# CricketLogic

> CricketLogic builds cricket software from ball-by-ball data. It makes three things: (1) the **CricketLogic engine** — an open-source Python library that turns Cricsheet ball-by-ball YAML into a queryable DuckDB cricket data warehouse (predefined analytical views, a Cricsheet downloader, player-registry integration, optional DSPy natural-language query agents; `pip install cricketlogic`, MIT); (2) **IPLTracker** (https://www.ipltracker.com) — the IPL companion whose champion/record/team/player pages are computed first-party by this engine from 19 IPL seasons of ball-by-ball data; and (3) **Crichit** (https://www.crichit.xyz) — arcade cricket games built in Unity. This site (cricketlogic.net) is the home of the open-source engine.

## Sister products

- [IPLTracker](https://www.ipltracker.com): IPL champions, records, teams, players, seasons, head-to-head — powered by the CricketLogic engine
- [Crichit](https://www.crichit.xyz): arcade cricket games (Unity)

## Core pages

- [Home](https://www.cricketlogic.net/): What it is, install, quickstart, feature grid, code sample
- [Features](https://www.cricketlogic.net/features/): Ingestion, predefined views, SQL querying, DSPy agents, downloader, player registry
- [Quickstart](https://www.cricketlogic.net/quickstart/): Install → download data → ingest → query (HowTo)
- [Use cases](https://www.cricketlogic.net/use-cases/): Fantasy modelling, player/team analytics, stat sites, research
- [Docs](https://www.cricketlogic.net/docs/): Data model, CLI/Python/HTTP entry points, links to GitHub
- [About](https://www.cricketlogic.net/about/): What CricketLogic is and is not

## Compare

- [CricketLogic vs pandas-only](https://www.cricketlogic.net/compare/vs-pandas/)
- [CricketLogic vs scraping ESPNcricinfo](https://www.cricketlogic.net/compare/vs-scraping-espncricinfo/)
- [CricketLogic vs paid cricket APIs](https://www.cricketlogic.net/compare/vs-paid-cricket-apis/)

## Blog

- [Loading Cricsheet YAML into DuckDB](https://www.cricketlogic.net/blog/loading-cricsheet-yaml-into-duckdb/)
- [Computing batting & bowling stats from ball-by-ball data](https://www.cricketlogic.net/blog/computing-batting-bowling-stats-from-ball-by-ball/)
- [Why DuckDB for cricket analytics](https://www.cricketlogic.net/blog/why-duckdb-for-cricket-analytics/)

## Key facts

- Language: Python (>= 3.11); storage engine: DuckDB
- Install: pip install cricketlogic
- Data source: Cricsheet ball-by-ball YAML/JSON (https://cricsheet.org)
- Core tables: matches, innings, deliveries, players, teams, player_registry, player_names
- Predefined views: batting_performance, bowling_performance, partnerships, match_summary, fantasy_points
- Interfaces: Python API (CricketDB), CLI (cricketlogic), HTTP service (cricketlogic serve)
- Player registry: 17,550+ players, 8,766 name variations, 12 external sources
- License: MIT
- Source: https://github.com/cricketlogic/cricketlogic