The goal of the ABL Wargus project is to develop an adaptive AI for real-time strategy games using a reactive planner. Based on an analysis of how skilled human players conceptualize RTS gameplay, we partition the problem
space into domains of competence seen in expert human play.The agent consists of the following components:
- Strategy Manager: Makes high-level strategy decisions
- Production Manager: Determines which units and buildings to produce
- Tactics Manager: Manages small-scale combat
- Income Manager: Controls worker units to harvest resources
- Scouting Manager: Handles reconnaissance
The agent has also been extended with case-based reasoning to learn build orders from replays. This reduces the amount of knowledge that needs to be hardcoded in the agent. The agent architecture is shown below:
- Josh McCoy and Michael Mateas. An integrated agent for playing real-time strategy games. In Proceedings of the 23rd AAAI Conference on Artificial Intelligence (AAAI 2008), Chicago, July 2008.
- Ben Weber and Michael Mateas. Case-Based Reasoning for Build Order in Real-Time Strategy Games, Artificial Intelligence and Interactive Digital Entertainment (AIIDE 2009), October 2009.
- Ben Weber and Michael Mateas. Conceptual Neighborhoods for Retrieval in Case-Based Reasoning, International Conference on Case-Based Reasoning (ICCBR 2009), July 2009.
The agent performing a rush on Garden of War:
The agent getting overrun on the map No Where to Run:
The agent scouting the opponent on Garden of War: