Open Datasets & Public Ledgers
Empirical football intelligence, match-by-match settlement ledgers, and multi-season expected goals (xG) metrics. Dual-archived on CERN Zenodo and Figshare with permanent international DOIs under CC-BY 4.0.
Benchmark Files & Downloads
Updated 2026-10-01 · Snapshot 01xgmind_prediction_ledger.csv
Row-for-row transcription of the live settled ledger at /performance/. Covers settled pre-match forecasts with multi-dimensional binary outcomes (1X2 direction, total goals, exact score, HT/FT, Over/Under 2.5). Settled rows are never edited or retrofitted.
premier_league_xg_audit_2025_26.csv
Season-level advanced metrics for all 20 Premier League clubs (2025–26 campaign): xG, xGA, xG differential, possession percentage, big chances created, shots on target, set-piece goal ratios, penalties, and defensive actions. True gaps are left empty as NaN — never zero-filled.
brier_calibration_demo.ipynb
Companion Python notebook that recomputes all published ledger hit rates from the CSV, asserts figures against the live dashboard, and teaches the mathematical decomposition of the Brier score (Reliability, Resolution, and Uncertainty).
Cite this resource
Attribution snippet for data journalists, newsletters, and academic papers
Terms of Use & Citation Guidelines
All datasets and notebooks provided on this hub are distributed under the Creative Commons Attribution 4.0 International (CC-BY 4.0) license.
Attribution Requirement: You are free to share, adapt, and build upon these materials for any purpose, including commercial projects and academic papers, provided you give appropriate credit to XG Mind AI and provide a direct hyperlink back to https://xgmind.com/data/.
For inquiries regarding custom data snapshots, real-time API integrations, or academic collaboration, please contact our quant desk at [email protected].