A positive expected-value bet — a "+EV bet" — is just a market mispriced relative to the sharpest available line. Most of the time on BettingLab that "sharpest line" is Pinnacle, sometimes Circa, occasionally a deep-liquidity exchange like Sporttrade.
What we mean by "fair value"
Every bet has an implied probability you can read straight off the price. American odds of −150 imply a 60% win rate; +150 implies 40%. The market sets those prices to clear flow, not to reflect truth — that's the gap we hunt.
When Pinnacle (or another sharp book) prices the same bet at −135, that's a 57.4% implied probability. Strip Pinnacle's vig — typically 2-3% — and you land on a fair line that reflects the consensus sharp read.
Where the edge actually lives
If a retail book is offering −135 on a bet whose Pinnacle-implied fair line is −150, the gap is your edge. Concretely: Pinnacle thinks this hits 60% of the time, the retail book is letting you take the same side at a price that only requires 57.4% to break even. That 3% delta is the EV.
Why it persists
Retail books don't move on every Pinnacle tick. They move on their flow. A line that's sat unmoved for 30 minutes while Pinnacle drifted 8 cents is a line you can pick off. That's all this is. There's no secret model.
Sizing
Kelly criterion is the textbook answer; fractional Kelly (typically 0.25x) is the practical one. The point of sizing is to survive variance, not maximize per-bet return.