We study decentralized agent coordination with performance guarantees by developing a primitive for mustering teams of agents of minimum acceptable team sizes for StarCraft using randomization to accurately estimate the

In order to experiment with machine learning and data mining techniques in the domain of Real-Time Strategy games such as StarCraft, a dataset is required that captures the complex detail of the interactions taking place between the players and the game. This paper describes a new extraction process by which game data is extracted both directly from game log (replay) files, and indirectly throu...

We describe a generative Bayesian model of tactical attacks in strategy games, which can be used both to predict attacks and to take tactical decisions. This model is designed to easily integrate and merge information from other (probabilistic) estimations and heuristics. In particular, it handles uncertainty in enemy units’ positions as well as their probable tech tree. We claim that learning,...

A big challenge for creating human-level game AI is building agents capable of operating in imperfect information environments. In real-time strategy games the technological progress of an opponent and locations of enemy units are partially observable. To overcome this limitation, we explore a particle-based approach for estimating the location of enemy units that have been encountered. We repr...

The game Starcraft is one of the most interesting arenas to test new machine learning and computational intelligence techniques; however, StarCraft matches take a long time and creating a good dataset for training can be hard. Besides, analyzing match logs to extract the main characteristics can also be done in many different ways to the point that extracting and processing data itself can take...

In this paper, we apply the idea of testing games by learning interactions with them that cause unwanted behavior of the game to test the competition entries for some of the scenarios of the 2010 StarCraft AI competition. By extending the previously published macro action concept to include macro action sequences for individual game units, by adjusting the concept to the realtime requirements o...

This paper investigates the problem of identifying a Real-Time Strategy game player and predicting what a player will do next based on previous actions in the game. Being able to recognize a player’s playing style and their strategies helps us learn the strengths and weaknesses of a specific player, devise counter-strategies that beat the player, and eventually helps us to build better game AI....

Macro-management is an important problem in StarCraft, which has been studied for a long time. Various datasets together with assorted methods have been proposed in the last few years. But these datasets have some defects for boosting the academic and industrial research: 1) There’re neither standard preprocessing, parsing and feature extraction procedures nor predefined training, validation an...

We utilize the distribution of early building decisions on Starcraft maps to predict the winning percentages in each racial matchup. Tournament replays from 5661 games on 7 maps were parsed. The win rate on a given map was predicted from a regression on the remaining maps’ build order distributions and logistic-space win rates. The method is only highly effective at predicting the outcomes for ...