The House Edge Varies by a Factor of Thirty Across a Single Casino Floor
Walk from one end of a casino to the other and the cost of an hour’s play changes by more than most people would believe.
Blackjack played correctly on good rules carries a house edge of around half a percent. The “any seven” bet on a craps table carries 16.67%. Both are on the same floor, both are perfectly legitimate, and one is roughly thirty times more expensive than the other per dollar wagered.
Nobody hides this. The figures are published, calculable and uncontested. They are simply never presented side by side, because there is no commercial reason for anyone to do so.
That gap between available and presented is the interesting part, and it survived the move online in a different form. On a floor, layout does the work: better-value games sit away from entrances and finding the single-zero wheel means knowing to look for it. Online, the equivalent levers are lobby ordering, default filters and promotional placement. According to the PokerTube site, which reviews platforms across a range of markets, game presentation and navigation now carry the weight the physical floor layout used to.
The Actual Numbers
Here is the spread, expressed as expected loss per $100 wagered.
Blackjack, basic strategy, favourable rules: 0.50%. The lowest number available on a table game, and it requires no card counting. It requires knowing when to hit a 16 against a dealer 7 and doing it every time. The underlying combinatorial analysis has been settled since the 1960s and is not in dispute anywhere.
Baccarat, banker bet: 1.06%. After the 5% commission on winning banker bets. The player bet is 1.24%, and the difference is why anyone who has read anything backs the banker.
Craps, pass line: 1.41%. One of the better bets in the building, and it sits on a table surrounded by some of the worst.
Roulette, single zero: 2.70%. European wheel.
Roulette, double zero: 5.26%. American wheel. Same game, same bets, one extra pocket, and the cost of playing it nearly doubles.
Baccarat, tie at 8 to 1: 14.36%. On the same table as the 1.06% bet.
Craps, any seven: 16.67%. The most expensive standard bet most floors offer.
Converted into an evening, the abstraction disappears. A hundred hands at $25 costs $12.50 in expectation on blackjack and $416.75 on the any-seven bet.
The Roulette Comparison Is the Cleanest
If you want one example that demonstrates the whole point, use the two wheels.
Single zero and double zero roulette are the same game. Identical bet types, identical payouts, identical procedure. The American wheel has one additional pocket.
That single pocket moves the edge from 2.70% to 5.26%. Over 100 spins at $10, that is $27.00 against $52.60, and the games are visually near-identical to anyone who is not counting pockets.
Which is the practical lesson: the difference between an expensive game and a reasonable one is frequently a detail that does not announce itself. Nobody puts a sign on the American wheel.
What Moves the Number Within a Game
Blackjack is the game where the published figure varies most with conditions, and the variation is larger than the difference between many separate games.
Blackjack paying 6:5 instead of 3:2 on a natural adds roughly 1.4 percentage points to the house edge. That is a bigger swing than the entire edge on the pass line, and it comes from a single rule printed on the felt.
Dealer hitting soft 17 rather than standing adds about 0.2 points.
Deck count, doubling rules, resplitting and surrender each move it by smaller amounts, and they compound.
So a blackjack table can be a 0.5% game or a 2% game depending on rules that are all disclosed and rarely read. The player who checks the payout on naturals before sitting down has done more for their expected cost than any betting system could.
What the Move Online Changed
The transition left the arithmetic intact and changed how it is encountered.
One genuine improvement came with it. Online games publish their return figures in the information screen, which no physical slot machine has ever done. A player who checks can know the number before playing, which was structurally impossible on a floor.
Those figures are not self-reported either. Independent test houses including eCOGRA verify that a deployed game’s mathematics match its stated return before it reaches a licensed market, which is a layer the physical floor never had.
Whether anyone checks is a separate question.
What This Is Not
Two clarifications, because the numbers invite a wrong conclusion.
A lower edge is not a winning bet. Every figure above is positive from the house’s side. Blackjack at 0.5% is still a game you lose at over time, just considerably more slowly. Nothing here identifies a profitable play, because none exists in the standard offering.
Edge is per wager, not per hour. A game with a low edge dealt quickly can cost more per hour than a high-edge game played slowly. Baccarat at 1.06% dealt at 70 hands an hour against roulette at 2.70% spun at 40 an hour is a genuinely close comparison, and hands per hour is the variable nobody tracks.
That second point matters more than most of the first section. Cost is edge multiplied by volume, and volume is the part under a player’s control.
The Short Version
Three things worth carrying.
The spread is enormous and public. Thirty-fold between the best and worst standard bets in the same building, with nothing hidden.
The conditions matter as much as the game. 6:5 blackjack is a worse proposition than 3:2 blackjack by more than the entire margin on several other games.
Volume is the multiplier. Whatever the percentage, it applies to every dollar wagered, and the number of dollars wagered per hour is the one input that is entirely yours.
Disclaimer: This article is intended for informational and educational purposes only. It explains differences in casino house edges and does not constitute gambling, financial, or investment advice. Gambling involves risk, and no strategy can guarantee winnings. Readers should gamble responsibly, follow applicable laws and regulations, and never wager more than they can afford to lose.