Learning-by-Doing and Preference Discovery in Video Game Play

50 Pages Posted: 18 Sep 2019 Last revised: 25 Sep 2019

See all articles by Charles Zou

Charles Zou

University of Chicago

Pradeep K. Chintagunta

University of Chicago

Date Written: September 20, 2019


In this paper we investigate how players learn about their preferences for different modes of play and how this influences their engagement in a popular franchise video game. An important aspect of players' session-level data from a specific generation of the game is that they budget their game time between competitive and non-competitive modes of play. Player behavior over time is consistent with learning and initial competition aversion; shares of time spent in competitive modes tend to drop upon adoption of a new game generation and then rise with time played. To study how the firm can increase player engagement, and to understand preferences for competitive and non-competitive play, we formulate and estimate a structural model that nests Bayesian learning within a multiple-discrete continuous framework. This allows us to explain behavior along the extensive (whether to play a mode) and intensive (how long to play that mode) margins for the multiple modes of play. With this model, we find that users can, broadly speaking, be categorized as high types ("hardcores") who tend to be competition-seeking and are more naturally engaged; and low types ("casuals") who are more likely to be competition-averse and who drop out before learning their true preferences for the modes. We consider actions the firm can undertake to improve consumer engagement--in particular, we perform counterfactual analysis on advertising and console switching (by bundling the game with a new console) policies. We find that low types tend to be more responsive to both policies and primarily respond by increasing usage along the extensive margin. Console switching has a significant positive effect on total play, but causes players to substitute away from competitive levels. This is consistent with the learning framework, where players pay a skill or familiarity "cost" when switching consoles. Finally, we discuss the value of engagement to the firm, both qualitatively and with respect to future purchasing (adoption of a new version).

Keywords: Bayesian learning, multiple discrete continuity, video game, advertising

JEL Classification: M31

Suggested Citation

Zou, Charles and Chintagunta, Pradeep K., Learning-by-Doing and Preference Discovery in Video Game Play (September 20, 2019). Available at SSRN: https://ssrn.com/abstract=3450807 or http://dx.doi.org/10.2139/ssrn.3450807

Charles Zou (Contact Author)

University of Chicago ( email )

1101 East 58th Street
Chicago, IL 60637
United States

Pradeep K. Chintagunta

University of Chicago ( email )

5807 S. Woodlawn Avenue
Chicago, IL 60637
United States
773-702-8015 (Phone)
773-702-0458 (Fax)

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

Paper statistics

Abstract Views
PlumX Metrics