Positive EV Betting
Stop asking who wins. Start asking which price is wrong.
A bet has positive expected value (+EV) when the odds you're offered pay better than the true probability justifies. It is the only betting strategy that profits by construction: make enough +EV bets and variance washes out, leaving the edge. Every other approach is a story bettors tell themselves.
▸ The formula
EV = (P(win) × profit) − (P(lose) × stake)
EV = (0.45 × 1.5u) − (0.55 × 1u) = 0.675u − 0.55u = +0.125u per bet — a 12.5% edge. The EV calculator runs this for any odds and probability.
▸ Step one is always devigging
The probabilities implied by betting odds sum to more than 100% — the excess is the vig. A -110/-110 line implies 52.4% + 52.4% = 104.8%. Removing the vig proportionally gives the market's fair probabilities (50/50 there). Every EV calculation starts from fair probabilities, not raw ones — the no-vig calculator devigs any two- or three-way market instantly.
▸ Where does the true probability come from?
This is the honest hard part. The main approaches, in rising order of effort:
- Sharp-book anchoring. Devig the line at the sharpest available book and treat that as truth; any softer book beating it by more than its own vig is +EV. This is the workhorse method of most +EV bettors because it outsources probability estimation to the most efficient market available.
- Market-consensus anchoring. Devig several books and average. Weaker than a sharp anchor, better than nothing.
- Your own model. The ceiling is highest and so is the failure rate — a model must beat the closing line over hundreds of bets before it deserves real stakes. Judge it by CLV, not by win-loss.
▸ Practical +EV workflow
- Anchor: devig the sharp line to get fair probability.
- Hunt: shop every book for prices beating fair — outliers, slow-moving lines, promos, boosted odds.
- Verify: run the EV formula; bet only clear positives, not rounding errors.
- Size: fractional Kelly on the calculated edge (Kelly calculator) — full Kelly on an overestimated edge is how +EV bettors go broke.
- Audit: track CLV monthly. Positive CLV confirms the workflow; negative CLV means your anchor or execution is broken.
▸ The failure modes
Three things kill +EV bettors: overestimating edges (fix: fractional Kelly and conservative anchors), betting stale numbers (fix: the price you actually get is the only price that counts), and impatience with variance (fix: +EV profits emerge over hundreds of bets — a losing +EV month is normal and changes nothing if CLV holds).
▸ FAQ
What is positive EV betting?
Betting only when the offered odds pay more than the fair probability justifies. Expected value = (win probability × profit) − (lose probability × stake); when that number is positive, the bet profits on average regardless of any single result.
How do you find +EV bets?
The standard method: devig the sharpest available line to get a fair probability, then shop other books for prices that beat it. Odds boosts, slow-moving lines, and off-market numbers at soft books are the usual sources.
Is positive EV betting actually profitable?
Mathematically yes, by definition — if your probability estimates are honest. The practical risks are overestimated edges, moved lines, and book limits on winning players. Tracking closing line value tells you whether your +EV process is real.