game AI

In this paper we try to determine the effectiveness of different AI techniques used in simple games. This effectiveness is measured by comparing game-play experience to implementation effort. Game-play experience is measured by letting a test panel play with the different kinds of AI techniques after which a questionnaire is filled in and the implementation effort is simply logged. The results showed that the increasing numbers of AI features is valued, but only until a certain level.

In this paper we try to determine the effectiveness of differ-ent AI techniques used in simple games. This effectiveness is measured by comparing game-play experience to implemen-tation effort. Game-play experience is measured by letting a test panel play with the different kinds of AI techniques af-ter which a questionnaire is filled in and the implementation effort is simply logged. The results showed that the increas-ing numbers of AI features is valued, but only until a certain level.

We present a fully procedural alternative to branching dialogue that is influenced by theories from linguistic pragmatics and technical work in the field of dialogue systems. Specifically, this is a dialogue manager that extends the Talk of the Town framework, in which non-player characters (NPCs) develop and propagate subjective knowledge of the gameworld. While previously knowledge exchange in this framework could only be expressed symbolically, such exchanges may now be rendered as naturalistic conversations between characters. The larger conversation engine currently lacks a player interface, so in this paper we demonstrate our dialogue manager through conversations between NPCs. From an evaluation task, we find that our system produces conversations that flow far more naturally than randomly assembled ones. As a design objective, we have endeavored to make this dialogue manager lightweight and agnostic to its particular application in Talk of the Town; it is our hope that interested readers will consider porting its straightforward design to their own game engines.

Real-time strategy games present an environment in which game AI is expected to behave realistically. One feature of realistic behaviour in game AI is the ability to recognise the strategy of the opponent player. This is known as opponent modeling. In this paper, we propose an approach of opponent modeling based on hierarchically structured models. The top-level of the hierarchy can classify the general play style of the opponent. The bottom-level of the hierarchy can classify specific strategies that further define the opponent’s behaviour. Experiments that test the approach are performed in the RTS game Spring. From our results we may conclude that the approach can be successfully used to classify the strategy of an opponent in the Spring game.

Procedural storytelling offers immense promise for games to offer more reactive narrative experiences that feel more deeply tailored to players' decisions. To date, interactive narrative systems have tended toward either a large emergent possibility space with less focus on narrative structure, or toward greater structure with smaller possibility spaces. In this paper, we introduce Lume, a system for procedural narrative generation that combines the best of these two approaches through a novel combinatorial scene architecture in which storylet scenes are comprised of parameterized node-trees. We detail how the system works and discuss how it moves toward creating reactive interactive narratives that are both dynamic and coherent. CCS CONCEPTS • Human-centered computing → Interactive systems and tools; Interaction design theory, concepts and paradigms; Systems and tools for interaction design; Hypertext / hypermedia.

—Multi-thread architectures are the current trends for both PCs (multi-core CPUs and GPUs) and game consoles such as the Microsoft Xbox 360 and Sony Playstation 3. GPUs (Graphics Processing Units) have evolved into extremely powerful and flexible processors, allowing its use for processing different data. This advantage can be used in game development to optimize the game loop. As reported in the literature, GPGPUs have been used in processing some steps of the game loop, while most of the game logic is still processed by the CPU. This proposal differs by presenting an architecture designed to process practically the entire game loop using the GPU. Two test cases, a crowd simulation and a 2D game shooter prototype called GpuWars, are presented to illustrate the proposed architecture.

To adapt game difficulty upon game character’s strength, Dynamic Difficulty Adjustment (DDA) and some other learning strategies have been applied in commercial game designs. However, most of the existing approaches could not ensure diversity in results, and rarely attempted to coordinate content generation and behaviour control together. This paper suggests a solution that is based on multi-level swarm model and ecosystem mechanism, in order to provide a more flexible way of game balance control.

Would it be possible to bring the promise of unlimited re-playability typically reserved for Roguelike games to competitive multiplayer shooters? This paper tries to address this issue by proposing a novel method to dynamically generate maps at run-time almost as soon as players press the Play button, while ensuring the features what players would expect from the genre. The procedures are simple and practically feasible to be employed in actual computer games. In addition, the work experiments the possibility of incorporating asynchronous game-play element into a multiplayer shooter with human imitating bots where the players can let their bot/avatar replace them when they are not around. The algorithms are implemented and evaluated with a playable game. The evaluations prove that playable 3D dynamic maps can be generated in order of seconds using game context data to initialise the parameters of the algorithm. The paper also shows that asynchronous game-play element is a possible feature for consideration in next generation multiplayer shooters.

This paper proposes a motion-gaming AI for health promotion that can adapt to the player's behavior change in an effective manner. Through modeling of the player's behavior and predicting of their counteraction, this AI learns how its actions can induce its opponent player to move. The proposed AI aims at suppressing health risks associated with motion gaming, by improving balancedness in use of body segments, as well as at increasing the level of calories consumption.

Open world games create a refreshing gaming experience by providing a dynamic environment with many interactable game systems and characters. These systems require special game artificial intelligence (AI) agents to create an illusion of an authentic, living world. The hypothesis is, By modeling game systems that are suitable for Multi-Agent Reinforcement Learning and Evolutionary Game Theory principles, it is possible to create game AI agents that can act rationally and adaptive to the dynamic environment. Doing so will hypothetically make open-worlds more compelling and immersive by making players think that they are in a world with intelligent living beings that have already existed long before. The focus will be on creating a single system with adaptive human-like game AI agents first, then gradually introducing multiple game systems and real players. Finally, the results of this research will be used in commercial open-world games and its effect will be observed.

Real-Time Strategy (RTS) games are well-known for their substantially large combinatorial decision and state spaces, responsible for creating significant challenges for search and machine learning techniques. Exploiting domain knowledge to assist in navigating the expansive decision and state spaces could facilitate the emergence of competitive RTS game-playing agents. Usually, domain knowledge can take the form of expert traces or expert-authored scripts. A script encodes a strategy conceived by a human expert and can be used to steer a search algorithm, such as Monte Carlo Tree Search (MCTS), towards high-value states. However, a script is coarse by nature, meaning that it could be subject to exploitation and poor low-level tactical performance. We propose to perceive scripts as a collection of heuristics that can be parameterized and combined to form a wide array of strategies. The parameterized heuristics mold and filter the decision space in favor of a strategy expressed in terms of parameters. The proposed agent, ParaMCTS, implements several common heuristics and uses NaïveMCTS to search the downsized decision space; however, it requires a preceding manual parameterization step. A genetic algorithm is proposed for use in an optimization phase that aims to replace manual tuning and find an optimal set of parameters for use by EvoPMCTS, the evolutionary counterpart of ParaMCTS. Experimentation results using the µRTS testbed show that EvoPMCTS outperforms several state-of-the-art agents across multiple maps of distinct layouts.

The impressive performance of Monte Carlo Tree Search (MCTS) based game-playing agents in high branching-factor domains such as Go, motivated researchers to apply and adapt MCTS to even more challenging domains. Real-time strategy (RTS) games feature a large combinatorial branching factor and a real-time aspect that pose significant challenges to a broad spectrum of AI techniques, including MCTS. Various MCTS enhancements were proposed, such as the combinatorial multi-armed bandit (CMAB) based sampling, state/action abstractions, and machine learning. In this paper, we propose to employ move pruning as a way to improve the performance of MCTS-based agents in the context of RTS games. We describe a class of possibly detrimental player-actions and propose several pruning approaches targeting it. The experimentation results in µRTS indicate that this could be a promising direction.

This paper proposes a motion gaming AI that encourages players to use their body parts in a well-balanced manner while promoting their enjoyment. The proposed AI uses time series forecasting to predict what actions its opponent human player will perform with respect to a candidate action of the AI, from which result it estimates the amount of movement (momentum) to be produced on each part of the body of the human player against its action. The AI finally selects an action with the goal of making the momentum of body parts on each side of the player body equal. In this AI, a Monte-Carlo Tree Search (MCTS) is employed for candidate action selection and is embedded with a dynamic difficulty adjustment (DDA) mechanism for enhancing enjoyment of the game. Our results offer a contingent evidence that an opponent gaming AI can be used to effectively improve the human player's balance, enjoyment, engrossment, personal gratification while playing the game.