Craps Myths: 10 Claims Checked Against the Math
Ten claims · Dice math · Systems · Casino fairness
Craps myths survive because random games regularly produce results that look meaningful. A streak can be real without being predictive, and a bet can win without being a good value. Here are ten common claims checked against the rules and the math.
The fast answer
No ritual, progression, hedge, or recent sequence changes the 36 combinations of two fair dice. Players can improve decisions by choosing lower-edge bets and limiting exposure, but ordinary casino craps remains a negative-expectation game.
1. “Seven is due after a long absence”
False. Seven has six combinations out of 36 on every independent roll. A 20-roll stretch without 7 is unusual, but once it has happened, roll 21 still has a 1/6 chance of 7. The dice do not compensate for a past shortage.
2. “A hot shooter is more likely to win the next roll”
False as a prediction. “Hot” accurately describes a long hand that has already occurred. It does not add information about the next independent throw. Increasing bets on a hot hand increases dollars exposed; it does not improve the dice.
3. “The Don’t player makes seven appear”
False. A wager cannot change a physical dice result. Betting Don’t Pass may conflict with the table’s social mood, but it does not jinx the shooter. Etiquette means avoiding celebration when others lose—not pretending chips influence probability.
4. “A betting system changes the house edge”
False. Martingale, regression, pressing, Iron Cross, and named systems change bet size or combine wagers. They can change volatility and the frequency of small wins, but each component keeps its own expected value. Table limits and finite bankrolls also prevent unlimited progression.
5. “Hedging protects a good bet for free”
False. An Any Craps bet can soften a come-out Pass loss on 2, 3, or 12, but it loses on the other 32 combinations and carries an 11.11% edge at a 7-to-1 payout. A hedge may narrow one loss while adding expected cost elsewhere.
6. “Free odds makes the whole Pass bet a zero-edge wager”
Only partly true. The odds portion pays at true odds and has zero house edge. The required flat Pass Line portion still carries about a 1.414% edge. Adding odds reduces the house edge as a percentage of the total money placed, but increases dollars at risk.
7. “A high payout means a good bet”
False. Payout must be compared with probability. Any 7 wins more often than many proposition bets but commonly pays only 4 to 1 despite true odds of 5 to 1, producing a 16.67% house edge. Large signs and exciting payouts do not reveal value by themselves.
8. “Dice control has been proved to beat casino craps”
Not established. Setting dice is observable; sustaining a profitable change in outcomes under rule-compliant casino conditions is a separate claim. Stewart N. Ethier’s 2025 paper, “Testing for Dice Control at Craps”, explains how shooter data could be tested but explicitly uses no actual shooter data and takes no position on efficacy.
9. “The casino must rig the dice to win”
False for a properly regulated game. Standard payouts already give the house an advantage on the flat bets. A licensed casino risks its license and reputation by cheating when the published game math already produces revenue. That does not mean players should trust an unlicensed room or an unverifiable app; jurisdiction and game rules matter.
10. “A stop-win locks in a long-term profit”
False. Leaving while ahead preserves that session’s result. It does not change expected value across repeated sessions. Limits are useful harm-reduction tools because they restrict time and money exposed—not because they defeat probability.
| Claim test | Evidence to request |
|---|---|
| “This system wins 80% of sessions.” | Average win, average loss, all sessions, total amount wagered, and table limits |
| “This shooter avoids sevens.” | Complete preregistered sample under rule-compliant conditions |
| “This hedge cannot lose.” | Every resolving outcome, payout, commission, and combined expected value |
| “This game is fair.” | Licensed jurisdiction, approved rules, dispute process, and current paytable |
Sam’s rule: if a claim relies on a winning story, ask for the denominator. How many rolls, sessions, and dollars were involved—including the failures?
See the deeper guides to dice control evidence, hedging craps bets, the gambler’s fallacy, and house edge.
Run a claim through four gates
- Define it: rewrite the claim so its trigger, wager, stopping point, and promised result are measurable.
- Price it: list every stake, payout, commission, push, and loss condition.
- Test it: use all eligible trials under a rule fixed before the results are known.
- Compare it: measure the claim against the same wagers made without its prediction or progression rule.
For “press after two wins because the table is hot,” define which two wins qualify and the exact press. Calculate the larger exposure, record every qualifying sequence, and compare it with keeping the original stake. A few successful presses show that the payoff can occur; they do not show that the hot-table explanation selected it better than chance.
A claim that changes after a loss has failed the definition gate. A claim that reports only wins has failed the test gate. A claim that ignores total dollars wagered has failed the price gate. These failures can be identified before arguing about whether the story sounds plausible.
