Georgia Tech researchers have written an algorithm that forecasts player performance in video games and adjusts the challenge accordingly to help players learn new skills more quickly.
The researchers say the model could be applied to areas outside of gaming and can scale to tens of thousands of users.
The team developed a simple turn-based game, then used participant scores to apply algorithms that predict how others with similar skillsets would perform. The researchers used the collaborative-filtering model often used for product ratings and recommendations to suggest the next challenge for players.
Although games currently use a technique called rubberbanding to adjust game difficulty in a reactionary way, the new algorithm adjusts difficulty based on in-game performance.
The approach also could be used for educational and training applications, for example, with students struggling with math concepts. "Our approach could allow novices to progress slowly and prevent them from abandoning a challenge right away," says Georgia Tech professor Mark Riedl, who helped develop the algorithm. "For those good at certain skills, the game can be tuned to their particular talents to provide the right challenge at the right time."
The researchers say they can forecast in-game player performance with up to 93-percent accuracy.
From Georgia Tech News
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA
No entries found