There's no craps strategy that beats the house long-term — the math doesn't allow it. What separates a good strategy from a bad one is how close you get to fair odds, how well you manage the money you're playing with, and how well you understand what you're actually signing up for over a full session. This page walks through that in three tiers: beginner, intermediate, and advanced. Start wherever your own game is and stop whenever you've got what you need — you don't have to read straight through to get value out of any one section.
Bankroll Management
No bankroll approach changes the house edge — that's fixed by the bet, not by how you manage your money. What bankroll management actually does is control how likely you are to go bust on a bad run before the math has a chance to even out, and how long you can keep playing without a single roll wrecking your night.
A few practical habits cover most of it:
- Set a session bankroll before you buy in — an amount you're genuinely fine losing, separate from money earmarked for anything else.
- Size your bets to your bankroll, not the other way around. A common guideline is keeping your bankroll at 30-40 times your typical pass line bet, so a cold streak doesn't end the session in ten minutes.
- Decide your stop-loss and your walk-away win number before you buy in — while you're thinking clearly, not mid-session when a loss or a win is already pulling at you.
- Don't chase losses by increasing bet size to "get even." The house edge doesn't care what you're down; a bigger bet just exposes more money to the same edge.
None of this improves your odds on any individual bet. It just keeps a bad run from turning into a bad night. This applies at every tier below, not just the beginner one — it's the one piece of strategy that never stops mattering no matter how advanced your betting gets.
Beginner Craps Strategy
The Pass Line Plus Free Odds
If you take away everything else on this page, take this: bet the pass line, and once a point is set, back it with free odds. That combination is the closest thing craps has to a foundation strategy, because it's the only path on the entire layout that gets you near a fair game. The pass line alone carries a 1.41% house edge. Odds carry zero house edge on their own. Blend the two at a table offering 3-4-5X odds and taking full odds every time, and the combined house edge on your total action drops to roughly 0.37% — one of the lowest numbers in any casino game, table or slot.
The tradeoff is that odds only exist as an add-on. You can't walk up and bet odds by itself, and you can't touch a pass line bet once it's down until the round resolves. If you want the full mechanics of how each piece works, the Every Bet Explained page covers it in detail. Everything past this section builds on this pair, so it's worth being genuinely comfortable with it before adding anything else.
Intermediate Craps Strategy
The Iron Cross
The Iron Cross combines a field bet with place bets on 5, 6, and 8, structured so that every number except 7 wins something. A common version bets $5 on the field, $5 on the place 5, and $6 each on the place 6 and 8 — roughly $22 total, covering 2, 3, 4, 5, 6, 8, 9, 10, 11, and 12. Roll anything but a 7, and one of those bets pays. Roll a 7, and you lose the entire $22 at once.
The appeal is obvious: you win on 30 of the 36 possible dice combinations. The catch is that the loss on the remaining 6 combinations is large enough to offset the frequent small wins — the blended house edge across the whole combination works out to roughly 1.14% per roll, which is a genuinely reasonable number as far as multi-bet systems go, but it's not better than pass line plus odds, and every roll it's exposed to the same 7 that ends any other bet. It rewards patience with steadier action, not better math.
The 3-Point Molly
The 3-Point Molly is a way of running the pass-line-plus-odds approach across three numbers at once instead of one. Start with a pass line bet and full odds once the point is set, then add two come bets, each backed with full odds once they establish their own points. Once all three are working, you effectively have three numbers with full odds riding at the same time.
This doesn't lower the house edge below what pass line plus odds already gives you — each individual bet still carries the same 1.41% and 0% split it always does. What it changes is your exposure: more of your bankroll is working on more numbers simultaneously, which means more total action per hour and more variance in both directions, not a better rate of return per dollar. It's a strategy for players who've got the pass-line-and-odds math down and want more of the table working for them, not a way to beat it.
Place Betting the 6 and 8
Rather than waiting for a pass line point to land on 6 or 8, you can place those numbers directly the moment you buy in, any time, point phase or not. The house edge on a place 6 or place 8 is 1.52%, close enough to the pass line's 1.41% that the difference barely matters, and you skip the wait for a specific come-out result.
The flexibility cuts both ways. Place bets can be turned off or taken down entirely between rolls, which pass line and come bets can't — useful if you want to pause action without walking away from the table. But that same flexibility means there's no free odds bump available the way there is on a pass line bet, so 1.52% is close to the floor for this approach rather than something you can push lower.
Advanced Craps Strategy
Hedging Bets (and Why the Math Says No)
Hedging in craps usually means adding a small bet on the numbers that would make your main bet lose — most commonly, tossing a few dollars on "any craps" (2, 3, or 12) alongside a pass line bet on the come-out roll, as insurance against the pass line crapping out. It feels like protection. It isn't.
The any craps bet itself carries an 11.11% house edge, several times worse than the pass line bet it's supposedly protecting. Because a pass line loss on come-out only happens on three specific rolls out of eleven possible outcomes, you're paying a persistently high-vig premium to insure against a relatively infrequent event, and that premium costs more on average than the loss it's covering. Combining any low-edge bet with a high-edge hedge doesn't lower your overall expected loss — it raises it, because you're now wagering more total money at a worse blended edge than either bet carried on its own. Hedging can smooth out how a loss feels in the moment, but it doesn't change what the math says over a session, and the math says no.
Playing the Don't Side
Betting the don't pass line instead of the pass line means betting with the house's own long-run advantage instead of against it, in a specific sense: the house edge on don't pass is 1.36%, a touch better than pass line's 1.41%, because of the 12 push on come-out. Lay odds behind a don't pass bet at max odds and the combined house edge can drop to roughly 0.27%, edging out even pass-line-plus-odds as one of the lowest-cost bets in the building.
The tradeoff isn't mathematical, it's social. Betting the don't side means rooting against the shooter everyone else at the rail is cheering for, and that can draw some real friction at a hot table. Nothing about the math cares how the table feels about it — the don't side is a legitimate, low-cost way to play craps, and the resistance to it is cultural, not statistical.
Session Variance
The house edge tells you what happens over an enormous number of bets. It tells you almost nothing about what happens over one session. Craps has a lot of variance built into it — even the best-priced version of the game, pass line plus full 3-4-5X odds, carries a standard deviation of roughly 4.92 units per bet resolved, dwarfing the 0.37% edge those bets carry. In plain terms: the swings around your average result are enormous compared to the size of that average, which is exactly why a single night can look nothing like what the house edge predicts, in either direction.
This cuts both ways. A short session with a hot shooter can produce a real, sizable win that has nothing to do with skill and everything to do with variance running favorably. A long cold stretch isn't a sign you're doing anything wrong, either — it's the same variance running the other way. Some players try to influence which way that variance breaks through a controlled throw, a contested and unproven idea covered on the dice setting page if you want to look into both sides of that debate. Whether or not you buy into it, treating any single session as proof the math is wrong, in either direction, is the mistake to avoid. The edge only shows up reliably over a much longer run than any one visit to the table.
If you want to see any of this play out before you bring real money to a table, the simulator lets you run pass line plus odds, the Iron Cross, or the don't side over as many rolls as you want, with nothing on the line but time. That's the fastest way to feel the difference between a strategy's edge and its variance, instead of just reading about it.