Craps Strategy Guide
What this guide covers
Craps has attracted more named betting strategies than almost any other casino game, from the simple (max odds) to the elaborate (the Iron Cross). This guide covers the most popular approaches honestly — what each one actually does, and why none of them change the fixed house edge built into the game's bets. For the underlying math these strategies operate on top of, see craps house edge explained; for the individual bet mechanics referenced throughout, see craps betting guide.
The one strategy that's genuinely mathematically sound: maximizing odds
Before covering the named betting systems, it's worth stating the one piece of craps strategy that isn't a matter of opinion: maximizing your odds bet relative to your pass line or come bet is the single most impactful thing you can do to lower your overall house edge, since odds carry zero house edge on their own. A pass line bet alone sits at 1.41%; pass line plus 5x odds blends down to roughly 0.33% across your total wagered amount. If you take away only one piece of strategy from this entire guide, it's this: always take maximum odds when your bankroll and the table's rules allow it.
The 3-point Molly strategy
The 3-point Molly (sometimes spelled "Molly" or "Molley") involves establishing a pass line bet with full odds, then adding come bets — also backed with full odds once each establishes its own point — until you have three total numbers actively working simultaneously (the original pass line point plus two come bets). The goal is broad coverage: with three numbers working, a 7 becomes relatively more likely to hit one of your covered numbers before it appears and ends your action, compared to having just a single point working.
Why the 3-point Molly doesn't beat the house edge
Despite its structured, almost systematic feel, the 3-point Molly is built entirely from bets already covered in craps betting guide — pass line, come, and odds — each carrying exactly the same house edge whether placed as part of this named strategy or individually without a system attached. Running three simultaneous come-bet-with-odds combinations doesn't create any new mathematical advantage; it simply distributes the same fixed-edge bets across more numbers at once, increasing your total action (and therefore your total expected cost) rather than improving your per-dollar value.
What the 3-point Molly actually changes
What this strategy genuinely does change is your session's variance and pacing: with three numbers working, you're more actively engaged with more of the table's rolls, and a single 7 that would otherwise just end one bet's point now potentially resolves three simultaneous bets at once. This can feel more exciting and involved than a single pass-line-and-odds bet, but it also means your total wagered amount — and correspondingly your total expected loss over a session — is considerably higher for the same "one bet" mental framing.
The Iron Cross strategy
The Iron Cross combines a field bet with place bets on 5, 6, and 8, structured so that every number except 7 is covered by at least one active bet — a 7 is the only roll that loses your entire combined wager under this structure, while every other possible roll wins at least one component. This is a genuinely clever combinatorial design, and it's popular specifically because it minimizes the "no result" rolls where nothing happens either way.
Why the Iron Cross doesn't beat the house edge either
Despite covering nearly every number, the Iron Cross combines bets with meaningfully different house edges — the field bet (2.7%-5.5%) and place bets on 5, 6, and 8 (4.0%, 1.52%, 1.52% respectively) — into a single structure, and the blended house edge across the whole combination lands somewhere in that range, not at zero. You're simply betting on more outcomes simultaneously using the same fixed-edge bets already available individually; the structure doesn't create new value, it just repackages existing bets into a wider-covering combination.
The real trade-off with the Iron Cross
The Iron Cross's appeal is genuinely about the psychological experience of winning something on nearly every roll rather than about superior expected value — losing only on a 7 (which happens on average once every six rolls) while winning something on every other roll creates a specific, pleasant win-frequency pattern. This is a legitimate reason to enjoy the strategy, but it should be understood as a variance and experience choice, not a value optimization, since the combined house edge is worse than simply running pass line plus maximum odds.
The "don't" side strategy
Betting the don't side exclusively (don't pass and don't come, backed with odds where possible) offers a very slightly better house edge than the equivalent "do" side bets — 1.36% versus 1.41% on the line bets specifically — at the social cost of betting against the shooter and the table's shared excitement, which some players find genuinely uncomfortable at a lively physical table. See pass line vs don't pass bets for the fuller mechanics and house-edge comparison behind this specific strategic choice.
Why the house edge difference between do and don't is nearly negligible
The 0.05 percentage point difference between pass line (1.41%) and don't pass (1.36%) is genuinely tiny in practical terms — over a typical session, this difference amounts to a few cents of expected value difference per $100 wagered. Choosing don't-side play purely for this marginal edge, while ignoring the genuine social and atmospheric trade-off (particularly at a physical or lively live-dealer table), is arguably prioritizing the wrong factor; if you're indifferent to betting against the table, the edge difference is a nice bonus, not a compelling reason on its own.
Place betting strategies: covering specific numbers
Some players build a strategy entirely around place betting their preferred numbers (commonly 6 and 8, given their favorable 1.52% house edge) rather than using the pass line at all, letting them enter and exit specific number coverage at will without waiting for a new come-out roll. This offers genuine flexibility — place bets can be added or removed at any time — at the cost of a slightly higher house edge than pass line plus odds, and the inability to add the zero-edge odds bet the way pass line and come bets allow.
Combining place bets with line bets
A common structured approach uses pass line plus odds as the "core" bet, with place bets on 6 and 8 added as a supplementary layer for additional coverage without waiting for a come bet to establish its own point. This blends the excellent value of the line-and-odds foundation with the flexibility and immediacy of place betting, at a modest overall house-edge cost compared to a pure line-and-odds-only approach.
The "dice control" or "rhythm rolling" claim
Some strategy content promotes "dice control" or "rhythm rolling" — the claim that a sufficiently practiced shooter can influence dice outcomes through a specific, consistent throwing technique, subtly reducing the probability of a 7 appearing. This claim isn't supported by rigorous, independently verified evidence, and casino table rules requiring dice to hit the back wall exist specifically to disrupt exactly this kind of attempted control. See can you beat craps? for a full, honest treatment of this claim and what legitimate historical advantage play in dice games has actually looked like.
Why dice control claims don't apply to online play at all
Even setting aside skepticism about dice control at a physical table, the claim is entirely inapplicable to RNG online craps, where there's no physical dice to throw with any particular technique, and to bubble craps, where a mechanical ball-and-air system generates results independent of any human throwing motion. Live dealer craps online similarly removes any player physical involvement in the actual roll. This is one of several reasons online craps, across formats, offers a particularly clean environment for understanding that outcomes are governed by fixed probability, not throwing technique.
Progressive betting systems applied to craps
Martingale, Fibonacci, and other progression systems developed originally for even-money roulette bets are sometimes applied to craps' even-money line bets by players familiar with those systems. The same fundamental math applies here as anywhere else: adjusting bet size based on previous results changes how wins and losses are distributed across a session without changing the underlying house edge of any individual bet. A progression system applied to a 1.41% pass line bet still operates on a 1.41%-edge bet at every step — the sequencing doesn't create value that wasn't already there.
Bankroll-based strategy: session budgeting as strategy
Arguably the most impactful "strategy" available to any craps player isn't about which bets to place at all — it's disciplined bankroll and session management: setting a budget in advance, sizing individual bets as a small percentage of that budget, and having predetermined stop-loss and win-target rules. See craps bankroll management guide for the full framework — this genuinely affects your session experience and risk of ruin in ways that bet-selection strategy alone doesn't.
Why no strategy overcomes the house edge
It's worth stating this plainly and directly: every bet at a craps table carries a fixed house edge (except odds, which carries none but is only available as a supplement to an edge-carrying bet), and no combination, sequence, or structure of these bets changes that fixed edge. The 3-point Molly, Iron Cross, don't-side play, and every other named strategy covered in this guide are all built entirely from the same bets covered in craps betting guide — they differ in variance, session pacing, and psychological experience, not in underlying expected value beyond what maximizing odds already provides.
Comparing strategies by what they actually optimize for
| Strategy | What it optimizes for | Underlying house edge |
|---|---|---|
| Pass line + max odds | Lowest possible cost | ~0.3%-0.8% blended, depending on odds multiple |
| 3-point Molly | Broad simultaneous coverage | Same as underlying bets, higher total action |
| Iron Cross | Winning on nearly every roll | 2.5%-4% blended, depending on table |
| Don't-side play | Marginally better base edge | 1.36% (line only), improved further with odds |
| Place 6 & 8 only | Flexibility without a point wait | 1.52% |
This table is the honest summary of craps strategy: nothing beats pass line plus maximum odds on pure cost, and everything else is a legitimate trade-off between that baseline and some other quality (coverage, win frequency, flexibility, or don't-side preference) you might personally value.
Building your own approach
Given that no strategy provides a mathematical edge beyond maximizing odds, the most useful exercise isn't finding the "best" system — it's honestly identifying what you actually want from a session (lowest cost, broadest coverage, frequent small wins, or occasional big payouts) and selecting bets that suit that goal, informed by the house-edge figures throughout this guide and craps betting guide. A player who understands this trade-off structure is better equipped than one chasing a system marketed as "beating" the game.
Adjusting your strategy for online formats specifically
RNG online craps' self-paced betting makes strategies requiring careful sequential tracking (like the 3-point Molly's multiple simultaneous come bets) considerably easier to manage than at a fast-moving physical table, since there's no risk of a rushed dealer interaction or a missed betting window. Live dealer craps online splits the difference — real-time pacing with the atmosphere of a physical table, but without needing to physically reach across a layout to place bets in specific positions. Consider which format suits your chosen strategy's complexity; see online craps vs live craps for the fuller comparison.
Session-length considerations for different strategies
Strategies involving more simultaneous bets (3-point Molly, Iron Cross) naturally involve more total money wagered per unit of time than a simple pass-line-only approach, which means your bankroll needs to be sized accordingly for the same intended session length. A player switching from pass-line-only to a full 3-point Molly without adjusting their bankroll plan may find their session ending considerably faster than expected, not because of bad luck specifically, but because of a straightforward increase in total action per round. See craps bankroll management guide for sizing your bankroll against your chosen strategy's total action.
Why understanding strategy trade-offs matters more than picking a "winning" system
The single most valuable takeaway from studying craps strategy isn't a specific system to adopt — it's the clear understanding that every approach reduces to the same underlying fixed-edge bets, differing only in how those bets are combined and sequenced. This understanding protects you from strategy marketing that implies a named system provides a mathematical edge it doesn't actually have, while still leaving plenty of room to choose an approach that genuinely suits how you want to experience the game.
The "hedge betting" temptation
New players sometimes discover that betting both pass line and don't pass simultaneously seems to "guarantee" a near-even outcome on the come-out roll, since one side wins whenever the other loses (except on the barred number, where don't pass pushes while pass line loses). This isn't a genuine edge or a clever hedge — you're simply paying the same combined house edge on two offsetting bets simultaneously, converting what would be one bet's expected loss into a smaller, but still present, expected loss spread across two bets. It doesn't create value; it mostly just reduces variance while keeping (or slightly increasing) your total action.
Why "hedging" pass line with any craps doesn't work either
A related idea involves betting pass line and simultaneously placing a small "any craps" bet specifically to offset the loss when a come-out roll lands on 2, 3, or 12. While this does soften that specific losing outcome, "any craps" carries an 11% house edge — considerably worse than pass line's 1.41% — so the hedge costs you more in expected value than it protects, especially once repeated across many come-out rolls over a session. This is a clear example of a strategy that feels protective in the moment but is mathematically worse than simply accepting pass line's occasional craps-roll loss.
Betting progressions specific to place bets
Some players apply a "regression" approach to place betting — starting with larger place bets on 6 and 8, then reducing ("regressing") the bet size after an initial win, aiming to lock in profit from the first hit while reducing future exposure. Like every other progression discussed in this guide, this reshapes the variance and timing of wins and losses without changing the underlying 1.52% house edge each individual place bet carries — it's a legitimate personal risk-management choice, not a value improvement.
The "press" strategy on place bets
The inverse of regression is "pressing" — increasing a place bet's size after a win, using profits to fund a larger subsequent bet on the same number, aiming to compound gains during a hot streak. This carries the same fundamental math as regression in reverse: it reshapes when and how much you might win or lose without changing the fixed house edge underlying each individual place bet. Pressing increases both your potential upside and your exposure to giving back accumulated winnings on a subsequent loss.
Strategy considerations for shorter versus longer sessions
If you're planning a short session (an hour or less), lower-variance approaches like pass line plus modest odds let you play a predictable number of rounds within a stable budget. If you're planning a longer session and want to stay engaged throughout, broader-coverage approaches like the 3-point Molly or Iron Cross — despite their slightly higher blended house edge — can make a longer stretch of play feel more consistently active, since something is resolving on a larger share of rolls. Matching your strategy's engagement pattern to your intended session length is a genuinely useful, non-mathematical consideration.
Why some players deliberately choose higher-house-edge bets
Not every rational craps strategy is about minimizing house edge — some players genuinely value the specific excitement of a large payout from a hardway or proposition bet enough to accept its worse math as the "cost" of that experience, similar to how someone might buy a lottery ticket knowing the expected value is negative. This is a legitimate personal choice as long as it's made with clear-eyed understanding of the actual cost involved, which is exactly what the house-edge figures throughout craps betting guide are meant to provide.
Strategy and table selection
Beyond bet selection, which specific table you choose affects your strategy's practical execution — a table's maximum odds multiple directly caps how far you can push the "maximize odds" principle, and table minimums affect whether strategies requiring multiple simultaneous bets (like the 3-point Molly) remain affordable within your bankroll. Comparing a few tables' posted odds multiples and minimums before settling in for a session is a small amount of effort that can meaningfully affect which strategies are practically available to you.
Strategy differences between crapless craps and standard craps
Because crapless craps explained removes the traditional craps-numbers-lose rule from the come-out roll, the relative value of different point numbers shifts, and standard strategies built around standard craps' point-number probabilities (like heavily favoring place bets on 6 and 8) need to be reconsidered for this variant's different math. If you're playing crapless craps specifically, don't assume standard-craps strategy conclusions transfer over unchanged.
Testing a strategy with free play first
Before committing a real-money bankroll to a more complex approach like the 3-point Molly or Iron Cross, running through several practice rounds in a free-play RNG demo — tracking how many simultaneous bets you're managing and how quickly your virtual balance moves — is a genuinely useful way to confirm the strategy's pacing and complexity actually suits you before real money is at stake. This is a low-cost way to validate that a strategy which sounds appealing in writing also feels manageable in practice.
The "Iron Cross" worked example
To make the Iron Cross concrete: suppose you bet $5 on the field, $6 each on place bets for 5, 6, and 8 — a total of $23 in action. If a 5 rolls, your place bet on 5 wins $7 (7:5 payout), while your field bet loses; net result: +$2. If a 6 rolls, your place bet on 6 wins $7 (7:6 payout), field loses; net: +$2. If a 4, 9, or 10 rolls, your field bet wins $5 at even money, place bets are unaffected; net: +$5. If a 2 or 12 rolls, your field bet wins a boosted payout (say $10 at 2:1); net: +$10, minus whatever place bets remain unaffected. If a 7 rolls, every bet loses simultaneously: −$23. This concrete breakdown shows exactly why the strategy feels good in the moment (frequent small-to-medium wins) while still carrying a real, calculable cost concentrated entirely on the relatively frequent 7.
Why "feels good" and "is good value" are different questions
The Iron Cross example above illustrates a pattern worth generalizing across every strategy in this guide: how a strategy feels moment-to-moment (frequency of wins, size of typical payout, sense of broad coverage) is a genuinely different question from its actual expected value. Both are legitimate things to care about, but conflating them — assuming a strategy that feels rewarding must therefore be mathematically sound — is exactly the trap that strategy marketing (in craps and in every other casino game) tends to exploit. Reading the house-edge figures directly, rather than trusting a strategy's name or reputation, is the reliable way to separate the two.
How strategy choice interacts with your risk tolerance
A player with a smaller bankroll relative to their desired session length is generally better served by lower-variance approaches (pass line plus modest odds) that make a stop-loss less likely to trigger prematurely. A player with a larger bankroll cushion relative to their session goals has more room to absorb the higher variance of broader-coverage strategies like the 3-point Molly without meaningful risk of an early, unplanned end to their session. This risk-tolerance matching is a genuinely useful strategic consideration layered on top of — not instead of — understanding each approach's underlying house edge.
Strategy and the shooter's perspective
Everything covered in this guide applies identically whether you're betting as a spectator or actively serving as the shooter yourself — being the shooter doesn't change which bets are available to you or their house edge, it only means it's your turn to roll the dice for that round. Some players enjoy specifically choosing to shoot as part of their session experience, while others are equally happy to never shoot and simply bet on other players' rolls; neither preference affects the strategic math covered throughout this guide.
Documenting your own strategy performance
If you want to evaluate how a chosen strategy is actually performing for you across multiple sessions, tracking basic figures (starting bankroll, ending bankroll, approximate total wagered, strategy used) session over session is more informative than relying on memory or impression, which tends to overweight particularly memorable wins or losses. This tracking won't reveal a hidden edge in any strategy — there isn't one to find — but it can help you honestly assess whether your chosen approach and bankroll sizing are actually sustainable for how you want to play over time.
Revisiting your strategy as you gain experience
There's no requirement to settle on one strategy permanently — many players start with simple pass-line-plus-odds play while learning the game's mechanics, then experiment with broader-coverage approaches like the 3-point Molly or Iron Cross once comfortable, and some eventually settle back into a simpler approach after determining the added complexity of managing several simultaneous bets doesn't add enough enjoyment to justify the extra attention required. This evolution is normal and doesn't reflect a "failure" to find the right system — it reflects the genuinely personal nature of what strategy trade-offs are worth making for you specifically.
Comparing craps strategy discipline to other table games
Players experienced with blackjack basic strategy or roulette betting systems will notice a genuine difference in craps: unlike blackjack, where strategy decisions measurably change your odds hand to hand based on visible information, craps strategy is entirely about which pre-existing, fixed-edge bets to combine and how much to stake on each — there's no equivalent to a blackjack "correct play" that changes the probability of a given roll. This distinction matters because it means craps strategy mastery is really about thorough understanding of the betting menu and disciplined bankroll management, not about in-the-moment decision skill the way card-counting-adjacent games can involve. See is craps a game of skill or luck? for a fuller treatment of this distinction.
Frequently asked questions
Does the 3-point Molly strategy reduce the house edge? No — it's built entirely from pass line, come, and odds bets, each carrying the same house edge whether used individually or combined into this named strategy; it changes coverage and variance, not underlying value.
Is the Iron Cross a good strategy? It offers a distinctive experience of winning on nearly every roll except a 7, but its blended house edge is worse than pass line plus maximum odds — a legitimate variance choice, not a value optimization.
What's the single most mathematically sound craps strategy? Maximizing your odds bet relative to your pass line or come bet, since odds carry zero house edge and directly lower your blended overall cost.
Does betting the don't side provide a meaningful edge over pass line? Only a marginal one (1.36% versus 1.41%) — genuinely negligible in practical terms, and probably not worth choosing purely for this reason if you're uncomfortable betting against the table.
Is dice control a legitimate strategy? It isn't supported by rigorous evidence, and casino rules requiring dice to hit the back wall exist specifically to disrupt attempted control; it's also entirely inapplicable to RNG or bubble craps formats.
Do progressive betting systems work on craps' even-money bets? No — like on any even-money bet, progression systems change how wins and losses are distributed across a session without altering the underlying house edge of any individual bet.
Should I use a complex strategy as a beginner? Not necessarily — pass line plus odds alone is simple, requires understanding just two bet types, and remains one of the best-value approaches available regardless of experience level.
Does a bigger bankroll require a different strategy? Not necessarily a different strategy, but it does open up higher odds multiples and more simultaneous bets, which is why bankroll planning and strategy choice are closely linked considerations.
Can I combine multiple named strategies? Yes, informally — since every named strategy is built from the same underlying bets, you can mix elements (say, a 3-point Molly core with an occasional Iron Cross round for variety) without any rule preventing it.
Why do craps strategies feel more numerous than roulette strategies? Craps' much wider bet menu, with meaningfully different house edges and persistence rules across bet types, naturally creates more combinatorial possibilities for named strategies than roulette's more uniform bet structure.
Does "hedging" pass line with a don't pass bet at the same time create value? No — it mostly cancels out variance rather than creating value, since you're paying the combined house edge on two offsetting bets rather than gaining any genuine protection.
Is pressing or regressing place bets a better approach? Neither is mathematically better — both reshape when and how much you might win or lose without changing the fixed house edge of the underlying place bet; the choice is a personal risk-management preference.
Should my strategy change based on how long I plan to play? It can reasonably — lower-variance approaches suit shorter, budget-constrained sessions, while broader-coverage strategies can keep a longer session feeling more consistently engaging, despite a slightly higher blended house edge.
