Positive vs. Negative Expectation

By the Riskr team · 6 min read

Strip away the lights and the lingo and every bet comes down to one number: its expected value, or EV — the average amount you'd win or lose per bet if you could repeat it forever. A bet is either positive expectation, where the math is on your side, or negative expectation, where it isn't. Casino games live entirely on the negative side, and understanding why is the single most useful idea in gambling.

What expected value actually is

Expected value is just a weighted average. List every outcome, multiply each by its probability, add them up. If the total is above zero, the bet is +EV. If it's below zero, it's −EV. The number doesn't predict any single result — it predicts the long-run average, which is the only thing that survives lots of repetition.

Take a clean coin flip where you win $1 on heads and lose $1 on tails. EV = (0.5 × +$1) + (0.5 × −$1) = $0. That's a fair bet — break-even forever. Now shift it so you win $0.95 on heads but still lose $1 on tails. EV = (0.5 × +$0.95) + (0.5 × −$1) = −$0.025 per dollar. You'll win about half your flips and still bleed two and a half cents on average every time. That tiny gap is what a casino sells.

Casino games are −EV by design

Every game on a casino floor is built with a payout structure slightly worse than the true odds. American roulette is the cleanest example. A single-number bet pays 35:1, which would be a fair price if the wheel had 36 slots. But the wheel has 38 — the 0 and 00 are the house's margin.

Every $1 spun on that bet is worth about 94.7 cents the moment it leaves your hand. The 5.3-cent gap is the house edge, and it's baked into the rules, not the luck. No spin, hot streak, or lucky number changes it.

What a +EV opportunity really looks like

Positive expectation does exist — but almost never at the games the house controls. It shows up where the player can find a price that's wrong:

• Card counting in blackjack. When a lot of high cards remain in the shoe, the odds briefly tilt toward the player, and a counter raises bets only during those windows. The edge is real but thin — often under 1.5% — and casinos shuffle, limit bets, or eject counters to kill it.
• Advantage play. Exploiting a biased wheel, a poorly designed promotion, or a video poker pay table that actually returns over 100% with perfect play. These edges are rare, get patched fast, and demand flawless execution.
• Mispriced markets. In poker you profit from opponents' mistakes, not the house. In a betting market a sharp bettor wins only when the posted odds are wrong more often than the bookmaker's margin costs them — a razor-thin proposition that almost everyone gets wrong.

The pattern is consistent: +EV requires either a skill the operator hasn't priced in or an error someone hasn't fixed yet. It's scarce, fragile, and actively hunted down. If an edge were easy and durable, it wouldn't survive.

Why time is the enemy of a −EV bet

Here's the cruel mechanic. In a −EV game, the more you play, the more certain your loss becomes. Over a handful of bets, variance dominates and you might be well ahead. But as the number of bets grows, the law of large numbers drags your actual results toward the expected average — which is negative. Volume is the house's best friend.

Play roulette for ten spins and anything can happen. Play it for ten thousand and you'll have donated almost exactly 5.3% of everything you wagered. The only winning move in a −EV game, mathematically, is to play less of it. That's the opposite of how the room is designed to make you feel.