APPLIED PROBABILITY INSTITUTE // OPEN SCIENCE
Applied Research & Open Data
Reproducible empirical studies, peer-reviewed methodology, and open-source datasets for sports market micro-structure analysis.
EMPIRICAL WHITEPAPER // 10,000 FIXTURES
PEER-GRADE
Overround Decomposition in European Football Markets: An Empirical Analysis of 10,000 Closing Lines
Comprehensive study isolating bookmaker vigorish from closing odds across 5 Premier League seasons using Multiplicative, Additive, and Shin probability algorithms.
SampleSize 10,000 matches
Overround 5.43% avg
BENCHMARK DATASET // 10K CLOSING LINES
RAW CSV
EPL Closing Lines & Fair Odds Dataset (2021–2026)
Complete 10,000-fixture dataset featuring bookmaker closing lines, computed overround percentage, and de-vigged fair probabilities across 20 English Premier League clubs.
Format CSV (974 KB)
Columns 14 Fields
PYTHON WORKBOOK // REPRODUCIBLE CODE
PYTHON 3
Reproducible Analysis Script (Python 3)
Open-source Python pipeline to independently replicate all tables, statistical models, overround distributions, and Shin parameter calculations.
Execution command:
python analysis_notebook.py Live Dataset Sample (First 5 Rows)
Sample fixtures demonstrating raw closing odds and true probability derivations (Multiplicative vs Shin)
| ID | Season | Date | Fixture | Closing Odds (1 / X / 2) | Overround | Mult Fair (H) | Shin True (H) |
|---|---|---|---|---|---|---|---|
| #1 | 2021-22 | 2022-05-21 | Brentford vs Luton Town | 2.02 / 3.97 / 3.40 | 4.11% | 47.55% | 49.50% |
| #2 | 2021-22 | 2021-12-18 | Wolves vs Tottenham | 1.71 / 5.05 / 3.67 | 5.53% | 55.42% | 58.48% |
| #3 | 2021-22 | 2021-12-03 | West Ham vs Newcastle | 1.60 / 3.06 / 9.77 | 5.42% | 59.29% | 62.50% |
| #4 | 2021-22 | 2021-10-12 | Nottm Forest vs Liverpool | 1.40 / 3.98 / 9.77 | 6.79% | 66.89% | 71.43% |
| #5 | 2021-22 | 2022-01-18 | Sheffield Utd vs Bournemouth | 2.80 / 3.85 / 2.38 | 3.71% | 34.44% | 35.71% |