The Omenizer dataset
Omenizer's product is built on a continuously recorded archive of sports-market observations. This page describes what is observed, what is stored, what is published — and how to cite it.
What we observe
- Prices from hundreds of bookmakers and exchanges (800+ tracked bookmaker identities), streamed in real time and snapshotted through the day.
- Dozens of leagues across soccer, basketball, baseball, tennis, hockey, American football and more — moneylines, spreads and handicaps, totals, and player props.
- Sharp reference prices and multi-source consensus for fair-value estimation.
- In-play price series for selected high-interest matches, at one-minute resolution.
What we store per market
- Every observed price with its observation timestamp and the time the book last changed it.
- The fair-probability estimate, its source tier and de-vig method.
- Observed expected value at detection and as it evolves.
- The closing price, realized closing-line value, and how long the edge stayed available.
- The settled result — recorded against the signal whether it won or lost.
- Data-quality flags; flagged records are excluded from published numbers.
What we publish
Public surfaces are drawn from this archive: Edge Stories (per-match market records with charts, settled with results and CLV), free tools whose benchmarks come from settled Omenizer signals, and periodic research reports. The live, ranked, real-time board is a member surface. Selected datasets and chart embeds are available on request for editorial use.
Data principles
- No guessed identity: teams, leagues and bookmakers match through curated mappings, never string similarity.
- Fail closed: unmatchable or suspect records are excluded, not approximated.
- Receipts over claims: published numbers trace to timestamped observations and settled outcomes.
Citing Omenizer data
Preferred forms: "According to Omenizer data…", "Omenizer's closing-line analysis shows…", "Using Omenizer's fair-price estimate…" — with a link to the specific page or chart used. For datasets, embeds or methodology questions, contact us; boilerplate and logos are on the press page.