An Experimental Approach to Identifying Prominent Factors in Video Game Difficulty

  • James Fraser
  • Michael Katchabaw
  • Robert E. Mercer
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8253)


This paper explores a full factorial analysis methodology to identify game factors with practical significance on the level of difficulty of a game. To evaluate this methodology, we designed an experimental testbed game, based on the classic game Pac-Man. Our experiment decomposes the evaluation of the level of difficulty of the game into a set of response variables, such as the score. Our offline experiment simulates the behaviour of Pac-Man and the ghosts to evaluate each game factor’s impact on a set of response variables. Our analysis highlights factors that significantly contribute to the game play of individual players as well as to general player strategies. This offline evaluation provides a benefit to commercial games as a useful tool for performing tasks such as game balancing, level tuning and identifying playability and usability issues.


Dynamic Difficulty Game Balancing Adaptive Game System 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Andrade, G., Ramalho, G., Santana, H., Corruble, V.: Challenge-sensitive action selection: An application to game balancing. Intelligent Agent Technology, 194–200 (2005)Google Scholar
  2. 2.
    Bateman, C., Boon, R.: 21st Century Game Design. Charles River Media (2006)Google Scholar
  3. 3.
    Fraser, I.J.: Game challenge: A factorial analysis approach. Master’s thesis, University Of Western Ontario (2012),
  4. 4.
    Hunicke, R., Chapman, V.: Ai for dynamic difficulty adjustment in games. In: Challenges in Game Artificial Intelligence AAAI Workshop, pp. 91–96 (2004)Google Scholar
  5. 5.
    Miller, S.: Auto-Dynamic Difficulty. Published in Scott Miller Game Matters Blog (2004),
  6. 6.
    Plaat, A., Schaeffer, J., Pijls, W., Bruin, A.: SSS*= AB+ TT (1995)Google Scholar
  7. 7.
    Reynolds, C.: Flocks, Herds and Schools: A Distributed Behavioral Model. ACM SIGGRAPH Computer Graphics, 25–34 (1987)CrossRefGoogle Scholar
  8. 8.
    Spronck, P., Sprinkhuizen-Kuyper, I., Postma, E.: Difficulty scaling of game ai. Intelligent Games and Simulation, 33–37 (2004)Google Scholar
  9. 9.
    Spronck, P.H.M.: Adaptive game ai. Ph.D. thesis (2005)Google Scholar
  10. 10.
    Szita, I., Lorincz, A.: Learning to Play Using Low-Complexity Rule-Based Policies: Illustrations through Ms. Pac-Man. Artificial Intelligence Research pp. 659–684 (2007)CrossRefGoogle Scholar
  11. 11.
    Togelius, J., De Nardi, R., Lucas, S.: Towards automatic personalised content creation for racing games. Computational Intelligence and Games, 252–259 (2007)Google Scholar
  12. 12.
    Togelius, J., Lucas, S.: Evolving controllers for simulated car racing. Evolutionary Computation, 1906–1913 (2005)Google Scholar
  13. 13.
    Yannakakis, G., Hallam, J.: Evolving opponents for interesting interactive computer games. In: From Animals to Animats, pp. 499–508 (2004)Google Scholar
  14. 14.
    Yannakakis, G., Hallam, J.: A Generic Approach for Generating Interesting Interactive Pac-Man Opponents. In: IEEE Symposium on Computational Intelligence and Games, pp. 94–101 (2005)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2013

Authors and Affiliations

  • James Fraser
    • 1
  • Michael Katchabaw
    • 1
  • Robert E. Mercer
    • 1
  1. 1.The University of Western OntarioLondonCanada

Personalised recommendations