Craps Probability and House Edge Explained
Why Craps Probability and House Edge Matter
Craps probability house edge isn’t abstract math—it’s the direct calculation of how much money you’re expected to lose per bet over time. Every single bet on the craps table has a house edge baked in, and that percentage determines your expected value loss per dollar wagered. Probability Analysis enables informed bet selection across the entire craps table, meaning players who understand the numbers make fundamentally different decisions than those guessing. Variance Awareness is equally critical—without it, players misread lucky runs as skill and increase bet sizes at exactly the wrong moment. If you’re playing an online craps game, these fundamentals apply just as much as at a physical casino.
How Dice Probability Works in Craps
Two six-sided dice produce 36 possible combinations. Dice probability drives house edge calculations for every craps bet, so understanding the distribution is non-negotiable.
- 7 appears 6 times out of 36 (16.67% probability)
- 6 and 8 each appear 5 times (13.89% each)
- 5 and 9 each appear 4 times (11.11% each)
- 4 and 10 each appear 3 times (8.33% each)
- 2 and 12 each appear just once (2.78% each)
Probability Analysis of these frequencies reveals why certain bets carry higher house edges—casinos pay less than true odds on outcomes, and the gap between true probability and casino payout is where the house edge lives. Previous rolls have zero effect on future outcomes. The dice don’t have memory.
Pass Line and Don’t Pass Craps House Edge
| Bet | House Edge | Notes |
|---|---|---|
| Pass Line | 1.41% | Best bet for beginners |
| Don’t Pass | 1.36% | Slightly better value |
Pass Line bets have a 1.41% house edge, making them among the most favorable wagers in any casino game, not just craps. Don’t Pass probability creates nearly equal value to Pass Line betting—the 0.05% difference comes from the come-out roll rules, where the 12 is a push rather than a win for Don’t Pass players. Both bets use near-true odds on point rolls. For a deeper breakdown of the mechanics, pass line vs don’t pass bets explained covers the structural differences clearly.
Come and Don’t Come Bet House Edge
Come bets duplicate the Pass Line house edge of 1.41% but apply after a point is established. Don’t Come holds at 1.36%. Mechanically, they’re identical to Pass/Don’t Pass—just triggered at a different phase of the roll sequence.
Session Optimization involves layering Come bets during active point play to spread risk across multiple numbers simultaneously. Instead of one bet riding on a single point, you can have three numbers working at once. This doesn’t change the house edge percentages on individual bets, but it does change your exposure and variance profile during a session.
Field and Place Bets House Edge Breakdown
| Bet | House Edge | Verdict |
|---|---|---|
| Place 6 or 8 | 1.52% | Solid value |
| Field (12 pays 3:1) | 2.78% | Marginal |
| Field (12 pays 2:1) | 5.56% | Avoid |
| Place 5 or 9 | 4.00% | Poor value |
| Place 4 or 10 | 6.67% | Avoid |
| Any Hardway | 9.09–16.67% | Never |
Place 6 and 8 bets have a 1.52% house edge, making them genuinely playable proposition wagers. The expected value calculation on Place 5 and 9 bets reveals a 4% edge—nearly three times worse than Place 6/8 for no compelling reason. Hardways are the worst bets on the table. Anyone suggesting hardways as a “fun side bet” is costing you real money. For help applying craps pattern recognition to understand when numbers are hitting, context matters—but the house edge on these bets never changes.
Taking and Laying Odds Reduces House Edge
Odds bets have zero house edge and pay true probability outcomes—they’re the only bets in the casino with no mathematical disadvantage. Taking odds on Pass/Come bets reduces your combined house edge below 0.8% when using 3x-4x-5x odds, which is what most Australian online casinos offer. Laying odds on Don’t Pass/Don’t Come pushes that figure even lower. The full mechanics are worth mastering through a dedicated guide on odds betting in craps.
Calculating Expected Value for Any Craps Bet
Expected value calculation reveals the mathematical advantage or disadvantage of each bet by multiplying your win probability by the payout, then subtracting the loss probability multiplied by your stake. The result is your average gain or loss per bet. For a Pass Line bet at $10: you’re losing an average of $0.141 per roll. That’s the 1.41% house edge in dollar terms. Negative expected value demonstrates how house edge accumulates over sessions—small percentages compound into significant losses across hundreds of bets. Only odds bets achieve zero expected value. Everything else favors the house. Pairing this with a solid craps betting strategy helps you apply these numbers practically.
Which Craps Bets Offer the Best Value
- Pass Line / Don’t Pass (1.36–1.41%): Your foundation. Always start here.
- Come / Don’t Come with odds: Matches Pass Line value; expands your coverage
- Place 6 and 8: Acceptable at 1.52% if you want numbers working between come-out rolls
- Maximum odds on any of the above: Drops effective house edge below 0.7%
- Field bets: Situational at best; avoid unless the casino pays 3:1 on 12
- Proposition bets and hardways: Never. The 9–16% house edge is indefensible
Probability Analysis consistently points to the same conclusion—Pass Line with maximum odds provides the best expected value for craps players.
Managing Variance While Understanding House Edge
Variance Awareness acknowledges a crucial reality: short-term luck operates independently of mathematically negative expected value. You can win 10 sessions in a row playing sub-optimal bets, and lose 5 in a row playing perfect strategy. That’s variance, not a flaw in the math. Session Optimization combines low house edge bet selection with proper bankroll management—sizing your bets so variance swings don’t wipe you out before the math has time to stabilize. A proper approach to craps bankroll management accounts for this directly. Build your session around Pass Line, Don’t Pass, and maximum odds. Ignore the rest.
