Scoring FAQ
Learn how Game Ranks calculates popularity scores for 26,638 ranked games.
How does Game Ranks rank games?
We collect a variety of user and critic scores, reviews, and community signal data. This data is compiled daily and processed through our proprietary algorithm to determine a daily popularity score. These are then updated, live, on our ranking pages.
What is the Popularity score?
The Popularity score measures long-term quality and adoption — a game's all-time standing. It combines quality ratings from both players and critics, rewards relevance, and factors in community engagement signals. Scores range from 0 to 100.
How is the Popularity score calculated?
The score has three components:
- Quality ratings are combined from a variety of sources using a weighted Bayesian average, which pulls games with fewer votes toward the mean to prevent small sample sizes from dominating. Games with only one rating source receive a small penalty.
- Age plays a factor. Newer games get a small boost that depreciates over time. Everyone knows The Legend of Zelda: Ocarina of Time is a masterpiece, but it's not necessarily the first thing you'd recommend someone in 2026.
- Engagement is derived from community signals that track how many people are interacting with any specific game. Each signal is log-scaled so diminishing returns apply and outliers don't dominate.
The final score is clamped between 0 and 100. A game needs at least one rating source to receive a Popularity score. This is why not every game in our database has a score.
What data sources power the rankings?
Some examples of metrics we collect include, but are not limited to:
- Player and critic ratings
- Game wishlists
- Web searches and traffic
- Player activity data
- Review sentiments
- Player counts
How often are scores updated?
Scores are fully recalculated every day at 4:00 AM UTC. Collected activity data syncs daily at 12:00 AM UTC. Game metadata is updated in real-time through webhooks as changes happen on IGDB.
Why doesn't a game have a score?
A game needs a minimum amount of data in order to properly calculate a score. If a game does not yet have enough data, it will remain unranked until it does.
What is a Bayesian average?
A Bayesian average adjusts a game's rating based on how many votes it has. Instead of a simple average, it blends the game's actual rating with the global average, weighted by vote count. A game with thousands of votes stays close to its true rating, while a game with only a few votes gets pulled toward the mean. This prevents games with very few ratings from dominating with artificially high or low scores.