How much a bet is really worth: your fair win probability against the price a book is offering.
Expected value is the long-run average profit of a bet, per dollar staked:
EV = p × (decimal odds − 1) − (1 − p)
where p is your estimate of the true win probability. If EV is positive, the price beats your probability and the bet makes money over many repetitions; if negative, the book's price already has you beat.
You make a side 52% to win and a book offers −105 (decimal 1.952). EV per $1 = 0.52 × 0.952 − 0.48 = +0.0152 — a +1.52% edge, or about $1.52 expected profit on a $100 bet. The breakeven probability at −105 is 51.22%, so any true probability above that is +EV.
The hard part isn't the arithmetic — it's the probability. Sharp bettors estimate it from the de-vigged consensus of sharp books rather than gut feel. That's exactly what EdgeFinder automates across 40+ books.
Deep dive: Expected value explained →
Anything reliably positive. Real edges at major sportsbooks are usually 1–5%; claims of consistent 10%+ edges on main markets are almost always probability-estimation error.
The most defensible source is the no-vig (fair) price from sharp market consensus — remove each book's margin and average across books. Model-based probabilities work when the model is graded publicly against results.
No. EV is a long-run average — variance dominates small samples. A +2% edge still loses on many nights; it wins over thousands of bets with disciplined bankroll management.
Strip the bookmaker margin from two- or three-way odds to reveal the fair price.
Check two or three prices for a guaranteed profit and split your stake exactly.
Optimal bet size from your edge and bankroll, with half- and quarter-Kelly presets.