Craps Sense for the Compulsive Tinkerer
A clear, practical look at craps probabilities, sensible bets, and how house edge, hit frequency, variance, and expected value shape your session.
If you like taking things apart to see how they work, craps is a delight: two dice, 36 equally likely outcomes, and a table full of choices whose math is simple once you stop being dazzled by the lights. This post gives you the useful probabilities, the bets worth your attention, and how to think in terms of house edge, hit frequency, variance, and expected value so your decisions are deliberate instead of emotional.
The dice basics (the real foundation)
Two dice produce 36 equally likely combinations. The totals occur with these frequencies:
- 2: 1/36
- 3: 2/36
- 4: 3/36
- 5: 4/36
- 6: 5/36
- 7: 6/36
- 8: 5/36
- 9: 4/36
- 10: 3/36
- 11: 2/36
- 12: 1/36
That 7 is the most common roll is the single most important fact in craps math. Everything else—payouts, risk, and frequency—flows from those counts.
What the technical terms mean (so you can speak fluent table-talk)
- House edge: the casino’s long-run percentage advantage on a particular bet. If a bet has a 1.41% house edge, you can expect to lose $1.41 per $100 wagered on average over many, many repetitions.
- Hit frequency: how often a bet wins (for example, a bet that wins when a 6 shows up has a hit frequency of 5/36 on any single roll).
- Variance (volatility): how wildly your results swing in the short term. High-variance bets win rarely but pay big; low-variance bets win more often with smaller payouts.
- Expected value (EV): the average win or loss on a bet over the long run, combining payouts and probabilities. EV per dollar = (probability of win × payout) − (probability of loss × amount lost).
Distinguishing these matters because a bet can win often (high hit frequency) but still have a poor EV (high house edge), or be a low-edge bet but volatile.
Bets that make sense for the tinkerer
-
Pass Line and Come: These are the classic starter bets. The pass line has a house edge around 1.41%. They win immediately on a 7 or 11 and lose on 2, 3, or 12 before a point is established; once a point is set, the shooter must roll the point again before a 7 for the bet to win.
-
Don’t Pass / Don’t Come: Slightly better edge than pass/come in most rule sets (~1.36%); these are the “betting against the shooter” options and behave oppositely. They can feel sour to play at a lively table, but math‑minded players like their slight edge.
-
Free Odds behind the Pass/Come: This is your best friend. After a point, you can place an additional odds bet that pays true mathematical odds (no house edge on that part). Because the odds portion has EV = 0, taking as much odds as the table allows lowers your overall house edge proportionally.
-
Place 6 and 8: If you prefer action on every roll, placing the 6 or 8 is sensible. They hit reasonably often (5/36 per roll) and have a modest house edge (about 1.52%). They combine decent hit frequency with lowish edge—good for players who want regular wins.
Avoid proposition bets (large one-roll wagers) if you care about long‑run math: they pay big when they hit but typically carry very high house edges—double-digit percentages are common—so they’re high variance and poor EV for most players.
Simple strategy principles (no systems promised)
- Favor bets with low house edge and reasonable hit frequency: pass/come with full odds and place 6/8 are a compact, sensible mix.
- Use odds to reduce the effective house edge: the more odds you can legally stake behind your pass/come bet, the closer your combined EV gets to neutral on that portion.
- Avoid high-house-edge prop bets unless you want a thrill and accept the likely loss over time.
- Flat-bet sizing beats complicated progressions: increasing bet sizes after wins or losses does not change EV and usually increases the chance of uncomfortable downswings.
A quick EV illustration
If a bet has a 1.41% house edge, then every $100 wagered on average loses $1.41 in the long run. That’s EV expressed as expected loss per dollar wagered—simple, robust, and why many players talk in house edge terms rather than “hit streaks.” Remember: EV tells you long-run averages; variance determines whether your session looks terrible or terrific.
Using tools without mistaking them for magic
Tink’s Toolchest can help you quickly reference strategy, compare bets, or log and review a session so you learn what worked for your style—without implying it predicts results or changes the odds.
Final practical takeaways
- Learn the dice probabilities and use them to choose bets that match your taste for action vs. risk.
- Take full odds when you can; avoid the flashy proposition bets if you care about minimizing expected loss.
- Manage your bankroll with flat staking and session limits so variance doesn’t surprise you.
Responsible-play reality check: the math here shows which bets are better or worse over time, but no strategy eliminates the house edge—play for fun, set limits, and never wager money you need.