What 'House Edge' Actually Means (and Why You Should Know Yours)

House edge is the most important number in gambling, and the one most players never check. Here's what it means, how it's calculated, and which games have hidden edges most players miss.

What 'House Edge' Actually Means (and Why You Should Know Yours)

The house edge is the single most important number in casino gambling. It's the difference between "I lost on a string of bad luck" and "I lost because the math said I would." And yet most players never check it before they bet — and most operators don't make it easy to find.

This piece is the basic-but-honest explanation. What house edge is, how it's calculated, which games have surprisingly high edges most players miss, and how to use the number once you know it.

The one-sentence definition

House edge = the percentage of every dollar wagered that the casino expects to keep, over a large number of bets.

If a game has a 2% house edge, then on average for every $100 you wager, you'll get back $98. The other $2 is the casino's expected long-run revenue from your play.

A few clarifications hidden inside that sentence:

• "Over a large number of bets" — house edge is a long-run statistical average. Any single bet can win or lose. Over hundreds of bets, your results will gravitate toward the expected return.
• "Of every dollar wagered" — note this is wagered, not deposited. If you deposit $100 and then play a $5 bet 50 times, you've wagered $250. The edge applies to that $250, not the $100 deposit.
• "On average" — house edge is the expectation, not the certainty. Variance can push your actual outcomes far above or below the expected value, especially over short sessions.

How it's calculated

For a game with discrete outcomes (like roulette, where there are 38 numbers on an American wheel):

House edge = (Total probability of operator winning × payout to operator)
           - (Total probability of player winning × payout to player)

For a single-number bet on American roulette:

• Probability player wins = 1/38
• If player wins, they get 36× their bet (35 profit + 1 stake back)
• Probability player loses = 37/38
• If player loses, they lose 1× their bet

Expected value = (1/38 × +35) + (37/38 × -1)
              = 0.921 - 0.974
              = -0.053

So the house edge on American roulette is 5.3%. For every dollar bet, the player expects to lose 5.3 cents on average.

On European roulette (single zero, 37 numbers instead of 38), the same calculation gives 2.7%. The difference between American and European roulette is one extra zero, and that one zero almost doubles the house edge.

Common games and their house edges

The good (under 2%):

• Blackjack (basic strategy): ~0.5%. The lowest-edge mainstream game.
• Baccarat (Banker bet): 1.06%. Slightly higher than blackjack but easier — no decisions.
• Plinko (depending on settings): 1-3%.
• Video poker (Jacks or Better, full pay): 0.5%.
• Backgammon (peer-to-peer with platform fee): the platform fee is the "edge", typically 5-10% of the rake, not the bet. Very low effective edge for skilled players.

The middle (2-5%):

• European Roulette: 2.7%.
• Craps (pass line + odds): 1.4% on pass, 0% on odds.
• Crash (typical implementations): 1-2%.

The high (5%+):

• American Roulette: 5.3%.
• Slots: 2-15% depending on the slot. Most popular slots: 4-8%.
• Keno: 25-35% (one of the worst).
• Sic Bo single-number bet: 7.87%.
• Lottery / scratch cards: 30-50% — the casino's gift to brand recognition.

The variance between games of "similar" type is huge. A keno spin is not "kind of like" a blackjack hand. They're an order of magnitude apart.

The edges most players miss

A few games where the house edge is sneakier than it looks:

Slot side-bets. A 4% slot turns into a 12-15% slot the moment you opt into the bonus side-bet feature. The bonus reels often have an edge twice the headline rate.

Insurance in blackjack. Insurance is a separate bet with a ~7% house edge. Even when offered with a "you have blackjack" hint, the insurance bet itself is bad math. Decline it.

Roulette "outside" bets. A bet on "red/black" feels safer than a single number — and the win frequency is higher — but the house edge is identical. Roulette has the same edge on every bet on the board; the variance differs.

Multi-game progressive jackpots. Operators sell these as "potentially infinite return" — and on rare occasions, they are. The effective edge during normal play is typically much higher than the headline rate, because the operator skims a percentage of every bet to fund the jackpot.

"Free" bonuses. Welcome bonuses come with wagering requirements (typically 30-50× the bonus). Playing through those requirements at the games allowed (usually only slots) means the effective edge on your bonus play is the slot edge × 30-50× compounded. The "free" money is rarely actually free.

How to use the number

Once you know the house edge of the game you're about to play, three things become useful:

1. Estimate your expected cost per session.

If you plan to play 200 hands of blackjack at $5 each, that's $1,000 wagered. At 0.5% edge, expected loss = $5. At 4% slot edge, the same 200 spins of $5 each = $40 expected loss. Both are reasonable entertainment costs for an evening, but they're 8× different.

2. Compare games on a fair basis.

A $1 minimum slot bet at 6% edge has a per-spin expected loss of $0.06. A $5 minimum blackjack hand at 0.5% edge has a per-hand expected loss of $0.025 — less than half. The blackjack stakes feel higher, but the expected cost per round is lower.

3. Notice when an operator hides the number.

Casinos that publish their full edge tables on their fair-play page are giving you the data you need. Casinos that don't, aren't. This is one of the cleanest filters for serious operators vs not.

The variance trap

The honest catch: house edge is the long-run expectation. Over a single session, variance — the noise around the expected value — dominates the outcome.

A 0.5% edge game can lose you $500 in a session. A 30% edge game can win you $500 in a session. Variance is bigger than edge over short timeframes. This is the source of every "I won big on slots" anecdote: short-term lucky outcomes that mask the long-term underlying tilt.

What edge controls is the direction of drift over many sessions. If you play 1,000 hours of slots at 6% edge, you will lose money — there's no statistical universe where you don't. If you play 1,000 hours of blackjack at 0.5% with correct basic strategy, you might break even, you might win, you might lose modestly. The math is dramatically kinder.

On provably fair operators

A house edge is meaningful only if you can verify the operator is actually delivering it. A casino can claim "1% edge" while running a rigged RNG that's secretly 5% — and you'd never know without forensic statistics over thousands of rounds.

Provably fair systems close that loophole. The cryptographic verification proves each round was honest. Combined with publicly-published edge tables, you get a complete picture: the published math is real, AND each round delivered against it. That's the standard we apply across every game on 6proclub.

In one paragraph

House edge is the percentage of every dollar wagered the casino expects to keep over a large number of bets. Mainstream games range from 0.5% (blackjack basic strategy) to 30%+ (keno). Edge controls long-run drift; variance dominates short sessions. The two things you should do: check the edge before you play, and only use operators that both publish the number AND let you cryptographically verify they're delivering it. Both are easy. Most players do neither.

Further reading

• Provably Fair Complete Guide — verify the edge
• House Edge vs Peer-to-Peer Play — backgammon's structurally different
• Plinko Probability Explained — the math behind one specific edge