Encyclopedia of Computer Graphics and Games

Living Edition
| Editors: Newton Lee

Game Player Modeling

  • Sehar Shahzad FarooqEmail author
  • Kyung-Joong Kim
Living reference work entry
DOI: https://doi.org/10.1007/978-3-319-08234-9_14-1



Game player modeling is the study of computational models to gain an abstracted description of players in games. This description helps to detect, predict, and express the behavior and feelings of players and personalizes games to their preferences.


Game player modeling is the study of computational models to gain an abstracted description of players in games. This description helps to detect, predict, and express the behavior and feelings of players and personalizes games to their preferences. These models can be automatically created using computational and artificial intelligence techniques which are often enhanced based on the theories derived from human interaction with the games (Yannakakis et al. 2013). It offers two major benefits. First, it helps in content customization to cover broader range of players with different skill levels and adapt challenges on the fly in response to the player’s actions (Bakkes et al. 2012). Second, it works as a form of feedback for the game developers and designers so that they may add new innovative features to the games as well as develop new games that advance knowledge, synthesize experience, and escalate the interest of the player (Yannakakis et al. 2013).

The very first instance of research on player modeling was reported in the 1970s where Slagle and Dixon attempted to model the behavior of opponent players in the domain of classical games by assuming the elementary fallibility of the opponent (Slagle and Dixon 1970). Later on, a search method based on knowledge about opponent players (i.e., strengths/weaknesses) was invented in 1993 (Carmel et al. 1993). In 2000, Donkers improved opponent modeling by taking into account the computer player’s uncertainty (Donkers 2003). Afterward, an increasing interest developed in the player modeling of modern video games to raise the entertainment factor (Charles and Black 2004). Recently, player modeling has extrapolated its perspective from opponent modeling to a number of other research topics including player satisfaction (Yannakakis 2008), modeling player’s preferences (Spronck and Teuling 2010), runtime challenge adaptation (Yannakakis et al. 2013), playing style, and learning effective game strategies (Lockett et al. 2007). A comprehensive history of player modeling is given in (Bakkes et al. 2012).

A player model can have three types of inputs: user data, sensory data, and in-game data (Yannakakis et al. 2013; Martinez and Shichuan 2012). User data includes personal assessment and third-person observation. The negligible limitations of user data are non-relevant data assessments, short-time memory, and player’s self-deception (Yannakakis 2012). Sensory data includes data collected from the sensors mounted on the player’s body or in the player’s surroundings. The most common sensor data includes biometrical (Gunes and Piccardi 2006), physiological (Drachen et al. 2010; Martinez et al. 2013), peripheral (Omar and Ali 2011), and nonverbal natural user interface with the games (Amelynck et al. 2012). However, the sensor’s interface with the player faces challenges when it comes to accuracy and performance. In-game data is based on the player’s actions taken within the game to infer performance, skills, strategies, behavior, and game contexts including level completion time, mission failure counts, resource utilization, situation handling, and target achievements (Nachbar 2013; Kim et al. 2012; Weber et al. 2011). The big challenge is to interpret the raw data correctly for high-level player modeling using limited amount of data.

Based on the type of the input data, several learning and data mining approaches are used for player modeling as can be seen in Table 1. The effectiveness of the modeling technique based on user data is calculated using demographic/stereotype approaches (Butler et al. 2010). The major challenge in such models is that they are limited to deal with situations where individuals greatly deviate from the average. The sensory data is correlated to the player’s behavior, emotions, preferences, cognitive, and affective states (Drachen et al. 2009). Physiological signals are correlated to arousal and valance using Plutchik’s emotion wheel and the valence-arousal scale by Russell (1980), facial expressions using continuous, categorical, and active appearance models, speech or psycholinguistic narrations using PERSONAGE, and psychological factors using Big Five model (Lankveld 2013). In-game data features collected during the game play are used to identify or predict the type of the players which can then be further used for personalized component generation or level modifications (Drachen et al. 2009). An overview of input data gathering, modeling approaches, computational analysis, and applications of game player modeling is shown in Fig. 1.
Table 1

Techniques used for game player modeling based on the input data types

Data type



Supervised learning

Unsupervised learning


User data

Supervised learning (shaker et al. 2010)

Probabilistic learning (Togelius et al. 2014)

Rating-based approach (Mandryk et al. 2006)

Neural network (Schmidhuber 2006)

Committee selection strategy (Togelius et al. 2014)

Clustering (Yannakakis et al. 2013)

Active learning (Togelius et al. 2014)

Classification and regression (Yannakakis et al. 2013)

Sensory data

Neuroevolution (Pedersen et al. 2009)


Cognitive appraisal theory (Frome 2007)

Usability theory (Isbister and Schaffer 2008)

Neural network and clustering (Charles and Black 2004)

Belief-desire intention (Ortega et al. 2013)

Facial action coding system (Ekman and Friesen 1978)

In-game data

Neural network (Charles and Black 2004; Pedersen et al. 2009)

Clustering (Drachen et al. 2009)


Supervised learning with labels (Togelius et al. 2014)

Multilayer perceptron (Togelius et al. 2006)

Sequential minimal optimization (Spronck and Teuling 2010)

Fig. 1

Inputtypes, modeling approaches, analysis, and applications of game player modeling

Although a lot of work has been done on the player modeling, several remaining issues need to be addressed. For instance, sensory data-based models lack non-obtrusive data assessment, data reliability, data validity, vigilance recognition, and quick reactivity. User data-based models exhibit low correlation with the data collection time and the particular situation. In-game data-based models are restricted to particular players’ personal interests in game, expert level, mood, enthusiasm, and surrounding environment, making it difficult to generalize for all players. However the generalization problem is resolved by continuously comparing and adjusting procedural personal behavior with human behavior and active player modeling (Holmgard et al. 2014; Togelius et al. 2014). Furthermore, hybrid approaches are used to overcome the issues of individual data-based player models (Arapakis et al. 2009; Kivikangas et al. 2011; Nogueira et al. 2013a, b). Game player modeling is also experimented in some commercial games (e.g., Tomb Raider, Civilization IV, and Left 4 Dead), but there are still some problems of generalization (Drachen et al. 2009; Spronck and Teuling 2010; Ambinder 2011). Even though player modeling can be generalized, there is still a gap between the player’s characteristics within a game and the real world which needs to be bridged in the future research (Holmgard et al. 2014).



  1. Ambinder, M.: Biofeedback in gameplay: how valve measures physiology to enhance gaming experience. In: Proceedings of the Game Developers Conference (2011)Google Scholar
  2. Amelynck, D., Grachten, M., Noorden, L.V., Leman, M.: Toward e-motion-based music retrieval a study of affective gesture recognition. IEEE Trans. Affect. Comput. 3(2), 250–259 (2012)CrossRefGoogle Scholar
  3. Arapakis, I., Konstas, I., Joemon, M. J.: Using facial expressions and peripheral physiological signals as implicit indicators of topical relevance. In: Proceedings of the seventeenth ACM International Conference on Multimedia, pp. 461–470. ACM Press, New York (2009)Google Scholar
  4. Bakkes, S.C., Spronck, P.H., Lankveld, G.V.: Player behavioural modelling for video games. Entertain. Comput. 3(3), 71–79 (2012)CrossRefGoogle Scholar
  5. Butler, S., Demiris, Y.: Using a cognitive architecture for opponent target prediction. In: Proceedings of the Third International Symposium on AI and Games, pp. 55–62. AISB, Leicester (2010)Google Scholar
  6. Carmel, D., Markovitch, S.: Learning models of opponent's strategy in game playing. In: Proceedings of AAAI Fall Symposium on Games Planning and Learning, pp. 140–147, Technion-Israel Institute of Technology, Israel (1993)Google Scholar
  7. Charles, D., Black, M..: Dynamic player modeling: a framework for player-centered digital games. In: Proceedings of the International Conference on Computer Games, Artificial Intelligence, Design and Education, pp. 29–35. Ulster University, Reading (2004)Google Scholar
  8. Donkers, H.H.L.M.: Searching with opponent models. PhD Thesis, Faculty of Humanities and Sciences, Maastricht University, Maastricht (2003)Google Scholar
  9. Drachen, A., Canossa, A., Yannakakis, G. N.: Player modeling using self-organization in Tomb Raider: underworld. In: Proceedings of the IEEE Symposium on Computational Intelligence and Games (CIG), pp. 1–8. IEEE, Milano (2009)Google Scholar
  10. Drachen, A. Nacke, E. L., Yannakakis, G., Pedersen, L.A.; Psychophysiological correlations with gameplay experience dimensions. In: Brain, Body and Bytes Workshop, CHI 2010, Boston (2010)Google Scholar
  11. Ekman, P., Friesen, W. V.: Facial action coding system: a technique for the measurement of facial movement. In: From Appraisal to Emotion: Differences among Unpleasant Feelings, Motivation and Emotion, vol. 183 12, pp. 271–302. Consulting Psychologist Press, Palo Alto (1978)Google Scholar
  12. Ekman, P., Friesen, W.V.: Facial action coding system: a technique for the measurement of facial movement. In: From Appraisal to Emotion: Differences among Unpleasant Feelings, Motivation and Emotion, vol. 12, pp. 271–302 (1978)Google Scholar
  13. Frome, J.: Eight ways videogames generate emotion. In: Proceedings of Digital Game Research Association (DiGRA), pp. 831–835. DIGRA, Tokyo (2007)Google Scholar
  14. Gunes, H., Piccardi, M.: A bimodal face and body gesture database for automatic analysis of human nonverbal affective behavior. In: Proceedings of the Eighteenth International Conference on Pattern Recognition, vol. 1, pp. 1148–1153 (2006)Google Scholar
  15. Holmgard, C., Liapis, A., Togelius, J., Yannakakis, G. N.: Evolving personas for player decision modeling. In: Proceedings of the IEEE Conference on Computational Intelligence and Games (CIG), pp. 1–8. IEEE Dortmund (2014)Google Scholar
  16. Isbister, K., Schaffer, N.: Game usability: advancing the player experience. A theory of fun for game design. CRC Press, Boca Raton (2008)Google Scholar
  17. Kim, K.-J., Seo, J.-H., Park, J.-G., Na, J.-C.: Generalization of TORCS car racing controllers with artificial neural networks and linear regression analysis. Neurocomputing 88, 87–99 (2012)CrossRefGoogle Scholar
  18. Kivikangas, J.M., Ekman, I., Chanel, G., Jarvela, S., Salminen, M., Cowley, B., Henttonen, P., Ravaja, N.: A review of the use of psychophysiological methods in game research. J. Gaming Virtual Worlds 3(3), 181–199 (2011)CrossRefGoogle Scholar
  19. Lankveld, G.V.: Quantifying individual player differences. PhD thesis, Tilburg University (2013)Google Scholar
  20. Lockett, A.J., Chen, C.L., Miikkulainen, R.: Evolving explicit opponent models in game playing. In: Proceedings of the Ninth Annual Conference on Genetic and Evolutionary Computation (GECCO), pp. 2106–2113. ACM, New York (2007)Google Scholar
  21. Mandryk, R.L., Inkpen, K.M., Calvert, T.W.: Using psychophysiological techniques to measure user experience with entertainment technologies. Behav. Inf. Technol. Spec. Issue User Experience 25(2), 141–158 (2006)CrossRefGoogle Scholar
  22. Martinez, A., Shichuan, D.: A model of the perception of facial expressions of emotion by humans: Research overview and perspectives. J. Mach. Learn. Res. 13(1):1589–1608 (2012)Google Scholar
  23. Martinez, H.P., Bengio, Y., Yannakakis, G.N.: Learning deep physiological models of affect. IEEE Comput. Intell. Mag. 8(2), 20–33 (2013)CrossRefGoogle Scholar
  24. Nachbar J.: Learning in games. In: Meyers R. (ed.) Encyclopedia of Complexity and Systems Science: SpringerReference (www.springerreference.com). Springer, Berlin (2013). 2013-04-30 11:57:51 UTC
  25. Nogueira, P.A., Rodrigues, R., Oliveira, E., Nacke, L. E.: A hybrid approach at emotional state detection: merging theoretical models of emotion with data-driven statistical classifiers. In: Proceedings of the IEEE/WIC/ACM International Joint Conference on Web Intelligence (WI) and Intelligent Agent Technologies (IAT), pp. 253–260. IEEE, Atlanta (2013a)Google Scholar
  26. Nogueira, P.A., Rodrigues, R., Oliveira, E.: Real-time psychophysiological emotional state estimation in digital gameplay scenarios. In: Engineering Applications of Neural Networks, pp. 243–252. Springer, Berlin/Heidelberg/New York (2013b)Google Scholar
  27. Omar, A., Ali, N.M.;Measuring flow in gaming platforms. In: Proceedings of the International Conference on semantic Technology and Information Retrieval (STAIR), pp. 302–305. IEEE, Putrajaya (2011)Google Scholar
  28. Ortega, J., Shaker, N., Togelius, J., Yannakakis, G.N.: Imitating human playing styles in Super Mario Bros. Entertain. Comput. 4(2), 93–104 (2013)CrossRefGoogle Scholar
  29. Pedersen, C., Togelius, J., Yannakakis, G.N.: Modeling player experience in super mario bros. In: Proceedings of IEEE Symposium on Computational Intelligence and Games (CIG), pp. 132–139. IEEE, Milano (2009)Google Scholar
  30. Russell, J.A.: A circumplex model of affect. J. Pers. Soc. Psychol. 39(6), 1161–1178 (1980)CrossRefGoogle Scholar
  31. Schmidhuber, J.: Developmental robotics, optimal artificial curiosity, creativity, music, and the fine arts. Connect. Sci. 18, 173–187 (2006)CrossRefGoogle Scholar
  32. Shaker, N., Yannakakis, G.N., Togelius, J.: Towards automatic personalized content generation for platform games. In: Proceedings of Artificial Intelligence and Interactive Digital Entertainment (AIIDE), pp. 63–68. AAAI Press, California (2010)Google Scholar
  33. Slagle, J.R., Dixon, J.K.: Experiments with the M & N tree-searching program. Commun. ACM 13(3), 147–154 (1970)zbMATHCrossRefGoogle Scholar
  34. Spronck, P.H., den Teuling, F.: Player modeling in Civilization IV. In: Proceedings of the Sixth Artificial Intelligence and Interactive Digital Entertainment Conference (AIIDE), pp. 180–185. AAAI Press, California (2010)Google Scholar
  35. Togelius, J., Nardi, R.D., Lucas, S.M.: Making racing fun through player modeling and track evolution. In: Workshop on Adaptive Approaches for Optimizing Player Satisfaction in Computer and Physical Games, pp. 61–71. CogPrints (2006)Google Scholar
  36. Togelius, J., Shaker, N., Yannakakis, G.N.: Active player modelling. In: Proceedings of the Ninth International Conference on Foundations of Digital Games (FDG) (2014)Google Scholar
  37. Weber, B.G., John, M., Mateas, M., Jhala, A.: Modeling player retention in Madden NFL 11. In: Proceedings of the Twenty-Third Innovative Applications of Artificial Intelligence Conference (IAAI) AAAI Press, San Francisco (2011)Google Scholar
  38. Yannakakis, G.N.: How to model and augment player satisfaction: a review. In: Proceedings of the First Workshop on Child, Computer and Interaction (WOCCI) (2008)Google Scholar
  39. Yannakakis, G.N.: Game AI revisited. In: Proceedings of the 9th Conference on Computing Frontiers. ACM (2012)Google Scholar
  40. Yannakakis, G.N., Spronck, P.H., Loiacono, D., Andre, E., Playermodeling, In: Dagstuhl Seminar on Artificial and Computational Intelligence in Games, pp. 45–55. Schloss Dagstuhl, Germany (2013)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  1. 1.Department of Computer Science and EngineeringSejong UniversitySeoulSouth Korea