How rule variations change the house edge

Every blackjack rule has a price, and most of those prices have been measured. Under six-deck rules where the dealer stands on soft 17, the player may double on any two cards and double after splitting, Wizard of Odds’ combinatorial analysis puts the basic-strategy house edge at about 0.40%. Each rule change moves that figure by a known amount, and the table below lets you estimate any table’s edge by adding those amounts to a baseline.

How the figures are measured

Wizard of Odds measures each variation as a single change from one fixed baseline: eight decks, dealer stands on soft 17 (S17), double on any first two cards, double after split (DAS) allowed, split to four hands. The site reports the effect on the player’s expected return; this page flips the sign so that a positive number means the house edge rises. All figures assume correct basic strategy for the rules in play.

Two consequences follow. First, the effects are approximately additive but not exactly: each was measured in isolation, so combining several introduces small interaction errors. Second, a rule’s value depends on the rest of the game. Late surrender, for example, is worth more when the dealer hits soft 17, because more surrender cells become correct.

Master table of rule effects

Figures are percentage points of house edge relative to the eight-deck S17 DAS baseline, from Wizard of Odds’ rule-variations page unless noted. Rounded to two decimals.

Continuous shuffling machines are a procedure rather than a rule. Wizard of Odds’ cut-card analysis found that a CSM lowers the basic-strategy edge by about 0.02 points in six decks and 0.11 points in single deck, because it removes the cut-card effect that favours the house in hand-shuffled games. The catch is speed: more hands per hour at the same edge means more expected loss per hour.

Worked example: building a table’s edge from the baseline

Start from a known total rather than the eight-deck baseline, because most casinos deal six decks. Wizard of Odds’ Appendix 9 gives exact combinatorial figures for six decks with resplit to four hands and no drawing to split aces:

Now add adjustments. Suppose a Las Vegas Strip table offers six decks, H17, DAS, no surrender and 6:5 blackjack.

1. Six-deck H17 DAS baseline: 0.62%.
2. Add 6:5 payout: +1.39 points.
3. Total: about 2.01%.

For a six-deck H17 game that also bans double after split, the direct route is 0.76% + 1.39 = 2.15%. BlackjackInfo quotes “about 2.1%” for the same combination, which is the kind of agreement you should expect from additive estimates.

Going the other way, a generous six-deck game with S17, DAS and late surrender: 0.40% − 0.07 = about 0.33%. BlackjackInfo quotes 0.36% for that package. The gap of a few hundredths of a percent is the interaction error mentioned above; it does not change any decision you would make about the table.

Which rules actually matter at the table

The table has a clear hierarchy. Payouts first, then the soft-17 rule, then doubling and splitting restrictions, then everything else.

A 6:5 payout costs more than every other common rule combined. The deck count, which casinos advertise loudly, is worth under half a point even at its extreme, and a single-deck 6:5 game (about −0.48 + 1.39) is worse than an eight-deck 3:2 game by nearly a full point. Treat “single deck” on a sign as a prompt to read the payout line, not as a benefit in itself.

H17 is the next most important rule because it is both expensive and nearly invisible: the sign says “dealer must hit soft 17” in small print and nothing else about the game looks different. Doubling and splitting rules follow, each worth a tenth or two of a point. Surrender is a small positive. The settlement rule in no-hole-card games, which decides whether you lose doubled and split bets to a dealer blackjack, is worth 0.11 points in its harshest form; European no-hole-card rules covers the strategy changes it forces.

Variant rules are not additive in the same way

The bottom rows of the table show why novelty games need their own analysis. “Dealer 22 pushes” is worth nearly seven points to the house on its own; “dealer wins all ties” nearly nine. No casino offers those rules in isolation. Blackjack Switch pairs the push-22 rule with the ability to swap cards between two hands; Free Bet Blackjack pairs it with free doubles and splits; Double Exposure pairs dealer-wins-ties with both dealer cards face up. The player-friendly half of each game gives most of that value back, and the published edge for the whole variant (typically 0.5% to 1%) is the only number worth remembering. See Blackjack Switch, Free Bet Blackjack and Double Exposure and other variants.

Using a calculator for a precise figure

When the estimate matters, use a tool that evaluates the full rule set at once. Wizard of Odds’ house edge calculator accepts deck count, soft-17 rule, DAS, doubling restrictions, resplit limits, resplitting and hitting split aces, original-bets-only, surrender and payout, and reports three figures: optimal composition-dependent play, total-dependent basic strategy with a cut card, and basic strategy with a continuous shuffler. The cut-card figure is the one that matches a hand-shuffled shoe game.

Whatever the number, it describes the long run under correct play. A 0.4% edge on a $10 flat bet is an expected loss of 4 cents a hand; the variance around that expectation is enormous, which is why bankroll and variance deserves as much attention as rule selection. No rule set offered by a casino gives a basic-strategy player an advantage; the goal of reading the sign is to lose less, not to win.

Related reading

• Choosing a blackjack table
• H17 vs S17
• 6:5 blackjack vs 3:2
• How the house edge is calculated