Casino House Edge Statistics: Every Major Game in One Table (2026)

Every casino game keeps a mathematically fixed share of what you bet. This page collects those figures in one place — ranked, translated into cost per $100 wagered, and, where we ran our own 44-million-bet simulation study, compared against what we actually measured. Every number links to the page where it comes from.

The ranked table: every major bet by house edge

Rank	Game / bet	House edge	Cost per $100 wagered	Basis
1	Blackjack, basic strategy (8 decks, S17, DAS, 3:2)	0.44%	$0.44	Measured, 8M hands
2	Baccarat — banker (after 5% commission)	1.04%	$1.04	Measured, 10M hands (published ~1.06%)
3	Crypto-casino originals (Plinko, Wheel, Dice and similar)	~1%	~$1	Measured, 24M rounds (RTP 97.99–99.53%)
4	Baccarat — player	1.26%	$1.26	Measured, 10M hands (published ~1.24%)
5	French roulette — even money with la partage	1.35%	$1.35	Analytic
6	Craps — don’t pass	1.36%	$1.36	Analytic
7	Craps — pass line	1.41%	$1.41	Analytic
8	European roulette — any bet	2.70%	$2.70	Analytic
9	Online slots (typical published RTP 94–97%)	3–6%	$3–6	Published RTP
10	American roulette — any bet	5.26%	$5.26	Analytic
11	Blackjack — “mimic the dealer”	5.68%	$5.68	Measured, 1M hands
12	Blackjack — “never bust”	5.94%	$5.94	Measured, 1M hands
13	Baccarat — tie at 8:1	14.29%	$14.29	Measured, 10M hands (published ~14.36%)
14	Craps — Any 7	16.67%	$16.67	Analytic

The span is a factor of 38 between the best and worst rows. Nothing else you control at a casino — bonuses included — moves your expected cost as much as which bet you place.

Measured vs published: does the maths hold up?

Published house edges deserve independent checking, so we checked. Our data study ran 44 million simulated bets with fixed seeds and exact published rules:

Game	Published	We measured	Sample
Baccarat banker	~1.06%	1.04%	10,000,000 hands
Baccarat player	~1.24%	1.26%	10,000,000 hands
Baccarat tie (8:1)	~14.36%	14.29%	10,000,000 hands
Plinko (all 9 configs)	~1% (RTP ≈ 99%)	0.97–2.01%	9,000,000 drops
Wheel (all 15 configs)	1.00% exactly	0.47–1.47%	15,000,000 spins
Blackjack, basic strategy	~0.43–0.45%	0.44%	8,000,000 hands

Every result landed within normal sampling variance of the published value. The visible wobble on high-variance configurations (Plinko high-risk, Wheel high-risk) is what extreme payout tables do to million-round samples — not a deviation from the maths.

Strategy changes the edge in exactly one game

Roulette, baccarat, slots, Plinko and Wheel have fixed odds: no system, progression or “due number” changes the expected loss. Blackjack is the exception, because you make decisions mid-hand — and the price of deciding badly is measurable:

How you play blackjack	House edge	Per $1,000 of bets
Perfect basic strategy	0.44%	$4.36
”Mimic the dealer”	5.68%	$56.78
”Never bust” (stand on hard 12+)	5.94%	$59.35

Playing by feel multiplies the cost of the same table by roughly 13. Note also that a 6:5 blackjack payout adds over a full percentage point to the house edge — a rule change that silently erases most of basic strategy’s benefit.

Rare-event frequencies, measured

Long odds are abstract; measured counts are not. From the same study:

• Plinko’s 1000× slot (high risk, 16 rows): 31 hits in 1,000,000 drops — one per 32,258 drops for either outer slot, matching the binomial 1-in-65,536 per side. At one drop per second, that is about nine hours per hit.
• Wheel jackpots: every high-risk jackpot landed within 0.5% of its theoretical 1-in-N frequency across 5 million high-risk spins. No segment was ever “due”.
• A dealt blackjack natural: 4.743% of 8-deck hands (theory: 4.745%) — about one hand in 21.
• Baccarat ties: 9.52% of hands, yet the 8:1 payout still leaves a 14.29% edge — the worst mainstream bet in the casino short of craps props.

Using these statistics

Journalists and researchers are welcome to reuse any figure on this page with attribution and a link. Measured figures come from single fixed-seed runs, reported in full — no cherry-picking — and the methodology contains everything needed to reproduce them: seeds, rules, payout tables and sample sizes.

For players, the practical reading order is: pick the game from the top of the ranked table, learn the one strategy that matters if it’s blackjack, treat bonuses as discounts with conditions rather than free money, and understand the house edge itself so no streak — hot or cold — surprises you. Play only what you can afford to lose, where it is legal for you.

Bonuses interact with these edges in a way the headline offer hides: at 40x wagering the same bonus is a 30.8% loss on 4%-edge slots and a 22.8% gain on a 1%-edge even-money game. We measured it across 1.3 million simulated claims in the bonus study.

And no staking pattern moves these numbers: we simulated 400,000 sessions of Martingale, D’Alembert, Fibonacci and flat betting, and every one lost the house edge per dollar wagered — see the betting-systems study.

RTP alone does not tell you how long your money survives, either: three models at identical 96% RTP produced median sessions of 2,315, 745 and 147 spins purely from volatility — see how long $100 lasts on slots.