Capstone Project 2026

Predicting Game Revenue on Steam: What Drives Financial Success Among Top-Selling Titles?

What we're doing: Using Gamalytic Steam analytics data, this project analyzes which game attributes are most associated with estimated revenue across 10,500+ Steam titles. Framed as a BI initiative for game developers, publishers, and platform operators.

Dataset: Gamalytic Steam Analytics (Spring 2026) — includes price, review score, review count, publisher class, average playtime, wishlists, and estimated revenue across 10,572 AAA, AA, and Indie titles.

Target Variable: Estimated Revenue (log-transformed)

Dimensions (one per member):

• Publisher Class: AAA vs. AA vs. Indie

• Price Tier: Under $20 / $20–$40 / $40+

• Review Score Tier: Mixed / Positive / Very Positive / Overwhelmingly Positive

Factors: Price, Review Score, Average Playtime, Publisher Class, Review Count, Wishlists

Methods: Multiple Linear Regression + Random Forest (500 trees), validated with K-Fold Cross Validation (k=5)

Key Limitation: Revenue figures are Gamalytic estimates, not officially reported values. Findings reflect patterns among established titles with 500+ reviews and should be interpreted as market-performance approximations rather than exact figures.