Machine Learning Analysis: Mobile Game In-App Purchases 📱🎮

Image credit

Introduction

In-game purchases have grown to be one of the largest sources of revenue in modern gaming industry. According to Wikipedia,

"They discovered that free games represented 39% of the total revenue from January, and that the number jumped to 65% by June, helped in part by the fact that over 75% of the 100 top grossing apps are games."

Despite many games being free-to-play, players continue to spend their money on in-app purchases. Understanding how and why players spend money in games is not only relevant for maximizing revenue but also for designing fair and sustainable gaming ecosystems. Insights into spending behavior can inform targeted promotions, balanced gameplay design, and better user experiences.

This analysis explores player spending patterns on in-app purchases, with a focus on the mobile game market. In-app purchases are a major revenue driver for the entire gaming industry and understanding spending patterns is essential to improve monetization strategies.

Objectives

• Understand the structure and contents of the dataset
• Clean and preprocess the data for analysis
• Perform exploratory data analysis to find spending trends
• Develop machine-learning models (Linear Regression and Random Forest) to predict in-app purchase amounts
• Evaluate model performance and accuracy

Questions

Based on an initial study of the dataset, there are several questions we can ask to guide our analysis:

• What is the overall distribution of in-app purchases among players?
• How does spending vary across different countries?
• How does player age relate to average in-app spending?
• What is the age distribution of the player base?
• Which game genres are associated with higher spending?
• Do players on different devices (iOS, Android, etc.) spend differently?
• How does player engagement (number/length of sessions) correlate with spending?
• How are players distributed across spending segments (Minnows, Dolphins, Whales)?
• What is the total revenue from each spending segment?
• How has total revenue evolved over time?
• How well can we predict their in-app purchase amount?

Tools

Python for Data Cleaning, Data Transformation, Data Visualisation and Data Analysis

Data Set

The data set is publicly available on Kaggle.