H17 House Edge — Cost of Dealer Hitting Soft 17

H17 means the dealer hits soft 17. If the dealer has Ace-6, or another soft total of 17, the hand is not finished: the dealer must draw at least one more card.

For the player, H17 is worse than S17 when the rest of the rules are identical. A standard blackjack rule-variation benchmark puts the cost at about 0.22 percentage points of house edge.

That does not mean every H17 table has the same edge, and it does not mean H17 is the worst rule you can encounter. It means the dealer’s extra draw changes the outcome distribution enough to move expectation toward the casino.

This page focuses on what the H17 surcharge costs and how it interacts with player strategy. For the direct side-by-side explanation of H17 and S17 procedures, see Hit Soft 17 vs Stand.

The first H17 draw is unusually safe for the dealer

A soft 17 contains an ace counted as 11. The standard example is Ace-6.

Under H17, the dealer draws.

The key point is that the ace can change from 11 to 1. That means the first additional card cannot bust Ace-6.

Examples:

Some draws improve the hand immediately. Others turn it into a lower hard total that may require further draws. Those later draws can bust the dealer, but the extra improvement opportunities are valuable enough that H17 favors the house overall.

That is why “the dealer hits more, so the dealer busts more” is incomplete. Bust frequency is only one part of expected value.

The approximate H17 rule cost

If two tables are identical except for the soft-17 procedure, a useful approximation is:

[ E_{H17} \approx E_{S17} + 0.22% ]

where the percentage is expressed in percentage points of house edge.

If an S17 game is estimated at 0.45%, the otherwise identical H17 game would be roughly:

[ 0.45% + 0.22% = 0.67% ]

The exact total can vary because blackjack rule effects are not perfectly independent. Use 0.22 points as a comparison benchmark, not as a promise that any particular table has a 0.67% edge.

For complete rule-package analysis, see House Edge by Rules.

What H17 costs per amount wagered

The cleanest way to understand 0.22 percentage points is to apply it to action.

[ \text{Extra expected loss from H17} \approx A \times 0.0022 ]

where (A) is comparable wagering action.

The word expected matters. You do not lose $22 every time you cycle $10,000 through an H17 table. In one session you may win or lose hundreds or thousands because blackjack variance is much larger than this small average-rule effect.

The H17 cost appears as a long-run difference across repeated comparable action.

One hour at a $25 table

Suppose you make about 70 initial $25 wagers in an hour.

Initial wager action:

[ $25 \times 70 = $1{,}750 ]

The incremental H17 cost, using 0.22 percentage points, is:

[ $1{,}750 \times 0.0022 = $3.85 ]

So if two otherwise identical tables differ only on H17/S17, the H17 table is roughly $3.85 worse per hour at that simplified action level.

This calculation deliberately treats the initial wager as the denominator for comparison. Actual session action can be higher because doubles and splits add money. If you want a complete hourly framework, use Blackjack Expected Loss Per Hour.

H17 changes some correct player decisions

The dealer’s rule changes not only the edge but also some basic-strategy decisions.

For common four-to-eight-deck U.S.-style games, H17 strategy modifications relative to a typical S17 chart can include:

• double 11 against a dealer ace;
• double soft 18 against a dealer 2;
• double soft 19 against a dealer 6;
• when late surrender is available, surrender additional hands against a dealer ace.

Exact strategy depends on deck count, surrender, DAS, hole-card procedure, and other rules. Do not use those examples as a complete strategy chart.

The lesson is that a player who notices “dealer hits soft 17” should also verify that the strategy chart being used is an H17 chart. Blackjack Basic Strategy is the appropriate starting point.

Why H17 is smaller than the 6:5 problem

Players sometimes focus on the soft-17 sign because it is visible and easy to understand. But H17 is often not the biggest issue on the felt.

A common benchmark comparison looks like this:

The exact values vary with the baseline, but the scale is useful.

A 3:2 H17 table can easily be a much better game than a 6:5 S17 table. If you have to choose what to inspect first, verify the blackjack payout before obsessing over H17.

3:2 vs 6:5 Blackjack covers that larger payout difference.

Percentage points are not the same as a 0.22% relative increase

This distinction is easy to miss.

If an S17 game has a 0.50% house edge and H17 adds about 0.22 percentage points, the H17 estimate becomes roughly 0.72%.

That is not a 0.22% relative increase. Relative to the original 0.50% edge, the increase is:

[ \frac{0.72 - 0.50}{0.50} = 0.44 ]

or 44% larger.

That does not mean the player loses 44% of bets. It means the long-run house-edge percentage rose from 0.50% to 0.72%.

For blackjack rules, percentage points are usually the clearer language because the baseline edge can be small.

Why dealer bust rate alone cannot price H17

A player watching the game may notice that H17 sometimes sends the dealer from soft 17 into a sequence that eventually busts.

That observation is true but insufficient.

To calculate the rule value, you must account for every possible outcome after the extra draw:

• dealer remains on 17 after the ace converts to 1;
• dealer finishes on 18;
• dealer finishes on 19;
• dealer finishes on 20;
• dealer finishes on 21;
• dealer draws again from a lower hard total;
• dealer eventually busts.

Each outcome affects different player totals differently. Improving from 17 to 20 is valuable against player 18 and 19; remaining on 17 creates pushes against player 17; busting pays every surviving player hand.

The house-edge effect is the probability-weighted sum of all those changes across all relevant starting hands and dealer upcards. That is why a rigorous value comes from combinatorial analysis or simulation, not from comparing one visible bust percentage.

H17 and the player’s blackjack payout are separate

H17 does not change the amount paid on a natural blackjack.

If the table pays 3:2, a $20 blackjack still wins $30 whether the dealer uses H17 or S17, subject to the usual dealer-blackjack and push rules. If the table pays 6:5, the same $20 natural wins only $24.

These are separate rule levers:

• blackjack payout changes the value of one of the player’s most favorable outcomes;
• H17 changes the dealer’s play after a non-blackjack soft 17 occurs.

The fact that both affect house edge does not make them interchangeable.

That distinction is especially useful when a casino markets an H17 game with otherwise strong rules. You can price the H17 disadvantage without confusing it with the payout schedule.

H17 does not change the meaning of a soft player hand

“Dealer hits soft 17” applies to the dealer procedure. It does not mean the player is required to hit every soft 17.

Player decisions remain strategy choices governed by the player’s cards, the dealer upcard, and the rules.

For example, a player soft 18 may stand, hit, or double depending on the dealer upcard and rules. The dealer has no comparable discretion: if H17 applies and the dealer reaches soft 17, the dealer draws.

This is why the table abbreviation H17 describes a dealer rule, not a general instruction about all soft hands. Soft Hand Strategy handles the player side separately.

H17 can coexist with very good rules

H17 is a negative rule for the player, but a table should be judged as a package.

For example, a game might offer:

• 3:2 blackjack;
• H17;
• double after split;
• late surrender;
• resplitting aces;
• a reasonable deck count.

Another table might use S17 but pay 6:5 and restrict doubling.

The first game can still be superior.

Do not classify every H17 table as “bad blackjack.” Classify it as 0.22 points worse than the same table would be under S17, then evaluate everything else.

H17 and surrender against an ace

The dealer ace is where H17 strategy differences become especially visible.

When late surrender is available, H17 can make surrender correct on some hands against a dealer ace that would continue under S17. In four-or-more-deck total-dependent strategy, for example, H17 can call for surrendering hard 15 and hard 17 against an ace in situations where the comparable S17 chart does not.

Pairs can add their own exceptions depending on decks and DAS.

The reason is not mystical. H17 changes the dealer’s conditional outcome distribution after the ace upcard survives the blackjack check.

Because surrender rules vary so much by casino and jurisdiction, use a chart that matches the exact game. Early Surrender vs Late Surrender covers the timing distinction.

H17 does not make insurance better

Insurance asks a different question: does the dealer have a ten-value card under the ace right now?

The H17 rule applies only if the dealer does not have blackjack and play continues. It does not make the insurance wager profitable for a basic-strategy player.

A non-counting player should not use the H17 sign as a reason to insure the hand. The relevant mathematics are the unseen ten-value cards, not whether the dealer would later draw on Ace-6.

See When to Take Insurance for the separate insurance analysis.

Dealer H17 procedure is a table rule, not discretion

A blackjack dealer should not decide from mood or judgment whether to hit Ace-6. The procedure is set by the game rules.

Massachusetts blackjack equipment standards, for example, require disclosure of the dealer draw rule and provide H17 wording that the dealer draws to 16 and soft 17 and stands on hard 17 and all 18s. The Massachusetts Gaming Commission rule text is a useful example of how the distinction is formally posted.

If the table sign says H17, expect the dealer to hit every soft 17 required by the rules. If the sign is unclear, ask before making the first wager.

The mathematical benchmark and strategy source

Wizard of Odds lists “dealer hits on soft 17” at about a 0.22% reduction in player return among common blackjack rule variations. Its multi-deck blackjack strategy guide also identifies several common H17 strategy modifications relative to S17.

The two facts belong together: H17 changes the dealer’s expectation and can change the best player response.

Because rule interactions exist, do not take a single 0.22 figure and reconstruct an entire table edge by arithmetic unless the rest of the assumptions are known.

H17, deck count, and “premium” tables

H17 is more common in some market segments than others, while S17 may be reserved for particular limits or pits. That can create the impression that the soft-17 rule itself determines game quality.

It does not.

A table’s mathematical price is a combination of:

• deck count;
• payout on natural blackjack;
• dealer soft-17 rule;
• DAS;
• resplitting rules;
• surrender;
• hole-card or no-hole-card procedure;
• doubling restrictions;
• strategy.

A casino can pair H17 with favorable rules or S17 with poor ones.

For deck-specific effects, see House Edge by Deck Count.

What H17 means for a card counter

Card counters are not exempt from the base rule.

H17 changes the off-the-top expectation of the game and some strategy indexes. A counter may still find a positive expectation in favorable shoe conditions, but H17 starts the game from a less favorable baseline than otherwise identical S17.

The practical effect then depends on penetration, bet spread, game speed, deck count, and the count system being used.

This page does not provide an advantage-play system. Card Counting Basics explains the distinction between changing card composition and ordinary basic-strategy expectation.

A table-selection example

You find three six-deck games:

Percentage quality probably favors A over B because of S17, while both have a major payout advantage over C.

But if your comfortable wager is $25, the $50 minimum at A may create more dollar exposure than the small percentage improvement justifies for you.

The correct decision can therefore have two layers:

1. Which table has the better percentage expectation?
2. Which stake and pace fit the amount you are actually willing to risk?

Good game selection requires both.

H17 and short-session results

The H17 penalty is a small long-run shift, not a short-run prediction.

Suppose you play 100 hands. The incremental expected cost associated with 0.22 percentage points is tiny compared with the normal standard deviation of blackjack results.

You might win substantially at H17 and lose substantially at S17. That does not reverse which rule is better.

The same distinction applies to every small rule effect. Expected value describes the average over repeated comparable play; Blackjack Variance Explained describes why individual sessions wander far from that average.

How to treat H17 in practice

When you see “dealer hits soft 17”:

1. confirm that blackjack still pays 3:2;
2. use an H17 strategy chart;
3. include roughly 0.22 percentage points as a negative rule factor when comparing an otherwise identical S17 game;
4. check DAS, surrender, deck count, and other restrictions;
5. compare table minimum and speed in dollars, not percentages alone.

H17 is worth noticing. It is not worth ignoring larger rules.

Bottom line

H17 costs the player roughly 0.22 percentage points compared with an otherwise identical S17 game, because the dealer gets an extra opportunity to improve soft 17; the rule also changes a small set of optimal player decisions.

This page prices H17; it does not merely define it

The abbreviation H17 can be defined in one line: the dealer hits soft 17. The indexing-worthy question here is different: what does that rule cost when compared with an otherwise identical S17 game? The approximate 0.22-percentage-point benchmark is a rule-value estimate, and it must be interpreted alongside payout, deck count, doubling, splitting, and surrender rules.

For terminology and dealer procedure, use H17. Use this page when comparing tables, translating the rule difference into expected dollars wagered, or checking which strategy deviations are triggered by H17.

Read S17 as the reverse side of the same rule-price comparison

There is no separate economic mechanism for S17. S17 is simply the player-friendlier side of the same H17/S17 comparison. If changing S17 to H17 increases the house edge by roughly 0.2 percentage points in a benchmark ruleset, changing H17 back to S17 reduces it by the same comparison amount when all other rules are held constant.

This page therefore prices both directions in one place. Use it to compare otherwise similar tables, then check payout, deck count, DAS, surrender, resplitting and other rules before deciding that S17 alone makes a table good.