Game Recommendation Engine from Player Behavior

Board Game Arena (BGA), owned by Asmodee, is the world's largest online board gaming platform with 11M+ users and 1,000+ games. Player habits generate massive amounts of behavioral data — repeated plays of the same game, frequent transitions between games — creating implicit links between titles.

User gaming habits on BGA generate natural connections between games, whether through multiple plays of a single title or frequent transitions from one game to another. How to analyze and visualize these connections in order to recommend personalized game journeys based on player profiles?

20GB+ of raw player behavior data from Board Game Arena — including game session logs, user play histories, game-to-game transition patterns, and player profile metadata. Required SQL-based processing due to the massive data volume.

Used SQL extensively to query, filter, and aggregate the 20GB+ dataset into workable structures. Mapped game-to-game transition networks from player sessions. Identified natural game clusters and player archetypes. Built visualization of game connections and designed a recommendation logic based on player behavior profiles.

Delivered a game connection analysis with visual mapping of player journeys and a recommendation framework for personalized game paths based on user profiles.

Directly applicable to BGA's product strategy — improving game discovery, increasing session time, and reducing churn through smarter recommendations. Demonstrated ability to handle large-scale real-world datasets.