Optimal Level Design in Video Games

55 Pages Posted: 28 Aug 2021

See all articles by Yifu Li

Yifu Li

University of Science and Technology of China

Christopher Thomas Ryan

University of Chicago - Booth School of Business

Lifei Sheng

University of Houston Clear Lake

Date Written: July 9, 2021

Abstract

An important problem in video game design is how to sequence game elements within a level (or "chunk") of a game. Each element has two critical features: a reward (e.g., earning an item or being able to watch a cinematic) and a degree of difficulty (e.g., how much energy or focus needed to interact with the game element). The latter property is a distinctive feature in video games, which unlike services (like a trip to the spa) or passive entertainment (like sports or movies), often require concerted effort to consume. We study how to sequence game elements to maximize the overall experienced utility of the level, subject to the dynamics of adaptation to rewards and difficulty. Utility from rewards wain as players become accustomed to them while less effort needs to be expended to overcome challenges as players become accustomed to difficulty.

We find that the optimal design depends on the relationship between rewards and difficulty, leading to qualitatively different level designs. For example, when the proportion of reward-to-difficulty is high, the optimal design mimics that of more passive experiences like that studied in Das Gupta et al (2016). By contrast, the optimal designs of games with low reward-to-difficulty ratios resemble work-out routines with "warm-ups" and "cool-downs" book-ending intense activity. Intermediate cases follow the classical "mini-boss, end-boss'' design where difficulty has two peaks. Where the peaks appear depends on the degree to which players are reward-seeking or difficulty averse. This raises a salient distinction between games designed for entertainment (attracting players seeking rewards) versus "serious" games designed for educational and training purposes (attracting players who benefit from an incentive structure to face challenges). In the former case, for example, the optimal difficulty pattern follows an N-shape: it starts out easy, reaches an internal peak, then follows a U-shaped pattern thereafter. The design does not always follow a classical crescendo or U-shaped pattern uncovered in the previous literature because these designs can stress out players. Level designs with multiple peaks of difficulty are ubiquitous in video games. In summary, this paper provides practical guidance to game designers on how to match level design to the relationship between reward and difficulty inherent in their game's mechanics.

Keywords: video games, level design, memory decay, adaptation

Suggested Citation

Li, Yifu and Ryan, Christopher Thomas and Sheng, Lifei, Optimal Level Design in Video Games (July 9, 2021). Available at SSRN: https://ssrn.com/abstract=3883557 or http://dx.doi.org/10.2139/ssrn.3883557

Yifu Li

University of Science and Technology of China ( email )

International Institute of Finance
School of Management
Hefei, 230026
China

Christopher Thomas Ryan

University of Chicago - Booth School of Business ( email )

5807 S. Woodlawn Avenue
Chicago, IL 60637
United States

Lifei Sheng (Contact Author)

University of Houston Clear Lake ( email )

2700 Bay Area Blvd
Houston, TX 77058

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
295
Abstract Views
764
rank
151,237
PlumX Metrics