The fortunes of real-time strategy have risen and fallen numerous times since the 1992 release of "Dune II," and the genre has persisted doggedly through it all. Though Blizzard is no longer releasing new content for "StarCraft II," the beloved RTS title and others like it are still enjoyed by millions of players across the globe. But the worldwide population of video game players has grown exponentially over the last few years, and the popularity of real-time strategy has not expanded accordingly. By adopting a user-friendly attitude while retaining the complexity and technicality that form the genre's core, Frost Giant's new RTS title might be the first step toward bringing real-time strategy to a new generation of gamers and esports fans.

This article reviews the two most recent IEEE Conference on Computational Intelligence and Games (CIG) StarCraft Artificial Intelligence (AI) Competitions organized by the authors; these were the fourth and fifth in a series of annual competitions initiated in 2011. StarCraft AI Competitions have been hosted in conjunction with three different events: the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE), CIG, and Student StarCraft AI Tournament (SSCAIT). The purpose of these competitions is to design bots that are able autonomously and successfully to play the StarCraft game by implementing real-time strategies. Recent results reveal the promising use of AI techniques in creating successful AI entries, but there is room for improvement with respect to the bots' ability to adapt and learn to defeat humans and scripted AI bots. It provides a dynamic environment in which several agents interact to build military units with which to fight against an opponent.

This literature review covers AI techniques used for real-time strategy video games, focusing specifically on StarCraft. It finds that the main areas of current academic research are in tactical and strategic decision-making, plan recognition, and learning, and it outlines the research contributions in each of these areas. The paper then contrasts the use of game AI in academia and industry, finding the academic research heavily focused on creating game-winning agents, while the indus- try aims to maximise player enjoyment. It finds the industry adoption of academic research is low because it is either in- applicable or too time-consuming and risky to implement in a new game, which highlights an area for potential investi- gation: bridging the gap between academia and industry.