Enjoyable AI: understanding the ideal level of challenge for video game enjoyment (Summer School 2019)

Providing a challenging Artificial Intelligent (AI) opponent is an important aspect of making many video games enjoyable and immersive. Yet how challenging should an AI opponent be? Too weak, and they may bore players. Too strong, and they may frustrate players. And not all players are equally skilled at the same game. By far the most common solution to this problem is so-called Dynamic Difficulty Adjustment: providing an opponent that tailors its challenge to an optimal level for each player.
People have developed many different AI approaches for tailoring challenge levels, but we know very little about what level of challenge individual players actually enjoy most. Some may want more, others a ‘casual’ experience. In this project, we want to develop a method for finding and providing the level of challenge for each player that is most enjoyable. 

This project will involve two major components:

  1. Developing a robust web-based digital version of a popular board game that incorporates a state-of-the-art Dynamic Difficult Adjustment agent
  2. A first experimental trial using the developed game to better understand the relationship between the level of challenge and players’ overall enjoyment.

The successful candidate will have the opportunity to work on cutting edge AI research in DC Labs.

If you are interested in this opportunity, please contact Simon Demediuk or Sebastian Deterding for more information.

Required skills

We are seeking excellent candidates for up to three roles:

Role 1

  • [Essential] Experience of game development e.g. using Unity or HTML5/JavaScript
  • [Essential] Knowledge of programming 

Role 2

  • [Essential] Experience of graphic design and animation in game development

Role 3

  • [Essential] Experience of experimental social sciences research
  • [Essential] Knowledge of behavioural sciences

How to apply

For more details on the summer school application process (including eligibility and funding) please see the overview page here.

Supervisor(s)

Simon Demediuk
Sebastian Deterding

Further reading