This is a study of how people make decisions in dynamic, social environments, achieving a methodological innovation by using online real-time strategy (RTS) games as laboratories for studying human behavior. Investigation into social interactions has been discussed in the context of virtual worlds and online role-playing games, but not in RTS games, nor have researchers yet developed the data-collection techniques or theoretical principles for doing research in this important sector of human-centered computing. In addition to their entertainment value, RTS games have emerged to become virtual platforms that simulate real-world, real-time physics, scenarios, characters, and strategies. Particularly, multi-player online RTS games are providing a new model of human interaction that is in line with decision theory, game theory, planning, learning, and other concepts from research fields such as Computer Science, Artificial Intelligence, Economics, and Behavioral Sciences.

This research will study users' social strategies in RTS games in three major ways: developing a gaming environment, a user study with experiments, and an automated learning approach. The game will present users with a series of missions to be accomplished, and users will receive points for successful completion. In early stages these missions can be accomplished alone and with little effort, but as the player progresses they will require alliances with others. The first phase of user studies will involve a qualitative analysis of user behaviors, identifying specific points in the game when users must make decisions where the social relationships will be important factors. This qualitative analysis will be followed by a quantitative analysis of the players' performance, developing techniques for measuring the payoffs from each action. Once players begin developing strong alliances, a set of experiments will analyze their reasoning when making strategic decisions. This will include measurements of social tie strength, structural social network features, the past history of interactions, and evolutionary simulations of strategies. The final phase of research will involve controlled experiments with users, presenting them with situations where they have to make a decision that requires consideration of the social structure.

This project will make software, test suites, documentation, and teaching materials freely available on the Internet. Decision making is an important problem in artificial intelligence, and effective decision-making is important in all kinds of organizations and real-world applications. This research could provide the theoretical and experimental basis for developing practical algorithms and applications for social decision making, to make it easier for the organizations and the users of such applications in their decision-making process.

National Science Foundation (NSF)
Division of Information and Intelligent Systems (IIS)
Standard Grant (Standard)
Application #
Program Officer
William Bainbridge
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Maryland College Park
College Park
United States
Zip Code