Posted on Nov 23, 2017 | Rating
   
  

Adaptation and Assessment (TwoA) component

Real-time automatic assessment of and adaptation of game difficulty to player expertise.

Short non-technical description: This component enables a real-time automatic adaptation of game difficulty to player's expertise level. The adaptation algorithm makes use of a stealth assessment algorithm that assigns difficulty ratings and expertise ratings to the players and the game modules respectively. The component tracks changes in these ratings allowing assessment of players' learning progress either by players themselves or by instructors.

Technical description:

This assessment process is non-intrusive and does not negatively affect player’s engagement with the game. Having a player with a known skill rating, the TwoA component can recommend a game module with a difficulty level comfortable to the player. Such adaptation provides a nice balance between player’s motivation and game challenge. The component is lightweight with minimal requirements for integrating with a game and minimal impact on the game’s performance. The component is also agnostic to the game content requiring only basic performance metrics with no need for explicit domain knowledge. There are two main components:

  1. The core adaptation and assessment component to be integrated with a game
  2. An exemplar standalone data visualization/analysis component. Manuals on using both components are available within the packages.

The component is available "as is" without warranties or conditions of any kind. The authors of the software are not obliged to provide any kind of technical support to third parties on any issue they may find with the component.

Detailed description:

This component enables a real-time automatic adaptation of game difficulty to player's expertise level. The component implements a variation of the Computer Adaptive Practice algorithm used in Math Garden, a web-based learning environment for school children. The TwoA component provides a portable and highly interoperable implementation of the publicly available version of the algorithm. It also introduces several improvements to the algorithm to meet specific needs of serious games.

The adaptation algorithm makes use of a stealth assessment algorithm that assigns difficulty ratings and expertise ratings to the players and the game modules respectively. The component tracks changes in these ratings allowing assessment of players' learning progress either by players themselves or by instructors. This assessment process is non-intrusive and does not negatively affect player’s engagement with the game. Having a player with a known skill rating, the TwoA component can recommend a game module with a difficulty level comfortable to the player. Such adaptation provides a nice balance between player’s motivation and game challenge. The component is lightweight with minimal requirements for integrating with a game and minimal impact on the game’s performance. The component is also agnostic to the game content requiring only basic performance metrics with no need for explicit domain knowledge. There are two main components:

  1. The core adaptation and assessment component to be integrated with a game
  2. An exemplar standalone data visualization/analysis component. Manuals on using both components are available within the packages.

Date: May 10, 2017

Language: English

Access URL: https://github.com/rageappliedgame/HatAsset

Download: Adaptation-and-Assessment-TwoA-component.zip

difficulty adaptation

assessment

real-time

automatic



Game development environment: Unity

Target platform: Windows

Programming language: C#

Version: 1.2.5

Version notes: Third iteration. Summary of most important changes from the previous version of the TwoA component:

  • Added a second adaptation module that requires only player accuracy. Accuracy can be any value between 0 and 1.
  • Remove dependency on external files. Now it is assumed that the game developer will add scenario and game data programatically instead of storing them in an xml file.
  • Extended API for greater flexibitility of managing player and scenario data.
  • Added methods to request recommended difficulty rating instead of scenario.
  • Added a parameter (K factor) for scaling changes in ratings in reassessment methods.

Development status: Under Development

Commit URL: https://github.com/rageappliedgame/HatAsset

Type: Apache 2.0 (Apache License 2.0)

URL: https://opensource.org/licenses/Apache-2.0

Personal skills
Computers in other domains Computer games
Education Computer-assisted instruction E-learning Interactive learning environments
Component Personalisation

Related Articles



Component
Motivation Assessment Component
TU Graz, Rage project, TU Graz, CSS

Component
Motivation-based Adaptation Component
TU Graz, Rage project, TU Graz, CSS


Component
Competence-based Adaptation Component
TU Graz, Rage project, TU Graz, CSS

Component
Competence Assessment Component
TU Graz, Rage project, TU Graz, CSS


Component
Real-time Facial Emotion Detection Software Component
RAGE Project, Open Universiteit, Kiavash Bahreini, Rage project, Kiavash Bahreini

Component
Player-centric rule-and-pattern-based adaptation asset
Dessislava Vassileva, Rage project, Sofia University

×