Guru
Updated
House edge is the percentage of each bet the casino expects to keep over the long run. For blackjack played with perfect basic strategy under good rules it is around 0.5 percent, the lowest of any standard table game. Played by feel it is closer to 2 percent, and at a 6:5 table it is 2 percent before you make a single mistake.
This guide explains where the number comes from, what moves it, and how to convert it into a realistic expected cost per hour so you can decide whether a session is worth its price.
The player and dealer bust at similar rates, around 28 percent each when both play to 17. The casino's advantage comes from order: you act first, and if you bust you lose immediately even if the dealer would have busted too. Both busting happens about 8 percent of the time, and the house wins all of those.
The player gets several things back: a 3 to 2 payout on naturals (worth about 2.3 percent), the choice to stand on stiff totals while the dealer must hit (about 3.5 percent), doubling (about 1.6 percent) and splitting (about 0.4 percent). Add it all up and the dealer's 8 percent edge from acting last shrinks to roughly 0.5 percent.
Four decks, dealer stands on soft 17, 3:2 blackjack, double on any two cards, double after split, no surrender: the house edge with perfect basic strategy is about 0.4 to 0.45 percent. The perfect book edge page on this site runs a simulation of exactly this game so you can see the number emerge from actual hands.
Every rule change moves it. H17 adds 0.2 percent, no DAS adds 0.14 percent, eight decks instead of four adds about 0.06 percent, and 6:5 adds 1.4 percent. Late surrender subtracts 0.08 percent and re-splitting aces subtracts about 0.07 percent.
Each incorrect decision costs a specific amount of expected value in that spot, and the house edge you actually face is the base edge plus the average of those costs weighted by how often you make them. Standing on 16 against a 10 costs about 4 cents per dollar bet each time. Never doubling costs about 1.6 percent of total action. Taking insurance costs about 7 percent of every insurance bet.
A player with a handful of habitual errors typically plays at 1.5 to 2 percent. The trainer tracks exactly which decisions you miss and how much each cost, so you can see your personal house edge shrink toward the floor as you fix them.
Expected loss per hour equals house edge times average bet times hands per hour. At 0.5 percent, $25 a hand and 70 hands an hour (a fairly full table) the expected cost is about $8.75 an hour. The same player at a 6:5 table faces about 1.9 percent, roughly $33 an hour.
The actual result in any one hour will be far from the expectation; a standard deviation for a flat-betting player is about 1.15 bets per hand, so a 70-hand hour at $25 swings roughly plus or minus $240 two thirds of the time. The expectation is what you pay on average; the variance is what you feel.
| Player | Approx House Edge | Expected Cost per Hour ($25, 70 hands) |
|---|---|---|
| Perfect basic strategy | 0.4-0.5% | About $8 |
| Basic strategy with a few leaks | 0.8-1.0% | About $16 |
| Typical intuitive player | 1.5-2.0% | About $30 |
| Perfect strategy, 6:5 table | 1.8-1.9% | About $33 |
| Mimics the dealer (hits to 17, never doubles or splits) | About 5.5% | About $96 |
No. Basic strategy reduces the house edge to its minimum, around 0.5 percent, but it stays positive for the house. Only card counting or other advantage techniques can make the player's expectation positive.
Return to player is 100 percent minus the house edge. A 0.5 percent house edge is a 99.5 percent RTP. Both describe the long-run average, not any individual session.
Variance. A session of a few hundred hands is dominated by luck; the house edge only becomes visible over tens of thousands of hands. Losing or winning 20 bets in an hour is normal at any edge.
Reading strategy helps, but repetition is what locks it in. Use the trainer to practice hand-by-hand decisions with immediate EV-focused feedback.