Blackjack RTP: how the rules, not a slider, set the edge

Where the edge comes from at all

The dealer's advantage has a single structural source: you act first. Bust before the dealer draws and you lose immediately, even if the dealer would have busted on the same shoe. That asymmetry is worth roughly 8% on its own.

Almost everything else in the rules is a rebate against that 8%. A natural blackjack paying 3:2 rather than even money, the option to double down, splitting pairs, and the dealer's obligation to follow a fixed drawing rule while you may use judgement — those give most of it back. The residue, on a good table played well, is under half a percent.

What each rule is worth

Rule effects are usually stated in percentage points of house edge. They are close to additive over the ranges that matter, so a table's edge can be estimated by starting from a reference game and adding the deltas.

A six-deck game where the dealer stands on soft 17, doubling after split is allowed and blackjack pays 3:2 comes out around 0.4% — a 99.6% RTP, better than almost anything else in the building. Change one thing, the blackjack payout, to 6:5, and the same table jumps to roughly 1.8%. The felt looks identical. The cost of playing it has more than quadrupled.

Basic strategy is part of the number

Every published blackjack edge assumes perfect basic strategy: the mathematically optimal action for every combination of your total and the dealer's upcard, derived by working out the expected value of hitting, standing, doubling and splitting in each case and taking the best.

It is a lookup table, not an intuition. Play by feel instead and the edge typically rises by one to two percentage points — which means an average player at a 0.4% table is really facing something closer to 2%, and has turned the best game on the floor into a mediocre one without touching the rules.

A handful of decisions carry most of that cost:

• Always split aces and eights. Two hands starting from an ace beat one soft 12. Sixteen is the worst total in the game; two hands starting from eight are better.
• Never split tens. Twenty is already a winning hand. Breaking it up is a losing trade in every published table.
• Hit hard 16 against a dealer 7 or higher. It feels terrible and it is correct. Standing on 16 against a strong upcard loses more often.
• Never take insurance. It is a side bet on the dealer holding a ten-value card, priced worse than the main game. On a six-deck shoe it carries an edge of roughly 7%.

Why deck count matters

Fewer decks favour the player, for a subtle reason: removal effects are larger. Drawing a card meaningfully changes what remains in a single deck and barely registers in an eight-deck shoe. This slightly increases the frequency of naturals and improves doubling outcomes.

It is also the reason blackjack is the one common casino game with genuine advantage play. Because the shoe has memory — cards removed do not come back until the shuffle — the composition of what remains shifts the edge as the shoe is dealt. Counting tracks that shift. Every other game on this site's list is memoryless, which is precisely why no betting system works on any of them.

Blackjack in a simulator

LarpCasino treats blackjack differently from its other five games for exactly this reason. Crash, Mines, Dice, Pinball and Slots have a house-edge slider. Blackjack does not, because its return is not something you can dial in — it falls out of the rules.

So the rules are what is tunable: deck count, whether the dealer hits or stands on soft 17, and the blackjack payout. The game quotes the RTP that perfect basic strategy returns for the table as configured, and the RTP simulator plays that same strategy against the same shoe over 10,000, 100,000 or a million hands.

That makes the table above testable rather than quotable. Run a million hands at 3:2, change nothing but the payout to 6:5, run it again, and read the difference off the empirical RTP. It is the cheapest possible way to learn what that rule costs, since the app has no real money in it at any point — credits are invented, nothing can be deposited or withdrawn, and it all runs on the device.

The short version

• The house edge starts around 8% from acting first, and the rules hand most of it back.
• A good six-deck table played perfectly sits near 0.4%; 6:5 blackjack alone pushes it past 1.8%.
• Published edges assume perfect basic strategy. Playing by feel costs one to two points.
• Insurance is a bad bet at any count you can see.
• Blackjack is the one common game with memory, which is why counting exists and betting systems still do not work anywhere else.