Portfolio
Esports Analytics Projects
Please click on the images to see more!
Champion Priority Analysis
One of the analytics tools I implemented used an algorithm to predict champion priority based on scrim, match and solo queue data. In combination with blind data, this proves to be a very accurate measure of what enemies are likely to draft. I would love to continue developing more tools like this using the available data.
Solo Queue Tracker
This tracker displays data in an easy-to-view format, allowing coaches and analysts to check what players practice in solo queue and easily scout opponents. On the backend, it’s tied into my main game database, so it was trivial to begin scouting our opponents for the Americas tournament even though our opponents were playing on boot camp accounts and were originally from a different region.
Automated Visualizations - Lvl 1 Scouting, Jungle Maps, Heatmaps
Here are some examples of visualizations that I worked on while with AOE. A lot of things broke during the transition from Bayes to GRID, these map data based tools being one of them, but it wouldn’t be too hard to revisit.
Software Development Projects
SQDBI - Solo Queue DataBase Ingest
Full-stack esports analytics backend using Python, SQLAlchemy, PostgreSQL and FastAPI. I’ve unified data from scrims, matches and solo queue games to allow for efficient querying and storage. I designed and built a REST API for efficient data analysis and retrieval, enabling automated match analysis and reporting and significantly improving efficiency in match preparation.
lol_bayes_auth
I developed a Python package for the Bayes Esports API, simplifying the obtaining and managing authentication tokens. The package applies caching and automatic refresh mechanisms, reducing unnecessary authentication requests. I implemented comprehensive unit tests using pytest, requests-mock, and freeze gun to ensure code reliability and ease of maintenance.
