Pub Date : 2017-09-01DOI: 10.1109/TCIAIG.2016.2546063
Ben A. Kybartas, Rafael Bidarra
Computers are often used as tools to design, implement, and even visualize a variety of narrative forms. Many researchers and artists are now further attempting to engage the computer actively throughout the development of the narrative itself. Any form of computational narrative authoring is at some level always mixed-initiative , meaning that the processing capabilities of the computer are utilized with a varying degree to automate certain features of the authoring process. We structure this survey by focusing on two key components of stories, plot and space, and more specifically the degree to which these are either automated by the computer or authored manually. By examining the successes of existing research, we identify potential new research directions in the field of computational narrative. We also identify the advantages of developing a standard model of narrative to allow for collaboration between plot and space automation techniques. This would likely benefit the field of automated space generation with the strengths in the field of automated plot generation.
{"title":"A Survey on Story Generation Techniques for Authoring Computational Narratives","authors":"Ben A. Kybartas, Rafael Bidarra","doi":"10.1109/TCIAIG.2016.2546063","DOIUrl":"https://doi.org/10.1109/TCIAIG.2016.2546063","url":null,"abstract":"Computers are often used as tools to design, implement, and even visualize a variety of narrative forms. Many researchers and artists are now further attempting to engage the computer actively throughout the development of the narrative itself. Any form of computational narrative authoring is at some level always mixed-initiative , meaning that the processing capabilities of the computer are utilized with a varying degree to automate certain features of the authoring process. We structure this survey by focusing on two key components of stories, plot and space, and more specifically the degree to which these are either automated by the computer or authored manually. By examining the successes of existing research, we identify potential new research directions in the field of computational narrative. We also identify the advantages of developing a standard model of narrative to allow for collaboration between plot and space automation techniques. This would likely benefit the field of automated space generation with the strengths in the field of automated plot generation.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"239-253"},"PeriodicalIF":0.0,"publicationDate":"2017-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2016.2546063","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49060034","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-09-01DOI: 10.1109/TCIAIG.2016.2543661
Miguel Nicolau, Diego Pérez-Liébana, M. O’Neill, A. Brabazon
Computer games are highly dynamic environments, where players are faced with a multitude of potentially unseen scenarios. In this paper, AI controllers are applied to the Mario AI benchmark platform, by using the grammatical evolution system to evolve behavior tree structures. These controllers are either evolved to both deal with navigation and reactiveness to elements of the game or used in conjunction with a dynamic A* approach. The results obtained highlight the applicability of behavior trees as representations for evolutionary computation and their flexibility for incorporation of diverse algorithms to deal with specific aspects of bot control in game environments.
{"title":"Evolutionary Behavior Tree Approaches for Navigating Platform Games","authors":"Miguel Nicolau, Diego Pérez-Liébana, M. O’Neill, A. Brabazon","doi":"10.1109/TCIAIG.2016.2543661","DOIUrl":"https://doi.org/10.1109/TCIAIG.2016.2543661","url":null,"abstract":"Computer games are highly dynamic environments, where players are faced with a multitude of potentially unseen scenarios. In this paper, AI controllers are applied to the Mario AI benchmark platform, by using the grammatical evolution system to evolve behavior tree structures. These controllers are either evolved to both deal with navigation and reactiveness to elements of the game or used in conjunction with a dynamic A* approach. The results obtained highlight the applicability of behavior trees as representations for evolutionary computation and their flexibility for incorporation of diverse algorithms to deal with specific aspects of bot control in game environments.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"227-238"},"PeriodicalIF":0.0,"publicationDate":"2017-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2016.2543661","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47678303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-09-01DOI: 10.1109/TCIAIG.2016.2541168
Linbo Luo, Haiyan Yin, Wentong Cai, J. Zhong, M. Lees
Generating suitable game scenarios that can cater for individual players has become an emerging challenge in procedural content generation. In this paper, we propose a data-driven scenario generation framework for game-based training. An evolutionary scenario generation process is designed with a fitness evaluation methodology that integrates the processes of AI player modeling, simulation and model training based on artificial neural networks. The fitness function for scenario evaluation can be automatically constructed based on the proposed methodology. To further enhance the evaluation of scenarios, we specifically study the impact of the timing of events in a scenario and propose a generic scenario representation model that characterizes individual scenario based on the types and timing of events in the scenario. We present an extensive evaluation of our framework by validating our AI player model, demonstrating the impact of timing of events in a scenario and comparing the effectiveness of our data-driven framework with our previous heuristic-based approach and a random baseline. The results show that it is necessary to consider the timing of events for scenario evaluation and the proposed framework works well in generating scenarios for game-based training.
{"title":"Design and Evaluation of a Data-Driven Scenario Generation Framework for Game-Based Training","authors":"Linbo Luo, Haiyan Yin, Wentong Cai, J. Zhong, M. Lees","doi":"10.1109/TCIAIG.2016.2541168","DOIUrl":"https://doi.org/10.1109/TCIAIG.2016.2541168","url":null,"abstract":"Generating suitable game scenarios that can cater for individual players has become an emerging challenge in procedural content generation. In this paper, we propose a data-driven scenario generation framework for game-based training. An evolutionary scenario generation process is designed with a fitness evaluation methodology that integrates the processes of AI player modeling, simulation and model training based on artificial neural networks. The fitness function for scenario evaluation can be automatically constructed based on the proposed methodology. To further enhance the evaluation of scenarios, we specifically study the impact of the timing of events in a scenario and propose a generic scenario representation model that characterizes individual scenario based on the types and timing of events in the scenario. We present an extensive evaluation of our framework by validating our AI player model, demonstrating the impact of timing of events in a scenario and comparing the effectiveness of our data-driven framework with our previous heuristic-based approach and a random baseline. The results show that it is necessary to consider the timing of events for scenario evaluation and the proposed framework works well in generating scenarios for game-based training.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"213-226"},"PeriodicalIF":0.0,"publicationDate":"2017-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2016.2541168","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43297658","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-09-01DOI: 10.1109/TCIAIG.2016.2560882
E. Lakshika, M. Barlow, A. Easton
Modern video games come with highly realistic graphics enabling the players to interact with visually rich virtual worlds. Realistic (life-like) animation of nonplayer characters (NPCs) in such virtual worlds is particularly important to enhance the gaming experience. Multiagent systems are one effective approach to synthesize life-like behaviors and interactions by codifying simple rules into NPCs (each NPC as an autonomous agent). However, such behaviors generally come at the cost of increasing computational expense and complexity in terms of aspects such as number of rules and parameters. Therefore, the desire for high fidelity (highly realistic) behaviors is often in conflict with the drive for low complexity. Multiobjective evolutionary algorithms provide a sophisticated mechanism to optimize two or more conflicting objectives simultaneously. However, evolutionary computing techniques need an appropriate objective function to drive the exploration in the correct direction. Pairing of evolutionary techniques and multiagent systems is challenging in the classes of problems in which the fitness is evaluated based on human aesthetic judgment rather than on objective forms of measurements. In this study, we present a multiobjective evolutionary framework to evolve low complexity and high fidelity multiagent systems by utilizing a machine learning system trained by bootstrapping human aesthetic judgment. We have gathered empirical data in three problem areas—simulation of conversational group dynamics, sheepdog herding behaviors, and traffic dynamics, and show the effectiveness of our approach in deriving low complexity and high fidelity multiagent systems. Further, we have identified common properties of the Pareto-optimal frontiers in the three problem areas that can ultimately lead to an understanding of a relationship between simulation model complexity and behavior fidelity. This understanding will be useful in deciding which level of behavioral fidelity is required for the characters in video games based on the distance to the camera, importance to the scene, and available computational resources.
{"title":"Understanding the Interplay of Model Complexity and Fidelity in Multiagent Systems via an Evolutionary Framework","authors":"E. Lakshika, M. Barlow, A. Easton","doi":"10.1109/TCIAIG.2016.2560882","DOIUrl":"https://doi.org/10.1109/TCIAIG.2016.2560882","url":null,"abstract":"Modern video games come with highly realistic graphics enabling the players to interact with visually rich virtual worlds. Realistic (life-like) animation of nonplayer characters (NPCs) in such virtual worlds is particularly important to enhance the gaming experience. Multiagent systems are one effective approach to synthesize life-like behaviors and interactions by codifying simple rules into NPCs (each NPC as an autonomous agent). However, such behaviors generally come at the cost of increasing computational expense and complexity in terms of aspects such as number of rules and parameters. Therefore, the desire for high fidelity (highly realistic) behaviors is often in conflict with the drive for low complexity. Multiobjective evolutionary algorithms provide a sophisticated mechanism to optimize two or more conflicting objectives simultaneously. However, evolutionary computing techniques need an appropriate objective function to drive the exploration in the correct direction. Pairing of evolutionary techniques and multiagent systems is challenging in the classes of problems in which the fitness is evaluated based on human aesthetic judgment rather than on objective forms of measurements. In this study, we present a multiobjective evolutionary framework to evolve low complexity and high fidelity multiagent systems by utilizing a machine learning system trained by bootstrapping human aesthetic judgment. We have gathered empirical data in three problem areas—simulation of conversational group dynamics, sheepdog herding behaviors, and traffic dynamics, and show the effectiveness of our approach in deriving low complexity and high fidelity multiagent systems. Further, we have identified common properties of the Pareto-optimal frontiers in the three problem areas that can ultimately lead to an understanding of a relationship between simulation model complexity and behavior fidelity. This understanding will be useful in deciding which level of behavioral fidelity is required for the characters in video games based on the distance to the camera, importance to the scene, and available computational resources.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"277-289"},"PeriodicalIF":0.0,"publicationDate":"2017-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2016.2560882","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48201586","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-09-01DOI: 10.1109/TCIAIG.2016.2592965
R. Cardoso, A. Gomes, M. Freire
In virtual worlds (including computer games), users develop trust relationships from their in-world interactions with others. However, these trust relationships end up not being represented in the data structures (or databases) of such virtual worlds, though they sometimes appear associated with reputation and recommendation systems. In addition, as far as we know, the user is not provided with a personal trust tool to sustain his/her decision-making while he/she interacts with other users in the virtual or game world. In order to come up with a computational formal representation of these personal trust relationships, we need to succeed in converting in-world interactions into reliable sources of trust-related data. In this paper, we develop the required formalisms to gather and represent in-world interactions—which are based on the activity theory—as well as a method to convert in-world interactions into trust networks. In the companion paper, we use these trust networks to produce a computational trust decision based on subjective logic. This solution aims at supporting in-world user (or avatar) decisions about others in the game world.
{"title":"A User Trust System for Online Games—Part I: An Activity Theory Approach for Trust Representation","authors":"R. Cardoso, A. Gomes, M. Freire","doi":"10.1109/TCIAIG.2016.2592965","DOIUrl":"https://doi.org/10.1109/TCIAIG.2016.2592965","url":null,"abstract":"In virtual worlds (including computer games), users develop trust relationships from their in-world interactions with others. However, these trust relationships end up not being represented in the data structures (or databases) of such virtual worlds, though they sometimes appear associated with reputation and recommendation systems. In addition, as far as we know, the user is not provided with a personal trust tool to sustain his/her decision-making while he/she interacts with other users in the virtual or game world. In order to come up with a computational formal representation of these personal trust relationships, we need to succeed in converting in-world interactions into reliable sources of trust-related data. In this paper, we develop the required formalisms to gather and represent in-world interactions—which are based on the activity theory—as well as a method to convert in-world interactions into trust networks. In the companion paper, we use these trust networks to produce a computational trust decision based on subjective logic. This solution aims at supporting in-world user (or avatar) decisions about others in the game world.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"305-320"},"PeriodicalIF":0.0,"publicationDate":"2017-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2016.2592965","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42770923","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-09-01DOI: 10.1109/TCIAIG.2016.2565661
M. Bonyadi, Z. Michalewicz, Samadhi Nallaperuma, F. Neumann
Designing automatic drivers for car racing is an active field of research in the area of robotics and artificial intelligence. A controller called Ahura (a heuristic-based racer) for the open racing car simulator is proposed in this paper. Ahura includes five modules, namely steer controller, speed controller, opponent manager, dynamic adjuster, and stuck handler. These modules have 23 parameters all together that are tuned using an evolutionary strategy for a particular car to ensure fast and safe drive on different tracks. These tuned parameters are further modified by the dynamic adjuster module during the run according to the width, friction, and dangerous zones of the track. The dynamic adjustment enables Ahura to decide on-the-fly based on the current situation; hence, it eliminates the need for prior knowledge about the characteristics of the track. The driving performance of Ahura is compared with other state-of-the-art controllers on 40 tracks when they drive identical cars. Our experiments indicate that Ahura performs significantly better than other controllers in terms of damage and completion time especially on complex tracks (road tracks). Also, experiments show that the overtaking strategy of Ahura is safer and more effective compared to other controllers.
{"title":"Ahura: A Heuristic-Based Racer for the Open Racing Car Simulator","authors":"M. Bonyadi, Z. Michalewicz, Samadhi Nallaperuma, F. Neumann","doi":"10.1109/TCIAIG.2016.2565661","DOIUrl":"https://doi.org/10.1109/TCIAIG.2016.2565661","url":null,"abstract":"Designing automatic drivers for car racing is an active field of research in the area of robotics and artificial intelligence. A controller called Ahura (a heuristic-based racer) for the open racing car simulator is proposed in this paper. Ahura includes five modules, namely steer controller, speed controller, opponent manager, dynamic adjuster, and stuck handler. These modules have 23 parameters all together that are tuned using an evolutionary strategy for a particular car to ensure fast and safe drive on different tracks. These tuned parameters are further modified by the dynamic adjuster module during the run according to the width, friction, and dangerous zones of the track. The dynamic adjustment enables Ahura to decide on-the-fly based on the current situation; hence, it eliminates the need for prior knowledge about the characteristics of the track. The driving performance of Ahura is compared with other state-of-the-art controllers on 40 tracks when they drive identical cars. Our experiments indicate that Ahura performs significantly better than other controllers in terms of damage and completion time especially on complex tracks (road tracks). Also, experiments show that the overtaking strategy of Ahura is safer and more effective compared to other controllers.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"290-304"},"PeriodicalIF":0.0,"publicationDate":"2017-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2016.2565661","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41760776","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-06-01DOI: 10.1109/TCIAIG.2015.2512592
Maite Frutos-Pascual, Begonya Garcia Zapirain
The video game market has become an established and ever-growing global industry. The health of the video and computer games industry, together with the variety of genres and technologies available, means that video game concepts and programmes are being applied in numerous different disciplines. One of these is the field known as serious games. The main goal of this paper is to collect all the relevant articles published during the last decade and create a trend analysis about the use of certain artificial intelligence algorithms related to decision making and learning in the field of serious games. A categorization framework was designed and outlined to classify the 129 papers that met the inclusion criteria. The authors made use of this categorization framework for drawing some conclusions regarding the actual use of intelligent serious games. The authors consider that over recent years enough knowledge has been gathered to create new intelligent serious games to consider not only the final aim but also the technologies and techniques used to provide players with a nearly real experience. However, researchers may need to improve their testing methodology for developed serious games, so as to ensure they meet their final purposes.
{"title":"Review of the Use of AI Techniques in Serious Games: Decision Making and Machine Learning","authors":"Maite Frutos-Pascual, Begonya Garcia Zapirain","doi":"10.1109/TCIAIG.2015.2512592","DOIUrl":"https://doi.org/10.1109/TCIAIG.2015.2512592","url":null,"abstract":"The video game market has become an established and ever-growing global industry. The health of the video and computer games industry, together with the variety of genres and technologies available, means that video game concepts and programmes are being applied in numerous different disciplines. One of these is the field known as serious games. The main goal of this paper is to collect all the relevant articles published during the last decade and create a trend analysis about the use of certain artificial intelligence algorithms related to decision making and learning in the field of serious games. A categorization framework was designed and outlined to classify the 129 papers that met the inclusion criteria. The authors made use of this categorization framework for drawing some conclusions regarding the actual use of intelligent serious games. The authors consider that over recent years enough knowledge has been gathered to create new intelligent serious games to consider not only the final aim but also the technologies and techniques used to provide players with a nearly real experience. However, researchers may need to improve their testing methodology for developed serious games, so as to ensure they meet their final purposes.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"133-152"},"PeriodicalIF":0.0,"publicationDate":"2017-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2015.2512592","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62593390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-06-01DOI: 10.1109/TCIAIG.2016.2523508
Greg Foderaro, Ashleigh Swingler, S. Ferrari
This paper presents a model-based approach for computing real-time optimal decision strategies in the pursuit-evasion game of Ms. Pac-Man. The game of Ms. Pac-Man is an excellent benchmark problem of pursuit-evasion game with multiple, active adversaries that adapt their pursuit policies based on Ms. Pac-Man’s state and decisions. In addition to evading the adversaries, the agent must pursue multiple fixed and moving targets in an obstacle-populated environment. This paper presents a novel approach by which a decision-tree representation of all possible strategies is derived from the maze geometry and the dynamic equations of the adversaries or ghosts. The proposed models of ghost dynamics and decisions are validated through extensive numerical simulations. During the game, the decision tree is updated and used to determine optimal strategies in real time based on state estimates and game predictions obtained iteratively over time. The results show that the artificial player obtained by this approach is able to achieve high game scores, and to handle high game levels in which the characters speeds and maze complexity become challenging even for human players.
{"title":"A Model-Based Approach to Optimizing Ms. Pac-Man Game Strategies in Real Time","authors":"Greg Foderaro, Ashleigh Swingler, S. Ferrari","doi":"10.1109/TCIAIG.2016.2523508","DOIUrl":"https://doi.org/10.1109/TCIAIG.2016.2523508","url":null,"abstract":"This paper presents a model-based approach for computing real-time optimal decision strategies in the pursuit-evasion game of Ms. Pac-Man. The game of Ms. Pac-Man is an excellent benchmark problem of pursuit-evasion game with multiple, active adversaries that adapt their pursuit policies based on Ms. Pac-Man’s state and decisions. In addition to evading the adversaries, the agent must pursue multiple fixed and moving targets in an obstacle-populated environment. This paper presents a novel approach by which a decision-tree representation of all possible strategies is derived from the maze geometry and the dynamic equations of the adversaries or ghosts. The proposed models of ghost dynamics and decisions are validated through extensive numerical simulations. During the game, the decision tree is updated and used to determine optimal strategies in real time based on state estimates and game predictions obtained iteratively over time. The results show that the artificial player obtained by this approach is able to achieve high game scores, and to handle high game levels in which the characters speeds and maze complexity become challenging even for human players.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"153-165"},"PeriodicalIF":0.0,"publicationDate":"2017-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2016.2523508","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47280599","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-06-01DOI: 10.1109/TCIAIG.2015.2508966
H. Kaindl, H. Horacek, A. Scheucher
There is a gap between theory and practice regarding the assessment of minimaxing versus product propagation. The use of minimaxing in real programs for certain two-player games like chess is more or less ubiquitous, due to the substantial search space reductions enabled by several pruning algorithms. In stark contrast, some theoretical work supported the view that product propagation could be a viable alternative, or even superior on theoretical grounds. In fact, these rules have different conceptual problems. While minimaxing treats heuristic values as true values, product propagation interprets them as independent probabilities. So, which is the better rule for backing up heuristic values in game trees, and under which circumstances? We present a systematic analysis and results of simulation studies that compare these backup rules in synthetic trees with properties found in certain real game trees, for a variety of situations with characteristic properties. Our results show yet unobserved complementary strengths in their respective capabilities, depending on the size of node score changes (“quiet” versus “nonquiet” positions), and on the degree of advantage of any player over the opponent. In particular, exhaustive analyses for shallow depths show that product propagation can indeed be better than minimaxing when both approaches search to the same depth, especially for making decisions from a huge amount of alternatives, where deep searches are still prohibitive. However, our results also provide some justification for the more or less ubiquitous use of minimaxing in chess programs, where deep searches prevail and the pruning algorithms available for minimaxing make the difference.
{"title":"Product Propagation: A Backup Rule Better Than Minimaxing?","authors":"H. Kaindl, H. Horacek, A. Scheucher","doi":"10.1109/TCIAIG.2015.2508966","DOIUrl":"https://doi.org/10.1109/TCIAIG.2015.2508966","url":null,"abstract":"There is a gap between theory and practice regarding the assessment of minimaxing versus product propagation. The use of minimaxing in real programs for certain two-player games like chess is more or less ubiquitous, due to the substantial search space reductions enabled by several pruning algorithms. In stark contrast, some theoretical work supported the view that product propagation could be a viable alternative, or even superior on theoretical grounds. In fact, these rules have different conceptual problems. While minimaxing treats heuristic values as true values, product propagation interprets them as independent probabilities. So, which is the better rule for backing up heuristic values in game trees, and under which circumstances? We present a systematic analysis and results of simulation studies that compare these backup rules in synthetic trees with properties found in certain real game trees, for a variety of situations with characteristic properties. Our results show yet unobserved complementary strengths in their respective capabilities, depending on the size of node score changes (“quiet” versus “nonquiet” positions), and on the degree of advantage of any player over the opponent. In particular, exhaustive analyses for shallow depths show that product propagation can indeed be better than minimaxing when both approaches search to the same depth, especially for making decisions from a huge amount of alternatives, where deep searches are still prohibitive. However, our results also provide some justification for the more or less ubiquitous use of minimaxing in chess programs, where deep searches prevail and the pruning algorithms available for minimaxing make the difference.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"109-122"},"PeriodicalIF":0.0,"publicationDate":"2017-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2015.2508966","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44552368","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-06-01DOI: 10.1109/TCIAIG.2016.2520256
Michael Cook, S. Colton, J. Gow
Automatically generating content for videogames has long been a staple of game development and the focus of much successful research. Such forays into content generation usually concern themselves with producing a specific game component, such as a level design. This has proven a rich and challenging area of research, but in focusing on creating separate parts of a larger game, we miss out on the most challenging and interesting aspects of game development. By expanding our scope to the automated design of entire games, we can investigate the relationship between the different creative tasks undertaken in game development, tackle the higher level creative challenges of game design, and ultimately build systems capable of much greater novelty, surprise, and quality in their output. This paper, the first in a series of two, describes two case studies in automating game design, proposing cooperative coevolution as a useful technique to use within systems that automate this process. We show how this technique allows essentially separate content generators to produce content that complements each other. We also describe systems that have used this to design games with subtle emergent effects. After introducing the technique and its technical basis in this paper, in the second paper in the series we discuss higher level issues in automated game design, such as potential overlap with computational creativity and the issue of evaluation.
{"title":"The ANGELINA Videogame Design System—Part I","authors":"Michael Cook, S. Colton, J. Gow","doi":"10.1109/TCIAIG.2016.2520256","DOIUrl":"https://doi.org/10.1109/TCIAIG.2016.2520256","url":null,"abstract":"Automatically generating content for videogames has long been a staple of game development and the focus of much successful research. Such forays into content generation usually concern themselves with producing a specific game component, such as a level design. This has proven a rich and challenging area of research, but in focusing on creating separate parts of a larger game, we miss out on the most challenging and interesting aspects of game development. By expanding our scope to the automated design of entire games, we can investigate the relationship between the different creative tasks undertaken in game development, tackle the higher level creative challenges of game design, and ultimately build systems capable of much greater novelty, surprise, and quality in their output. This paper, the first in a series of two, describes two case studies in automating game design, proposing cooperative coevolution as a useful technique to use within systems that automate this process. We show how this technique allows essentially separate content generators to produce content that complements each other. We also describe systems that have used this to design games with subtle emergent effects. After introducing the technique and its technical basis in this paper, in the second paper in the series we discuss higher level issues in automated game design, such as potential overlap with computational creativity and the issue of evaluation.","PeriodicalId":49192,"journal":{"name":"IEEE Transactions on Computational Intelligence and AI in Games","volume":"9 1","pages":"192-203"},"PeriodicalIF":0.0,"publicationDate":"2017-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/TCIAIG.2016.2520256","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48663047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：