Script- and cluster-based UCT for StarCraft

Justesen, Niels; Tillman, Balint; Togelius, Julian; Risi, Sebastian

doi:10.1109/cig.2014.6932900

Cited by 38 publications

(28 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Churchill and Buro's idea was to only consider during search the actions returned by a strategy from a set of expert-designed options, with all the other actions being ignored. Several algorithms were introduced for searching in action-abstracted state spaces (Justesen et al 2014;Wang et al 2016;Lelis 2017;.…”

Section: Introductionmentioning

confidence: 99%

Evolving Action Abstractions for Real-Time Planning in Extensive-Form Games

Mariño

Moraes

Toledo

et al. 2019

AAAI

View full text Add to dashboard Cite

A key challenge for planning systems in real-time multiagent domains is to search in large action spaces to decide an agent’s next action. Previous works showed that handcrafted action abstractions allow planning systems to focus their search on a subset of promising actions. In this paper we show that the problem of generating action abstractions can be cast as a problem of selecting a subset of pure strategies from a pool of options. We model the selection of a subset of pure strategies as a two-player game in which the strategy set of the players is the powerset of the pool of options— we call this game the subset selection game. We then present an evolutionary algorithm for solving such a game. Empirical results on small matches of µRTS show that our evolutionary approach is able to converge to a Nash equilibrium for the subset selection game. Also, results on larger matches show that search algorithms using action abstractions derived by our evolutionary approach are able to substantially outperform all state-of-the-art planning systems tested.

show abstract

Section: Introductionmentioning

confidence: 99%

Evolving Action Abstractions for Real-Time Planning in Extensive-Form Games

Mariño

Moraes

Toledo

et al. 2019

AAAI

View full text Add to dashboard Cite

show abstract

“…Real-Time Strategy (RTS) games in general, and STAR-CRAFT in particular, have emerged as a fruitful testbed for new AI algorithms [3], [4]. One of the most recurrent techniques for tactical decisions are those based on game tree search, like alpha-beta search [5] or MCTS [6]- [9]. Game tree search algorithms require some representation of the game state, a forward model that gives us the game state resulting from applying an action to another game state, and an evaluation function that assigns reward scores to game states.…”

Section: Introductionmentioning

confidence: 99%

Combat Models for RTS Games

Uriarte

Ontañón

2018

IEEE Trans. Games

View full text Add to dashboard Cite

Game tree search algorithms, such as Monte Carlo Tree Search (MCTS), require access to a forward model (or "simulator") of the game at hand. However, in some games such forward model is not readily available. This paper presents three forward models for two-player attrition games, which we call "combat models", and show how they can be used to simulate combat in RTS games. We also show how these combat models can be learned from replay data. We use STARCRAFT as our application domain. We report experiments comparing our combat models predicting a combat output and their impact when used for tactical decisions during a real game.

show abstract

“…The work of Justesen et al [2014] also deals with the script assignment problem. Their algorithm is similar to ours because it also uses a partition of the units to guide its search.…”

Section: Related Workmentioning

confidence: 99%

Stratified Strategy Selection for Unit Control in Real-Time Strategy Games

Lelis

2017

Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence

View full text Add to dashboard Cite

In this paper we introduce Stratified Strategy Selection (SSS), a novel search algorithm for micromanaging units in real-time strategy (RTS) games. SSS uses a type system to partition the player's units into types and assumes that units of the same type must follow the same strategy. SSS searches in the state space induced by the type system to select, from a pool of options, a strategy for each unit. Empirical results on a simulator of an RTS game shows that SSS employing either fixed or adaptive type systems is able to substantially outperform stateof-the-art search-based algorithms in combat scenarios with up to 100 units.

show abstract

Script- and cluster-based UCT for StarCraft

Cited by 38 publications

References 7 publications

Evolving Action Abstractions for Real-Time Planning in Extensive-Form Games

Evolving Action Abstractions for Real-Time Planning in Extensive-Form Games

Combat Models for RTS Games

Stratified Strategy Selection for Unit Control in Real-Time Strategy Games

Contact Info

Product

Resources

About