Skip to main content

GameTailor

Overview

Many computer games, training scenarios, and educational games pit the player against a succession of increasingly difficult challenges such as combat with computer-controlled enemies, puzzles, or tasks. In this project we explore techniques to procedurally generate computer game and training "missions" that are appropriately tailored to an individual player's level of skill mastery.

We introduce two problems:

  • Challenge tailoring: the problem of matching the difficulty of skill-based events (combats, puzzles, etcs.) over the course of a game to a specific player’s abilities. Challenge tailoring is a generalization of dynamic difficulty adjustment.
  • Challenge contextualization: the problem of constructing game events that set up the conditions for skill-based events that motivate and explain their occurrence to the player. Challenge contextualization may be solved by story or quest generation.

This project combines data-driven player modeling as a means of predicting player performance on future tasks and using optimization techniques to select (and contextualize) skill-based events of the appropriate level of difficulty.

The following is a screenshot of one of our data collection test beds in which players lead a team in spell-based combat against procedurally generated non-player characters. This is part of a larger game engine test bed based roughly on computer role-playing games.

Sponsors

Army Research Lab (ARL)

Publications

2014

Alexander Zook and Mark O. Riedl. Temporal Game Challenge Tailoring. IEEE Transactions on Computational Intelligence and Artificial Intelligence in Games, forthcoming

Mark O. Riedl and R. Michael Young. The Importance of Narrative as an Affective Instructional Strategy. In R. Sottilare, A. Graesser, X. Hu, and B. Goldberg (Eds.) Design Recommendations for Adaptive Intelligent Tutoring Systems: Adaptive Instructional Strategies, volume 2. Army Science Laboratory, 2014.

2012

Alexander Zook and Mark O. Riedl. A Temporal Data-Driven Player Model for Dynamic Difficulty Adjustment. Proceedings of the 8th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, Palo Alto, California, 2012

Alexander Zook, Steven Lee-Urban, Michael Drinkwater, and Mark O. Riedl. Skill-based Mission Generation: A Data-driven Temporal Player Modeling Approach. Proceedings of the 3rd Workshop on Procedural Content Generation in Games, Raleigh, North Carolina, 2012.

Alexander Zook, Stephen Lee-Urban, Mark O. Riedl, Heather K. Holden, Robert A. Sottilare, and Keith W. Brawner. Automated Scenario Generation: Toward Tailored and Optimized Military Training in Virtual Environments. Proceedings of the 7th International Conference on the Foundations of Digital Games, Raleigh, North Carolina, 2012.