Florian Herber
Supervisor:Prof. Gudrun Klinker, Ph.D.
Advisor:Dipl.-Inf. David A. Plecher, M.A.
Submission Date:14.10.2020


In this thesis a new viewpoint on Serious Game balancing is developed by shedding light to the differentiation between learning and gaming domain. Since the skills of a player can differ significantly in the two domains, it is necessary to treat them individually. Different levels of starting experiences and different skill acquiring rates require dynamic solutions in order to make Serious Games applicable to a broad range of users. Serious Games have the possibility to enhance learning processes around the world. Especially but not only, in situations where formal teaching is not available due to time, money or locations constraints, they can become powerful tools. For them to become personalized learning environments dynamic difficulty adjustment based on the individual treatment of the learning and gaming domain is indispensable. To accomplish this a theoretical model for the Componentwise Serious Game Balance (CSGB) was conceptualized. Additionally, Componentwise Dynamic Difficulty Adjustment (CDDA), a dynamic difficulty adjustment based on this model, was developed to make it applicable to Serious Game design. The CDDA was then implemented in the already existing Serious Game HieroQuest, which is dedicated towards teaching the Middle Egyptian language. During this, another game mode, which makes use of the dynamic properties to create the game world, was created. Possible effects on the learning success and player experience can now be investigated in a short- and long-term user study.

Componentwise Serious Game Balance

The Componentwise Serious Game Balance (CSGB) model allows for the individual consideration of learning and gaming domain within a Serious Game.

Componentwise Dynamic Difficulty Adjustment

The Componentwise Dynamic Difficulty Adjustment (CDDA) makes use of the CSGB model to enhance Serious Games with a dynamic difficulty adjustment, that considers both domains individually. This is achieved by a five step plan:

  1. Identify the dynamically adjustable elements and assign them to one of the domains
  2. Define clear measurements for the player's performance in both domains
  3. Define difficulty levels for each of the elements identified in Step 1
  4. Define intervals in which the difficulties are updated
  5. Define update functions for both domains to change the difficulties according tothe measurements

For details on both concepts please refer to the thesis: