Volleyball Analytics Engine

An analytics engine that turns raw match logs into readable performance signals: efficiency by rotation, attack distribution by zone, serve pressure, and opponent tendencies.

Match statistics are usually recorded but rarely interrogated. Coaches and players end up with totals instead of patterns, which makes it hard to prepare for a specific opponent.

A Pandas pipeline that normalizes rally-level data, then a React dashboard that renders it as court heat maps, rotation efficiency, and per-opponent tendency profiles.

Ingest → transform → serve. A pure-Python analysis core with a thin presentation layer keeps the statistics testable in isolation.

Python, Pandas, NumPy, React, Chart.js.

Keeping the analysis layer framework-free meant the same code powers both notebook exploration and the web dashboard without duplication.

Small-sample noise. Rates over few attempts are misleading, so the engine reports confidence context alongside every derived metric.

Roster data is treated as private: no third-party analytics, local-first processing, and no personally identifying fields in exported datasets.

Match preparation moved from raw box scores to zone-level tendency reads.

Data products live or die on how the number is framed, not on how the number is computed.