The big rules at a blackjack table are the easy ones to find: the payout, the soft 17 rule, the number of decks. Everything else lives in smaller type, on the felt or on the rule card. May you resplit aces? Does the dealer take doubles when the table plays without a hole card? What happens after the third split?
Those questions sound like trivia. Each has a price, computed to a hundredth of a percent, and once the headline rules match, those prices are most of what still separates one table from another. Our piece on using a house edge calculator priced six headline rules. This article prices the smaller print.
How to read the prices
All figures below come from the blackjack rule variation tables Wizard of Odds publishes. They are effects on the player's expected return, with basic strategy adjusted for the rule in question, measured against one fixed baseline: eight decks, dealer stands on soft 17, double on any first two cards, double after split, split to four hands. A plus sign helps the player, a minus sign hurts, and every figure is in percent.
Two cautions before the tables. Each figure was computed with every other rule held at the baseline, so adding them is an approximation, the same caveat our calculator article makes. For the same reason, a figure quoted against a different baseline can differ slightly from the one here.
Splitting, priced
Splitting has the most separate prices of any group here. The baseline allows splitting to four hands, so each restriction takes something away, and the ace rules sit on top.
| Splitting rule | Effect on player return |
|---|---|
| Split to only 3 hands | -0.01 |
| Player may not resplit | -0.10 |
| Player may not split aces | -0.18 |
| Player may not split at all | -0.57 |
| Player may resplit aces | +0.08 |
| Player may draw to split aces | +0.19 |
| Player may double or stand after splitting aces | +0.08 |
| Ace and ten after splitting aces counts as a blackjack paying 3:2 | +0.21 |
Read the top of the table against the middle. Splitting to only three hands sounds strict and costs 0.01 percent, effectively nothing; our article on resplit limits covers how rarely the fourth hand even comes up. Not being allowed to split aces at all costs 0.18 percent, about eighteen times as much on these rounded figures, from a rule that sounds no more restrictive in a sentence. The price follows how often a rule binds and how much the lost option was worth, not how it sounds.
On the gains side, drawing to split aces (+0.19) and counting an ace and ten after splitting aces as a 3:2 blackjack (+0.21) are the most generous entries in the group. The resplit aces row, +0.08, is the subject of its own article.
Doubling, priced
Doubling restrictions come in steps. The baseline allows doubling on any first two cards.
| Doubling rule | Effect on player return |
|---|---|
| Double on 9 through 11 only | -0.09 |
| Double on 10 and 11 only | -0.18 |
| No double after split | -0.14 |
| Player may double on any number of cards | +0.23 |
| Player may not double at all | -1.48 |
Notice the shape. Cutting the doubling window from any two cards down to 9 through 11 costs 0.09, and cutting it further to 10 and 11 only doubles the cost to 0.18. Those are the real world cases, and our piece on tables that restrict doubling explains what goes at each step: mainly the soft doubles first, then doubling hard 9. The bottom row is the thought experiment: with doubling removed entirely, the player loses 1.48 percent, more than the 6 to 5 payout costs. The no double after split row, -0.14, is the rule DAS and why it matters discusses.
21 TrainerFree on iPhone and Android. Full rule customization, a house edge calculator, and charts that redraw for whatever rules you pick.
Get the app
Surrender, priced one card at a time
The source prices surrender one dealer upcard at a time, a ten and an ace. The baseline has no surrender, so every row here is a gain or a zero.
| Surrender rule | Effect on player return |
|---|---|
| Late surrender against a ten | +0.07 |
| Late surrender against an ace | +0.00 |
| Early surrender against a ten | +0.24 |
| Early surrender against an ace | +0.39 |
| Late surrender after splitting | +0.00 |
Some arithmetic the table rewards. The two late surrender rows are 0.07 against a ten plus 0.00 against an ace, which totals 0.07; elsewhere on this site the rule is quoted at about 0.08, and the two figures agree to rounding. The two early surrender rows are 0.24 plus 0.39, which totals 0.63. Dividing 0.63 by 0.07 gives 9: on these figures early surrender is worth about nine times late surrender (nearly eight on the rounder 0.6 and 0.08). A concession that size did not last, a story told in its own article and compared rule by rule in early versus late surrender. The two zero rows are zeros to two decimal places: whatever late surrender against an ace and surrender after splitting are worth under these rules, it rounds to nothing.
The dealer's side: soft 17, the peek, and the hole card
| Dealer and deal rule | Effect on player return |
|---|---|
| Dealer hits soft 17 | -0.22 |
| Dealer must stop with six cards | +0.00 |
| Dealer does not peek with an ace showing | -0.01 |
| Dealer does not peek with a ten showing | -0.10 |
| European no hole card | -0.11 |
Hitting soft 17, worth -0.22, is the largest item in this group, and S17 versus H17 gets its own treatment. The peek rows reward a second look. In both, the source's condition is that a dealer blackjack the peek would have caught takes all bets made, doubles and splits included. On that footing, not peeking with an ace up costs almost nothing (-0.01), while not peeking with a ten up costs about ten times as much (-0.10). Under the European no hole card rule, -0.11, the dealer takes no hole card or never peeks at it, and a dealer blackjack takes all bets including doubles and splits, except that a player blackjack pushes; our article on no hole card variants untangles the several ways casinos settle those bets.
Deck count, and where the gains actually sit
| Deck count | Effect on player return |
|---|---|
| Six decks (from eight) | +0.02 |
| Five decks | +0.03 |
| Four decks | +0.06 |
| Two decks | +0.19 |
| One deck | +0.48 |
The distribution is the interesting part. Going from eight decks to six is worth 0.02 percent: real, but barely. Going from eight decks to one is worth 0.48, and two decks alone are worth 0.19. Most of the deck effect sits at one and two decks: the step from two decks to one is 0.48 minus 0.19, or 0.29, while the whole slide from eight decks to four is 0.06. Single deck versus eight decks covers the deck effect in more depth, along with the classic trap of pairing one deck with 6 to 5.
Stacking the small print
Individual prices are trivia until you add them. A poor table might combine three rules from this list: no resplitting (-0.10), doubling on 10 and 11 only (-0.18), and a dealer who hits soft 17 (-0.22). Stacked, that is -0.10 plus -0.18 plus -0.22, a total of -0.50. Dividing 0.50 by 1.39 shows the stack costs a bit more than a third of what the 6 to 5 payout costs.
Remember the additive caveat: -0.50 is an estimate built from figures computed one at a time, and the true combined effect can differ slightly because the rules interact. For understanding the game's structure, the estimate is the right tool.
21 Trainer includes a house edge calculator. The app also offers full rule customization (deck count, the soft 17 rule, double after split, surrender on or off, split limits and the 3:2 or 6:5 payout), and its strategy charts redraw for the rules you select, a reminder that a rule set is a set of playing instructions as well as a price.
Frequently asked questions
Which blackjack rules cost the player the least?
Against the source's baseline, some restrictions are nearly free. Splitting to only three hands instead of four costs 0.01 percent, and the no peek rule with an ace showing costs the same. Three rules are priced at 0.00, meaning any effect rounds to nothing: late surrender against an ace, late surrender after splitting, and a dealer who must stop with six cards.
How much more is early surrender worth than late surrender?
Per the Wizard of Odds tables, early surrender is worth 0.24 percent against a ten and 0.39 against an ace, for 0.63 in total. Late surrender is worth 0.07 against a ten and 0.00 against an ace, for 0.07 in total. Dividing 0.63 by 0.07 gives 9, so on these figures early surrender is worth about nine times late surrender. Our early surrender article tells how a concession that size came to be withdrawn.
Is losing the double option worse than a 6 to 5 payout?
Slightly, yes. Removing doubling entirely costs the player 1.48 percent against the baseline, while the 6 to 5 blackjack payout costs 1.39 percent. Both are far larger than any other figure in this article. The doubling restrictions this site covers, doubling on 9 through 11 only or on 10 and 11 only, cost 0.09 and 0.18 percent.
Does each deck you remove help by the same amount?
No. Relative to eight decks, six decks give the player back only 0.02 percent and five decks 0.03. The gains concentrate at the bottom of the range: two decks are worth 0.19 and a single deck 0.48. Most of the deck effect sits at one and two decks, and even the 0.48 a single deck gives back is far less than the 1.39 a 6 to 5 payout takes.
Can you test how a rule combination changes the odds?
Yes, approximately. Add the published effect of each rule that differs from the baseline, remembering that the sum is an estimate because rules interact. 21 Trainer includes a house edge calculator. The app's full rule customization covers deck count, the soft 17 rule, double after split, surrender on or off, split limits and the 3:2 or 6:5 payout, and its strategy charts redraw for the rules you select. Free on iPhone and Android.