of remarkable combinatorial complexity Recently, the concept of natural strategies has been proposed to model more human-like reasoning about simple plans and their outcomes So far, the theory of such simple strategic play was only considered in scenarios where all the agents have perfect information about the state of the game In this paper, we extend the notion of natural strategies to games with imperfect information We also show that almost all the complexity results for model checking carry over from the perfect to imperfect information setting That is, verification of natural strategies is usually no more complex for agents with uncertainty This tells games of natural strategic ability clearly apart from most results in game theory and multi-agent logics.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02428208
Contributor : Frédéric Davesne <>
Submitted on : Sunday, January 5, 2020 - 2:55:50 PM
Last modification on : Wednesday, January 8, 2020 - 1:34:07 AM

Identifiers

  • HAL Id : hal-02428208, version 1

Citation

Wojciech Jamroga, Vadim Malvone, Aniello Murano. Natural strategic ability under imperfect information. 18th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2019), May 2019, Montreal, Canada. pp.962--970. ⟨hal-02428208⟩

Share

Metrics

Record views

16