| # | Market | Book A | Book B | Book C | Combined | Signal | Stake split |
|---|
What you're looking at
- Implied probability
- Every decimal odds number is a probability in disguise:
p = 1 / odds. Odds of 2.00 imply a 50% chance. Odds of 1.25 imply 80%. Lower odds mean the book thinks it's more likely. - Arbitrage
- Take the best available price for each side of a market, even if they come from different books, and add up the implied probabilities. A single honest book always prices this over 100% (that gap is their fee). When three books disagree enough, the best-of-three can slip under 100%. Split a bankroll across every side in proportion to each price and you lock in the same profit no matter who wins.
- +EV
- No guarantee, just an edge. If our model's fair probability for an outcome is higher than what a book's price implies, that single bet has positive expected value over the long run, even though any one bet can still lose.
Odds are generated by a random walk around a made-up "fair" probability for each market, with each book carrying its own drifting bias and margin. It's built to occasionally cross into arb and +EV territory, the way real disagreeing books do. No real markets, teams, or books are represented.