Gameplay data plays nicer when divided into cohorts
J. Huang Brown University, Providence, RI, United States
Abstract
Gameplay data is inherently noisy due to many potential environmental factors, but cohort analysis can help extract the signal. Form cohorts to control for different conditions, then compare these cohorts to find differences that may help provide insight about the players. Cohorts based on the install date of the game make it possible to find out whether changes in the software have a long-term effect. Forming cohorts based on some condition of the players, such as skill level, can reveal how these groups are different and why they are different.
Keywords
Gameplay data; Cohort analysis; Cohorts; Retention; Elo rating system; ...
Get Perspectives on Data Science for Software Engineering now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
