Non-player character decision-making in computer games

Uludağlı, Çağkan; Oğuz, Kaya

doi:10.1007/s10462-023-10491-7

Cited by 6 publications

(2 citation statements)

References 68 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A typical depth-first algorithm traverses its surroundings and assigns a score to each direction, picking the cell with the highest score to move [11]. This algorithm can find the shortest path from the beginning to the end of the maze [12]. We regard the stop position after the reinforcement learning movement as the starting point, and the NPC departure zone as the endpoint.…”

Section: Improvement Based On Reinforcement Learningmentioning

confidence: 99%

A model for automatic agent generation based on machine learning in simulation games

Wang

2024

ACE

View full text Add to dashboard Cite

With the continuous promotion of Artificial Intelligence (AI) technology in games, researchers are generally willing to pay attention to the effect of AI on game picture quality and man-machine games. However, how to generate a large number of non-player characters (NPC) in a game, especially in a simulation game, is still a problem. Even if some people use machine learning to discover solutions, most answers tend to be limited to copying finite-state machines established by programmers. This paper explores techniques, such as machine learning and maze algorithms, for 2D in-game environments to construct a model that can quickly generate NPCs with 'personality' and adapt to the environment when developing a game. In the process of model construction, the temporal difference algorithm and the depth-first algorithm are used to customize the decision structure for the NPC. At the same time, a simple node-style map generation algorithm is added to restrain NPC. Moreover, an optional neural network model gives the NPC a memory that can be updated. This model successfully realizes the rapid construction of NPC in developing some simple 2D games. In addition, NPCs have a fair amount of intelligence. It can be used during game development or while the game is running and generate a certain number of NPCs in an unfamiliar environment with differences in behavior.

show abstract

Section: Improvement Based On Reinforcement Learningmentioning

confidence: 99%

A model for automatic agent generation based on machine learning in simulation games

Wang

2024

ACE

View full text Add to dashboard Cite

show abstract

“…In modern game design, the speed of development in virtual battles and the intensity of actions [1] offer players a variety of strategies and tactical maneuvers [2]. A key aspect is the development of game mechanics that allow players to manage virtual artillery units, adapt to changing conditions, and execute tasks with minimal losses.…”

Section: Introductionmentioning

confidence: 99%

Development and optimization of simulation models and methods for controlling virtual artillery units in game scenarios

Maksymova,

Boltyonkov,

Maksymov

et al. 2023

HAIT

View full text Add to dashboard Cite

In the realm of modern video game development, special attention is given to the simulation of artillery systems, which play a crucial role in various military-themed games. This research presents a mathematical model for simulating the actions of a virtual artillery system. The model is designed to manage the execution of combat tasks, including targeting destruction with a specified number of shells and incorporating the strategic movement between firing positions to minimize detection and attack by enemy forces in the game. The model presumes that all shots are effective and equates the number of firing positions to the number of shots, with a minimum of one shot per position. The model's dynamics do not allow for returning to previous positions, adding a layer of complexity and realism to the gameplay. Movement simulations between positions are designed along virtual roads of varying quality, enhancing the strategic elements of the game. A method for determining the optimal strategy for the artillery system's actions has been developed, introducing the concept of the current structure of combat task execution. This problem-solving approach falls within the realm of Pareto-oriented tasks or dynamic programming challenges. The computational method of the model is based on a general algorithm, underpinned by specialized additional algorithms. Results from this model demonstrate the feasibility of completing combat tasks effectively, with a maximum of two shots per firing position. The research differentiates between defensive and offensive tactics in gameplay, suggesting that while a strategy involving ten shots per target aligns with defensive gameplay, a strategy with four shots per target aligns with offensive actions. Consequently, the "shot-and-scoot" tactic in an offensive context can be aptly termed as “hid-and-shot”.

show abstract

Improved Non-Player Character (NPC) behavior using evolutionary algorithm—A systematic review

Armanto,

Rosyid,

Muladi

et al. 2025

Entertainment Computing

View full text Add to dashboard Cite

Non-player character decision-making in computer games

Cited by 6 publications

References 68 publications

A model for automatic agent generation based on machine learning in simulation games

A model for automatic agent generation based on machine learning in simulation games

Development and optimization of simulation models and methods for controlling virtual artillery units in game scenarios

Improved Non-Player Character (NPC) behavior using evolutionary algorithm—A systematic review

Contact Info

Product

Resources

About