Continual Online Evolutionary Planning for In-Game Build Order Adaptation in StarCraft

Niels Orsleff Justesen; Sebastian Risi

doi:10.1145/3071178.3071210

Continual Online Evolutionary Planning for In-Game Build Order Adaptation in StarCraft

Niels Orsleff Justesen, Sebastian Risi

Research output: Conference Article in Proceeding or Book/Report chapter › Article in proceedings › Research › peer-review

Abstract

The real-time strategy game StarCraft has become an important benchmark for AI research as it poses a complex environment with numerous challenges. An important strategic aspect in this game is to decide what buildings and units to produce. StarCraft bots playing in AI competitions today are only able to switch between predefined strategies, which makes it hard to adapt to new situations. This paper introduces an evolutionary-based method to overcome this challenge, called Continual Online Evolutionary Planning (COEP), which is able to perform in-game adaptive build-order planning. COEP was added to an open source StarCraft bot called UAlbertaBot and is able to outperform the built-in bots in the game as well as being competitive against a number of scripted opening strategies. The COEP augmented bot can change its build order dynamically and quickly adapt to the opponent’s strategy.

Original language	English
Title of host publication	GECCO ’17 Proceedings of the Genetic and Evolutionary Computation Conference
Publisher	Association for Computing Machinery
Publication date	16 Jul 2017
Pages	187-194
ISBN (Print)	978-1-4503-4920-8
DOIs	https://doi.org/10.1145/3071178.3071210
Publication status	Published - 16 Jul 2017
Event	The Genetic and Evolutionary Computation Conference - Berlin, Germany, Germany Duration: 15 Jul 2017 → 19 Jul 2017 http://gecco-2017.sigevo.org/index.html/HomePage

Conference

Conference	The Genetic and Evolutionary Computation Conference
Location	Berlin
Country/Territory	Germany
City	Germany
Period	15/07/2017 → 19/07/2017
Internet address	http://gecco-2017.sigevo.org/index.html/HomePage

Keywords

StarCraft
real-time strategy
adaptive AI
evolutionary planning
build-order optimization

Access to Document

10.1145/3071178.3071210

Continual Online Evolutionary Planning for In-Game Build Order Adaptation in StarCraftFinal published version, 6.41 MB

Cite this

@inproceedings{0ea9a6bed10646bd8d1cde77b358f870,

title = "Continual Online Evolutionary Planning for In-Game Build Order Adaptation in StarCraft",

abstract = "The real-time strategy game StarCraft has become an important benchmark for AI research as it poses a complex environment with numerous challenges. An important strategic aspect in this game is to decide what buildings and units to produce. StarCraft bots playing in AI competitions today are only able to switch between predefined strategies, which makes it hard to adapt to new situations. This paper introduces an evolutionary-based method to overcome this challenge, called Continual Online Evolutionary Planning (COEP), which is able to perform in-game adaptive build-order planning. COEP was added to an open source StarCraft bot called UAlbertaBot and is able to outperform the built-in bots in the game as well as being competitive against a number of scripted opening strategies. The COEP augmented bot can change its build order dynamically and quickly adapt to the opponent{\textquoteright}s strategy.",

keywords = "StarCraft, real-time strategy, adaptive AI, evolutionary planning, build-order optimization, StarCraft, real-time strategy, adaptive AI, evolutionary planning, build-order optimization",

author = "Justesen, {Niels Orsleff} and Sebastian Risi",

year = "2017",

month = jul,

day = "16",

doi = "10.1145/3071178.3071210",

language = "English",

isbn = "978-1-4503-4920-8",

pages = "187--194",

booktitle = "GECCO {\textquoteright}17 Proceedings of the Genetic and Evolutionary Computation Conference",

publisher = "Association for Computing Machinery",

address = "United States",

note = "The Genetic and Evolutionary Computation Conference, GECCO ; Conference date: 15-07-2017 Through 19-07-2017",

url = "http://gecco-2017.sigevo.org/index.html/HomePage",

}

Justesen, NO & Risi, S 2017, Continual Online Evolutionary Planning for In-Game Build Order Adaptation in StarCraft. in GECCO ’17 Proceedings of the Genetic and Evolutionary Computation Conference. Association for Computing Machinery, pp. 187-194, The Genetic and Evolutionary Computation Conference, Germany, Germany, 15/07/2017. https://doi.org/10.1145/3071178.3071210

Continual Online Evolutionary Planning for In-Game Build Order Adaptation in StarCraft. / Justesen, Niels Orsleff; Risi, Sebastian.
GECCO ’17 Proceedings of the Genetic and Evolutionary Computation Conference. Association for Computing Machinery, 2017. p. 187-194.

Research output: Conference Article in Proceeding or Book/Report chapter › Article in proceedings › Research › peer-review

TY - GEN

T1 - Continual Online Evolutionary Planning for In-Game Build Order Adaptation in StarCraft

AU - Justesen, Niels Orsleff

AU - Risi, Sebastian

PY - 2017/7/16

Y1 - 2017/7/16

N2 - The real-time strategy game StarCraft has become an important benchmark for AI research as it poses a complex environment with numerous challenges. An important strategic aspect in this game is to decide what buildings and units to produce. StarCraft bots playing in AI competitions today are only able to switch between predefined strategies, which makes it hard to adapt to new situations. This paper introduces an evolutionary-based method to overcome this challenge, called Continual Online Evolutionary Planning (COEP), which is able to perform in-game adaptive build-order planning. COEP was added to an open source StarCraft bot called UAlbertaBot and is able to outperform the built-in bots in the game as well as being competitive against a number of scripted opening strategies. The COEP augmented bot can change its build order dynamically and quickly adapt to the opponent’s strategy.

AB - The real-time strategy game StarCraft has become an important benchmark for AI research as it poses a complex environment with numerous challenges. An important strategic aspect in this game is to decide what buildings and units to produce. StarCraft bots playing in AI competitions today are only able to switch between predefined strategies, which makes it hard to adapt to new situations. This paper introduces an evolutionary-based method to overcome this challenge, called Continual Online Evolutionary Planning (COEP), which is able to perform in-game adaptive build-order planning. COEP was added to an open source StarCraft bot called UAlbertaBot and is able to outperform the built-in bots in the game as well as being competitive against a number of scripted opening strategies. The COEP augmented bot can change its build order dynamically and quickly adapt to the opponent’s strategy.

KW - StarCraft

KW - real-time strategy

KW - adaptive AI

KW - evolutionary planning

KW - build-order optimization

KW - StarCraft

KW - real-time strategy

KW - adaptive AI

KW - evolutionary planning

KW - build-order optimization

U2 - 10.1145/3071178.3071210

DO - 10.1145/3071178.3071210

M3 - Article in proceedings

SN - 978-1-4503-4920-8

SP - 187

EP - 194

BT - GECCO ’17 Proceedings of the Genetic and Evolutionary Computation Conference

PB - Association for Computing Machinery

T2 - The Genetic and Evolutionary Computation Conference

Y2 - 15 July 2017 through 19 July 2017

ER -

Continual Online Evolutionary Planning for In-Game Build Order Adaptation in StarCraft

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this