After finishing a great game, I always struggled to find what to play next. Game stores show popular titles, not games that actually match my taste.
So I built GamesPeek — a similarity engine that analyzes 45,000+ games across 6 criteria.
How the algorithm works
| Criteria | Weight |
|---|---|
| Genres | 30% |
| Keywords | 25% |
| Themes | 20% |
| Developers | 10% |
| Player perspectives | 8% |
| Game modes | 7% |
Features
- One-click "Find Similar" → ranked recommendations from most to least similar
- Advanced filters: platform, genre, release year, rating
- Side-by-side game comparison
- 45,000+ games database
- Completely free, no registration required
Tech stack
- Python + Django
- PostgreSQL (materialized vectors for fast similarity queries)
- Redis for caching
- IGDB API for game data
- Bootstrap 5 for frontend
Try it
Open source
The entire codebase is available on GitHub: https://github.com/AlexeyZam15/GamesPeek_site
MIT license. Contributions welcome.
Feedback welcome
I'd love to hear your thoughts on the algorithm approach or any suggestions for improvement. What similar game features have you found useful in other projects?
Stop guessing. Start discovering.
Top comments (0)