What it does This is an awesome-list repository that catalogs resources for applying AI and data mining to digital games. It collects links to game datasets (League of Legends matches, Steam reviews, Animal Crossing catalogs), public APIs (Battle.net, IGDB, OpenDota), AI competition platforms (StarCraft AI, VizDoom, microRTS), and academic books on the topic.

The interesting bit The list is unusually practical about where the data lives — it points to Kaggle CSVs, official publisher APIs, and even discontinued competitions like OpenAI’s Retro Contest. Someone building a training pipeline could actually start here instead of hunting through Reddit threads.

Key highlights

• 20+ game APIs including Steam, Riot, Blizzard, and niche sources like Super Mario Maker save parsers
• 30+ AI competition environments from Minecraft (Malmo) to Doom (ViZDoom) to Hanabi
• Curated academic reading list with Russell & Norvig, Yannakakis & Togelius, and game analytics texts
• Datasets span esports, mobile games, board games, and price/review aggregators
• CC-BY-4.0 license; accepts contributions via issues and PRs

Caveats

• The list appears to be link curation only — no code, no unified tooling, no quality ratings on the datasets themselves
• Some entries are just GitHub repos of unknown maintenance status; freshness is unclear

Verdict Worth bookmarking if you’re a researcher or indie dev building game-playing agents and need to find where the data hides. Skip it if you want a framework — this is a map, not a vehicle.