The house edge on standard blackjack — 6-deck, dealer stands on soft 17, 3:2 payout, double after split — sits at roughly 0.5% with basic strategy. That number is the lowest of any casino game, and the reason blackjack rewards the effort of learning strategy. This page explains what the number actually means, how each rule change moves it, and what it means for your bankroll per hour.
Assumes: 6 decks, dealer stands on soft 17 (S17), blackjack pays 3:2, double after split allowed (DAS), player uses perfect basic strategy. Every rule variation changes this number — usually against the player.
"House edge" is one of the most misunderstood concepts in casino gambling. It sounds intuitive — the casino's built-in advantage — but the number that gets quoted (0.5% for blackjack) has specific assumptions baked in, and misapplying it produces expectations that don't match reality. Some players think 0.5% means "I lose 50 cents for every $100 I bet" — which is technically true only over infinite hands with perfect play. Others think low house edge means blackjack is basically fair — which underestimates how much a few rule changes can multiply the number.
This page covers: what house edge actually measures (and doesn't), how the 0.5% baseline is calculated, how each blackjack rule variation moves the number, why blackjack has the lowest edge of any casino game, and the practical implications for how much a blackjack session actually costs you per hour. Everything here assumes you're playing basic strategy — if you're not, the effective house edge is much higher. See our basic strategy guide for the foundation.
House edge is the casino's expected return on each dollar wagered, calculated as a percentage over the long run. A 0.5% house edge means: for every $100 wagered, the casino expects to keep 50 cents on average, over millions of hands. The key words are "expected" and "long run" — any individual session can win or lose dramatically due to variance.
Mathematically, house edge is:
If perfect basic strategy generates an expected return of $0.995 for every $1 wagered on standard blackjack, the house edge is 1 − 0.995 = 0.005, or 0.5%.
This is different from hold percentage (what the casino actually keeps, which includes losses from bad play), variance (how much individual results deviate from the expected value), and return-to-player or RTP (the flip side of house edge — 99.5% RTP means 0.5% house edge). We cover RTP specifically in our RTP article; here we focus on the edge side of the equation.
The 0.5% figure comes from computer simulations of billions of hands played under specific rules with perfect basic strategy. The math accounts for every possible dealer-player combination, the probability of each, and the expected value of every strategic decision (hit, stand, double, split, surrender).
Peter Griffin's The Theory of Blackjack (1979) and later Don Schlesinger's Blackjack Attack refined these calculations to fractional-percentage precision. Modern figures come from combinatorial analysis rather than simulation — the game has been solved to many decimal places for every common rule set.
The baseline assumptions for the 0.5% figure:
Change any of these and the house edge moves. Some changes are dramatic — going from 3:2 to 6:5 blackjack payouts adds about 1.4% to the house edge, tripling it before you consider any other variation.
Each rule change adds or subtracts from the baseline house edge. All figures approximate, based on published mathematical analysis:
| Rule variation | Effect on house edge | Notes |
|---|---|---|
| Baseline (6-deck, S17, 3:2, DAS) | 0.50% | Standard reference point |
| Blackjack pays 6:5 instead of 3:2 | +1.39% | Single biggest player-hostile rule |
| Dealer hits soft 17 (H17) | +0.22% | Common on Vegas Strip low-limit |
| Single deck (typical H17 6:5) | +1.45% | Marketed as "single deck" but hostile |
| 8 decks instead of 6 | +0.02% | Marginal impact — nearly negligible |
| Double after split (DAS) not allowed | +0.14% | Removes a profitable player option |
| No re-splitting aces | +0.03% | Small but real |
| Late surrender allowed | −0.07% | Player-favorable if used correctly |
| Early surrender allowed | −0.62% | Rare but very player-favorable |
| Double on any two cards | −0.20% | Standard on best rules |
| Double restricted to 9-11 only | +0.09% | Common European variant |
| Continuous shuffle machine (CSM) | +0.05% | Removes deck-composition edge |
The pattern to notice: the 3:2 vs 6:5 payout is the single most important rule. A 6:5 game with all other rules favorable is still worse than a 3:2 game with several unfavorable rules. If you have any choice, choose 3:2 every time.
Combine the player-hostile rules — 6:5 payout, H17, no DAS, no surrender, CSM — and the house edge climbs toward 2% or higher. That's four times the baseline. It's still lower than slots (5-12%), roulette (~2.7%), or American keno (25-30%), but the game stops being the mathematical near-fair-play it's known for.
6-deck S17 3:2 DAS with basic strategy. The number you see quoted everywhere.
Typical downtown / Strip low-min table: 6:5, H17. Nearly 4× the baseline.
Everything hostile: 6:5, H17, CSM, restricted doubling. Avoid.
Downtown Vegas single-deck 3:2 H17: 0.19-0.26% edge. Rare but exists.
Compare blackjack against other casino games under standard rules and basic play:
| Game | House edge | Notes |
|---|---|---|
| Blackjack (basic strategy) | 0.5% | Standard 6-deck S17 3:2 DAS |
| Baccarat (Banker bet) | 1.06% | Best pure luck game |
| Craps (Pass line) | 1.41% | With odds bet, effective edge lower |
| Video poker (best variants) | 0.5% | Full-pay Jacks or Better, perfect play |
| European roulette | 2.70% | Single zero |
| American roulette | 5.26% | Double zero |
| Slots (average) | 5-15% | Varies widely; often not published |
| Keno | 25-30% | Worst edge in casino |
The reason blackjack is lowest: the player has strategic decisions to make. In roulette you spin and hope. In slots the outcome is determined the moment you press the button. In blackjack, every hand involves choices — hit, stand, double, split, surrender — and each choice affects the outcome. Games with player decision-making tend to have lower theoretical house edges, because the game is designed assuming players make optimal choices.
Video poker matches blackjack's edge because it also involves strategic decisions (which cards to hold). But video poker requires perfect play on every hand to achieve the theoretical edge — a mistake rate of 1% raises the effective edge substantially. Blackjack basic strategy is more forgiving: perfect play is achievable with a chart and a few weeks of practice.
"0.5% means I'll lose 50 cents for every $100." Only over an infinite number of hands, on average. Any individual session can win or lose thousands of dollars due to variance. A 1,000-hand session might see swings of 30-50 units in either direction. The house edge tells you the expected long-run outcome; variance tells you the range of short-run outcomes.
"Casinos rig the shoe so the edge is higher." No — the mathematical edge comes from rules (dealer plays last, players bust first losing immediately), not from any card manipulation. Regulated casinos are audited on RNG (for online blackjack) or dealer procedures (for physical tables). See our RNG article for the technical detail on online games.
"House edge is fixed." No — it depends entirely on the rule set. The exact same "blackjack" table can have edges from 0.19% to over 2% depending on which rules the casino sets. Always check the specific rules before playing. Look for: what does dealer do on soft 17? What's the blackjack payout? Is DAS allowed? Surrender available?
"With counting I can eliminate the house edge." Counting shifts the edge in the player's favor by 0.5-1.5% depending on system and rule set. That's a meaningful edge, but requires significant bankroll, betting variation, and practice. Modern countermeasures (CSMs, penetration limits, backing-off skilled players) make it harder than it was in the 1970s-80s. See our card counting guide for the realistic picture.
"More decks = worse edge." Only slightly. Going from 1 deck to 8 decks increases the edge by about 0.6% total, mostly from the reduction in split re-splitting opportunities and the shift in card composition. Individual deck count is a minor factor compared to payout rules and dealer soft-17 policy.
The 0.5% figure assumes perfect basic strategy on every hand. Real players make mistakes. Studies of casino data suggest average recreational players play blackjack at an effective house edge of 1.5-2%, three or four times the theoretical minimum — not because the rules are worse, but because the player deviates from basic strategy.
Common deviations that add to effective edge:
Each mistake typically costs 0.1-0.5% in effective edge. Combine several routine mistakes and a player at a 0.5% theoretical table is actually playing against a 1.5-2% effective edge. This is why basic strategy is worth the study time.
The house edge is a percentage per dollar wagered. To get to a practical dollar-per-hour cost, multiply house edge × average bet × hands per hour:
Example — standard $25 bet, 100 hands/hour, 0.5% edge:
Example — 6:5 game, $25 bet, 100 hands/hour, 1.89% edge:
Example — same game with average player mistakes (effective 1.5% edge):
Hands per hour vary significantly. Head-to-head with dealer: 200+ hands/hour. Full table (7 players): 50-70 hands/hour. Online RNG blackjack: 300+ hands/hour if you play fast. Live dealer online: similar to physical table pace.
These are expected hourly costs. Actual results in any given hour will vary widely due to variance. The math is: the more hands you play, the closer your actual results converge to the expected value. Play 100 hands and your outcome is nearly random around expectation. Play 100,000 hands (across many sessions) and you'll be very close to the expected loss.
Understanding house edge helps set realistic expectations for bankroll planning:
The house edge doesn't make blackjack a bad game — it's the lowest-edge game in the casino. But it's not zero. Play blackjack for the game itself, budget the expected cost, and don't play at 6:5 tables when 3:2 is available.
No. 0.5% is the baseline for standard 6-deck S17 3:2 DAS with basic strategy. Change any rule and the number changes. 6:5 payout adds 1.39%. Dealer hitting soft 17 adds 0.22%. Combine several unfavorable rules and the edge can exceed 2%. Always check the specific rules of the game you're playing.
They're the same information from opposite angles. 0.5% house edge = 99.5% return-to-player (RTP). Casinos and game providers may quote either number. See our RTP article for the RTP-focused view.
Per dollar wagered, over the long run. A single hand or single session can win or lose by amounts that have nothing to do with the theoretical edge. The math converges toward the expected value as you play more hands — in the short run, variance dominates.
Yes, but only under specific conditions. Counting can shift the edge from 0.5% against the player to 0.5-1.5% for the player, depending on system, penetration, betting spread, and rule set. Requires significant bankroll to weather variance, and casinos actively counter-measure counters. See our counting guide for the realistic picture.
Same math applies — depends on rules. Most licensed online blackjack uses 6-deck or 8-deck S17 3:2 rules, generating theoretical edges of 0.4-0.6% with basic strategy. Live dealer online games follow the same math as physical tables. See our RNG article for how online games handle the shoe.
Yes — taking insurance always hurts you unless you're counting cards and know the count justifies it. The insurance bet has its own house edge of about 7% at neutral count. Basic strategy is clear: never take insurance. This includes "even money" on your own blackjack — mathematically the same bad bet in disguise.
Side bets have much higher house edges — typically 3-7% depending on the specific payout structure. See our side bets article and 21+3 breakdown. As entertainment they're fine; as a bet strategy they're much worse EV than the main game.
Only if other rules stay the same. A "single deck" game with 6:5 payout and H17 has a much higher edge than a 6-deck game with 3:2 and S17. Modern "single deck" tables are almost always 6:5 with hostile rules — the fewer-decks marketing masks a worse game. Check rules before choosing based on deck count.