LoLA is a LoL (League of Legends) game data analysis / analytics project. See report.
The data crawling part is based on Riot API and a Python wrapper Cassiopeia (There is a In-Memory cache problem in Cass, refer to here). A SQLite database is designed and used in this project, which remodels and stores game objects for our analysis objectives. The database I/O part involves sqlite3 and pandas.
This part has been well tested with Python 3.5, though in some environments (e.g. Windows cmd) a decode/encode error may occur in print functions due to multi-language issue; you can just comment out all print codes without any influcence on crawling itself. Python 2.X may also run well with a few edits.
We have obtained data of over 220,000 Ranked-SOLO-5x5 matches with details in the North American region, Pre-Season 2016.
We are doing analyses such as:
- Champion Rank
- Champion Clustering
- Champion Recommendation
- Match Prediction
- Cheating Detection
Our results will be uploaded continuously. As we are doing many experiments, code in this part is quite messy now and will be refined later.
If you are interested in this project or have any problem, feel free to participate in.
