Blackjack
The only game on the floor where a disciplined player gets within half a percent of the house. Here is the full rule set we deal, the math behind it, and the decisions that separate a long night from a short one.
By The Lucky's Desk · Published · Updated
The rule set we deal
Blackjack rules vary wildly from property to property, and those variations are worth more than any strategy adjustment you will ever make. The table at Lucky's runs the following, and every number on this page is derived from exactly this configuration:
- Six decks, dealt from a shoe, shuffled at roughly 75% penetration.
- A natural blackjack pays 3:2. Never 6:5 — that single change would triple the house edge.
- Dealer stands on soft 17.
- Double down permitted on any two cards.
- Split up to three hands; split aces receive one card each.
- No surrender. Insurance offered when the dealer shows an ace.
Under that rule set, perfect basic strategy leaves the house holding roughly a 0.46% edge. That means a player betting $100 a hand and never making a strategy error expects to lose about 46 cents per hand on average — a rounding error compared to what the slots on the same floor will take from you.
A hand, worked through
Rules read like a legal document until you watch one hand play out, so here is a real sequence from the table.
You bet $200. You are dealt the eight of hearts and the three of clubs — a hard eleven. The dealer's up card is the six of diamonds.
Eleven against a six is the single strongest double-down in the game. The reasoning is two-sided. On your side, eleven cannot bust with one card, and roughly 31% of the shoe (any ten, jack, queen or king) makes you twenty-one. On the dealer's side, a six is the weakest card they can show: they must draw to at least seventeen, and they will bust that hand about 42% of the time. You are putting more money out when both halves of the equation favour you.
You double to $400. You catch the ten of spades — twenty-one, though not a natural, so it pays even money rather than 3:2. The dealer flips a nine for fifteen, draws a seven, and busts at twenty-two. You are up $400 on a hand you would have won for $200 by playing it timidly.
Now change one card. Same eleven, but the dealer shows an ace. Basic strategy says hit, do not double. The ace is the dealer's strongest card; they make a pat hand of seventeen or better more than 80% of the time once they do not have blackjack outright. Doubling there quietly hands back most of what you earned doubling against the six.
What each decision is actually worth
The table below shows the expected value of the main plays, expressed as return per dollar wagered. Anything above 1.00 is profitable for you.
| Situation | Correct play | Return per $1 |
|---|---|---|
| Natural blackjack | Stand (auto) | $2.50 returned |
| Hard 11 vs dealer 6 | Double | ~$1.67 |
| A,7 (soft 18) vs dealer 3 | Double | ~$1.13 |
| Pair of 8s vs dealer 10 | Split | ~$0.52 (least-bad option) |
| Hard 16 vs dealer 10 | Hit | ~$0.46 |
| Insurance | Decline | ~$0.93 |
| Pair of 10s vs dealer 6 | Stand, never split | ~$1.67 |
Two rows there deserve a second look. Sixteen against a ten is the most hated hand in blackjack and it loses money whatever you do — hitting is simply the option that loses less. And splitting eights against a ten looks insane until you notice that a hard sixteen is nearly worthless; two hands starting at eight are each worth more than one hand starting at sixteen, even against the dealer's best cards.
Insurance, and why the answer is always no
When the dealer shows an ace, you are offered insurance at 2:1 for half your bet. It sounds like protection. It is a side bet on whether the dealer's hole card is a ten, and the math is unambiguous: in a six-deck shoe there are 96 ten-value cards among 311 unseen cards, so the hole card is a ten about 30.9% of the time. A 2:1 payout needs 33.3% to break even. You are giving up roughly seven cents on the dollar every time you take it, and you take it precisely when you feel most anxious — which is how the bet was designed.
The only players for whom insurance is correct are card counters who know the remaining shoe is ten-rich. If you are not tracking the count, decline it every single time, including when you hold a blackjack yourself and the dealer offers you "even money." Even money is insurance wearing a different hat.
The mistakes that cost the most
- Standing on soft hands. A,6 is not a seventeen, it is a seven that cannot bust. Players who stand on soft seventeen against a nine or ten give away more edge than almost any other single error.
- Never splitting aces. Two hands each starting with an ace are enormously valuable. Playing A,A as a soft twelve is throwing away money.
- Playing by the dealer's rules. "Never bust" sounds safe. It costs about 6% — twelve times the house edge — because you stand on stiff hands when the dealer is strong.
- Chasing losses with bet size. Doubling your bet after a loss does not change the odds of the next hand. It changes how fast you reach the table maximum, which is where the strategy dies.
How our table differs from a real casino
Two deliberate departures. First, our shoe is reshuffled more often than a real six-deck game, which suppresses deep counting — the point of the game here is the story unfolding around the table, not a grind. Second, winning several large hands in a row raises your Suspicion meter. In a real casino the pit boss watches your bet spread; here the same instinct is modelled directly. A player who runs $500 into $40,000 at one table will find the floor gets very interested in who they are, which has consequences the math on this page does not cover.
Everything else — the payouts, the dealer's drawing rules, the effect of each decision — matches a real six-deck game, so the strategy you learn here is the strategy that applies at a real table.
Common questions
Does card counting work at Lucky's?
Partially. Tracking high and low cards will improve your results within a shoe, but frequent reshuffles limit how far the count can run, and the Suspicion system punishes an obvious bet spread.
Should I take even money on a blackjack?
No. It guarantees a 1:1 return on a hand that averages better than 1:1 if you simply let it play out.
Is a five-card hand worth anything?
No. "Five card Charlie" is a house rule at some properties; we do not deal it. Twenty-one beats twenty regardless of how many cards it took.
Why did my twenty lose to the dealer's twenty?
It did not lose — equal totals push and your stake is returned. Naturals are the exception: a two-card twenty-one beats a three-card twenty-one.
Next: the full basic strategy walkthrough, or compare this table against the rest of the floor on the house edge page.