Craps proposition bets can be ranked, but only after the payout and the way the bet resolves are specified. Under the common paytables used in the calculations below, the Field is the least costly one-roll option at 2.78% or 5.56%, hard 6 and hard 8 carry 9.09%, and Any Seven is the worst standard center-table wager at 16.67%.
That does not make the top-ranked proposition a strong foundation bet. A Pass Line wager has a much lower edge, and correctly paid odds have no house edge. This page answers a narrower question: if a player insists on proposition action, which common choices cost less or more per dollar wagered?
The ranking at a glance
The figures below assume two fair dice and the listed payouts stated to one. A different paytable produces a different edge.
| Rank | Wager or wager group | Assumed payout | House edge | What decides it |
|---|---|---|---|---|
| 1 | Field, with both 2 and 12 paying 3 to 1 | Mixed | 2.78% | One roll |
| 2 | Field, with both 2 and 12 paying 2 to 1 | Mixed | 5.56% | One roll |
| 3 | Hard 6 or hard 8 | 9 to 1 | 9.09% | Hard number before easy number or 7 |
| 4 | Any Craps; 3 or 11; C&E; two-combination hop | 7 to 1 or 15 to 1 | 11.11% | One roll |
| 5 | Hard 4 or hard 10 | 7 to 1 | 11.11% | Hard number before easy number or 7 |
| 6 | Horn, using 30-to-1 and 15-to-1 component payouts | Split wager | 12.50% | One roll |
| 7 | Whirl or World, using Horn plus Any Seven | Split wager | 13.33% | One roll |
| 8 | 2 or 12 at 30 to 1; one-combination hard hop at 30 to 1 | 30 to 1 | 13.89% | One roll |
| 9 | Any Seven | 4 to 1 | 16.67% | One roll |
Some casinos use different Field awards, hop payouts, or minimum betting units. The table sign and layout control the actual wager. “Ranked fourth” here means lower mathematical cost than the rows below it under these assumptions; it does not mean recommended.
How the ranking is calculated
Two dice produce 36 equally likely ordered combinations. For a one-roll wager with:
- $W$ winning combinations;
- $L = 36-W$ losing combinations; and
- a payout of $R$ to 1,
its expected value per unit wagered is:
$$EV = \frac{W(R)-L}{36}$$
The house edge is the negative expected value divided by the initial stake:
$$House\ edge = -EV$$
For a $1 Any Seven bet, six combinations win and 30 lose. At 4 to 1:
$$EV = \frac{6(4)-30}{36} = \frac{-6}{36} = -0.1667$$
The expected loss is 16.67 cents per $1 wagered, so the house edge is 16.67%. Seven is the most frequent total, but the common 4-to-1 payout is below the fair 5-to-1 price.
The one-roll bet guide derives the individual wagers in more detail. This page compares them on the same scale.
Why the Field can rank first without being a low-edge craps strategy
A standard Field wins on 2, 3, 4, 9, 10, 11, or 12 and loses on 5, 6, 7, or 8. There are 16 winning combinations and 20 losing combinations.
If 3, 4, 9, 10, and 11 pay even money while both 2 and 12 pay 2 to 1, the expected value of a $1 Field bet is:
$$EV = \frac{14(1)+2(2)-20}{36} = \frac{-2}{36} = -5.56%$$
If one of the extreme totals pays 3 to 1 while the other pays 2 to 1, the edge falls to 2.78%. If both pay 3 to 1, the arithmetic is even more favorable to the player, although that layout is not universal.
The practical lesson is not “bet the Field.” It is read the awards for 2 and 12 before comparing it with another proposition. The Field bet page explains the common paytable variants.
Hardways: lower edge, slower resolution
Hardways are proposition bets, but they are not one-roll bets. Hard 6 wins if 3-3 appears before an easy 6 or any 7. Hard 8 follows the same structure with 4-4. At 9 to 1, each carries a 9.09% house edge.
Hard 4 and hard 10 pay 7 to 1 and carry 11.11%. Other totals leave a hardway unresolved, so the wager may remain on the layout through several rolls.
That creates an important comparison issue:
- house edge measures expected loss per amount resolved;
- resolution speed influences how often the stake can be lost and replaced;
- working status determines whether the wager is active on a come-out roll.
A hard 6 has a lower edge than a Horn, but its hourly cost cannot be predicted from the percentage alone. The number of resolving decisions and whether the player repeatedly presses or replaces it matter. See Hardways House Edge for the complete resolving-combination calculation.
The 11.11% group is not one identical bet
Several common propositions land at 11.11% under the stated paytables:
- 3 or 11 paying 15 to 1;
- Any Craps paying 7 to 1;
- a two-combination hop paying 15 to 1;
- C&E when half the wager is treated as Any Craps and half as 11;
- hard 4 or hard 10 paying 7 to 1.
They share a percentage, but not a playing experience. Hard 4 may remain unresolved. Yo 11 disappears on the next roll. C&E requires a divisible stake and is settled as two component wagers. A player comparing them should therefore consider both the edge and how quickly another wager will be made.
For example, a $2 C&E is commonly split into $1 on Any Craps and $1 on 11. On a craps total, the Any Craps component wins and the 11 component loses. On 11, the opposite happens. On any other total, both lose. The edge is not calculated by pretending C&E is one new dice event; it comes from adding the two component expected values.
Horn, Horn High, and Whirl need component accounting
A Horn divides the stake across 2, 3, 11, and 12. At 30 to 1 on 2 and 12 and 15 to 1 on 3 and 11, a four-unit Horn has a 12.50% house edge.
The calculation must include the losing components on a winning roll. If 2 appears, the unit on 2 wins 30 units, while the other three units lose. The net profit is 27 units, not 30. A hit on 3 or 11 produces 12 units of net profit after the other three components are lost.
A Horn High adds an extra unit to one of those four totals. Its exact expected loss depends on which number is “high” and on the posted component payouts. It should not be assigned one universal percentage without defining the split.
A Whirl, also called a World bet in some casinos, combines a four-unit Horn with one unit on Any Seven. Under the payouts used here, its house edge is 13.33%. A 7 may return the five-unit stake with no profit because the Any Seven win is offset by the four losing Horn components. That apparent “save” does not make the combined wager efficient.
Exact totals and hop bets: confirm whether order matters
A bet on 2 or 12 has one winning combination out of 36. At 30 to 1, the house edge is 13.89%. A one-combination hard hop such as 3-3 at 30 to 1 has the same mathematics.
An easy hop such as 2-4 or 4-2 has two ordered winning combinations. At 15 to 1, its edge is 11.11%. The phrase used at the table must identify the intended combination clearly. A total-6 bet, hard 6, and a specific 2-4 hop are different wagers with different resolution rules.
The official Massachusetts craps rules illustrate this distinction by listing hardways, one-roll totals, Horn, C&E, Whirl, and specific hop combinations separately, with stated minimum payout odds. Other jurisdictions and casinos may approve different offerings, so the local layout remains decisive.
Ranking by percentage is not the same as ranking session danger
Suppose a player makes a $5 Any Seven wager on each of 60 rolls. At a 16.67% edge:
$$Expected\ loss = 60 \times $5 \times 0.1667 \approx $50$$
The player may finish ahead or behind by much more than $50. Expected loss is an average over repeated trials, not a prediction of the next session. But the example shows how a small center bet can become $300 of total action quickly.
Now compare a $5 hard 6. Its 9.09% edge applies when the bet is resolved, not automatically to all 60 physical rolls. Many rolls may leave it standing. To estimate session cost, count the actual decisions, replacements, and presses rather than multiplying the edge by every throw.
This is why a useful ranking has two columns in the player’s mind:
- price per dollar decided; and
- how often another dollar is put at risk.
The expected-loss guide and expected-loss calculator are better tools for session planning than remembering one dramatic payout.
Table procedure can change the result of a verbal bet dispute
Center action is often booked verbally through the stickperson. The safest sequence is:
- call the wager before the dice are released;
- state the amount and exact bet;
- wait for the dealer or stickperson to repeat or confirm it;
- check how the chips were placed;
- clarify whether a standing proposition is working on the come-out roll.
“Horn high,” “C&E,” “hop the sixes,” and “hard six” are not interchangeable calls. A late, incomplete, or unheard call may not become an accepted wager. If a dispute occurs, the useful evidence is the precise call, chip placement, dealer confirmation, dice status, posted paytable, and surveillance record—not the player’s intended meaning after the result is known.
A practical decision rule
For players who want the lowest mathematical cost, proposition bets should not replace the lower-edge core of the game. Pass, Don’t Pass, Come, Don’t Come, and correctly paid odds deserve comparison before center action. The best craps bets page covers that foundation.
For someone choosing a proposition strictly for entertainment:
- check the paytable first;
- use a fixed, small amount rather than pressing after a hit;
- distinguish a one-roll bet from a standing hardway;
- track the total amount wagered, not only the largest payout;
- stop treating “least bad” as “good value.”
The central truth is simple: a large payout tells you what happens when the wager wins. The ranking comes from what happens across all 36 combinations, including the many quiet losses between the loud calls.