Io - Basketball Github

GitHub-hosted basketball games are characterized by their simple mechanics, retro aesthetics, and local multiplayer capabilities.

We used the YOLOv3 model, pre-trained on the COCO dataset, to detect players on the court. We fine-tuned the model on a basketball dataset to improve detection accuracy. basketball github io

Example features: Hexbin density maps (hot zones), filter by shot clock time, compare two players side-by-side.
Tech stack: React, D3, Express (for data fetching).

basketball.github.io

If you visit today, here are three immediate wins: Example features: Hexbin density maps (hot zones), filter

index.html — main landing page and navigation
assets/ — images, icons, CSS
js/ — visualization and UI scripts
data/ — CSV/JSON datasets (or links to external large datasets)
docs/ — markdown pages for tutorials, playbooks, and drill descriptions
_config.yml (optional) — for Jekyll-based site generation

Example Project: projectile-hoops.github.io
Features: A slider-controlled model showing launch angle (arc) and initial velocity. As you adjust the sliders, the ball's parabolic arc redraws in real-time.
Takeaway: It teaches that a 45-degree launch angle theoretically provides the largest margin for error—a classic lesson from physics applied to basketball.

The code for this project is available on GitHub: basketball