Measuring Player Retention and Monetization Using the Mean Cumulative Function

Abstract

Game analytics supports game development by providing direct quantitative feedback about player experience. Player retention and monetization in particular have become central business statistics in free-to-play game development. Many metrics have been used for this purpose. However, game developers often want to perform analytics in a timely manner before all users have churned from the game. This causes data censoring which makes many metrics biased. In this work, we introduce how the Mean Cumulative Function (MCF) can be used to generalize many academic metrics to censored data. The MCF allows us to estimate the expected value of a metric over time, which for example may be the number of game sessions, number of purchases, total playtime and lifetime value. Furthermore, the popular retention rate metric is the derivative of this estimate applied to the expected number of distinct days played. Statistical tools based on the MCF allow game developers to determine whether a given change improves a game, or whether a game is yet good enough for public release. The advantages of this approach are demonstrated on a real in-development free-to-play mobile game, the Hipster Sheep.

Deep Learning Analysis and Age Prediction from Shoeprints

“…For a regression problem of age prediction, we apply cumulative score (CS) and mean cumulative score (MCS) as evaluation metrics to accommodate the nature of the problem. CS and MCS imitate the existing studies [50], [51], [52], and are used to assess accuracies in a range of age groups. CS (or CS j ) and MCS (or M CS − J) give more weight to the smaller ranges of match windows.…”

Micro- and macro-level churn analysis of large-scale mobile games

“…Most existing works [7,8,9,10,11,12,13,14,15,16,17] rely on traditional machine learning models to solve the micro-level churn prediction problem, which suffer from several major limitations as mentioned in Section 1. In view of the recent progress in graph embedding, a promising way is to adopt graph embedding frameworks for churn prediction.…”

“…There have been several previous studies [7,8,9,10,11,12,13,14,15,16, 17] on micro-level mobile game churn prediction by using traditional machine learning models (e.g., logistic regression, random forests, Cox regression). However, we observe several major limitations of these studies.…”

Player retention in league of legends

“…The discovery of patterns in playtime, including the importance of metrics such as inter-session intervals, session lengths, total playtime, etc., matched the increased adoption of Freemium business models in the games industry and introduced the idea of using behavioral telemetry to predict player behavior. This in turn has recently introduced the idea of performing predictive analysis on players behaviour in games [9,17,23,27,35,39], including recently survival analysis [34,36], and the insights that might be gained through this type of investigation. In the literature on these topics, player departure has been termed "churn" and departing players as "churners" [9], adopting terminology from the telecommunications industry.…”