About
A reference engine for cricket fans
This is a context engine, not a tipping site. We aggregate what's already known about cricketers — their stats, the people they've played with, the people who got them out, what's been said about them in the press and on podcasts — and present it cleanly so fans can make their own deductions.
The data
- Cricsheet — ball-by-ball data for every IPL match and every international match in the last decade. The skeleton every player and match page is built on.
- Wikidata — birth dates, birthplaces, country of citizenship. Cross-referenced via ESPNcricinfo player IDs.
- Wikipedia — biographical relationships (family, mentors, coaches) extracted via LLM and verified verbatim against the source.
- Google News — recent mentions of players in the cricket press, surfaced as an outbound link to the original article.
- Podcast transcripts — direct quotes from cricketers on cricket podcasts, transcribed via Whisper and extracted via LLM. Each quote has a verbatim source link.
- CricAPI — current and recent matches for the homepage's live section, refreshed every 15 minutes.
The principles
- Every claim has a source. Click through to where it came from.
- We describe; we don't predict. Win probabilities, "expect a low scorer", "favour the chasers" — that's tipping, not journalism. We stay on the journalism side of the line.
- No bookmaker partnerships. Ever. Even if it would pay better.
- The prediction journal is private. Your picks aren't on a leaderboard, aren't sold, aren't part of any incentive scheme. It's a personal log.
Contact
The site's a personal project. There's no support email yet — if something looks wrong (a player misattributed, a quote misextracted, a stat off), the kindest thing you can do is wait a release and it'll probably get fixed.