A player with experience in first-person shooter games will in most cases perform better in a new first-person shooter game than one without any experience and can even perform better than a player with some experience in the new game, indicating that it is possible to apply the behavior learned for one game in another game featuring similar concepts to perform well without knowing the details of the latter. Moreover, human players can often effortlessly use the experience acquired from one video game in another of the same genre. Such solutions already exist for certain navigation problems for instance and are used across many video games. These concepts are common to many first-person shooter games and are enough to define effective behavior regardless of the details of their interpretation. All of these problems can be reasoned about on a conceptual level using data such as the rate of fire of a weapon, the current health of the opponent and the location of health packs. Each moment, a player needs to evaluate the situation and switch to the most appropriate weapon, predict where the opponent likely is or is heading and find the best route to get there. For example, in a first-person shooter one-on-one match, players face problems such as weapon selection, opponent position prediction and navigation. These similar challenges then involve common problems for which basic behavior can be defined and applied regardless of the problem instance. On a conceptual level, video games of the same genre typically feature similar challenges based on the same concepts. Genres are used to categorize video games according to the way players interact with them as well as their rules. Although each video game is unique, they can share a number of concepts depending on their genre. This makes it difficult to create thoroughly robust AI because its development is constrained to the scope of an individual game project. On the other hand, AI is usually independently designed for each game. Thus, creating a truly smart and fully autonomous agent for a complex video game can be as challenging as replicating a large part of the complete human intelligence. The richer and more complex a game is, the more skills and abilities it requires. Since video games are designed for human beings, it is only natural that they focus on their cognitive skills and physical abilities. Thus, the scope of discussion is limited to the game aspect in this work.
Draw a cost curve for starcraft ii wings of liberty series#
Conversely, context AI would deal with context-specific tasks such as making a character perform a series of actions to advance the plot or reacting to player choices. This work focuses on game AI, that is, AI which is concerned with solving the problems in the game such as defeating an opponent in combat or navigating in a maze. On the other hand, the context encompasses all the elements that make up the setting in which these problems appear, such as characters and plot. The game includes the elements that define the actual challenges players face and the problems they have to solve, such as rules and objectives. A video game can be considered to have two main aspects, the context and the game. Introductionīecause artificial intelligence (AI) is a broad notion in video games, it is important to start by defining the scope of this work. The approach is illustrated using two video games, Which relies on conceptual views and actions toĭefine basic yet reasonable and robust behavior. To enable the development of conceptual AI Inspired by the human ability to detect analogiesīetween games and apply similar behavior on aĬonceptual level, this paper suggests an approachīased on the use of a unified conceptual framework Genre, but of different genres too, resulting in aĭifficulty to handle the many aspects of a complexĮnvironment independently for each video game. One issue with this approach is that it does notĮfficiently exploit the numerous similarities thatĮxist between video games not only of the same To quickly and efficiently create specificĪI. Tools currently focus on allowing video game developers Specifically designed for each game, video game AI The various aspects of complex environments With modern video games frequently featuring sophisticatedįor smart and comprehensive agents that understand
0 kommentar(er)
