Data analysis · Machine learning
Behind the play
Finding the patterns behind the players.
The problem
Aggregate engagement can hide meaningful differences between players. This analysis asks how behavior relates to retention and progression in a mobile puzzle game.
How I approached it
The pipeline combines data quality checks, exploratory analysis, SQL analytics, feature engineering, and K-Means clustering. It examines D1 and D7 retention and compares progression gates across player segments.
What came out of it
The repository documents analysis of 31,331 valid player records and four behavioral clusters, with visualizations and a written report connecting the findings to product questions.
What I took away
The useful output of an analysis is a decision someone can make, supported by transparent assumptions.
Project details: repository and documentation . Personal contributions are based on my résumé and project notes.