Craps Expected Loss Per Hour

A craps table can look inexpensive because the minimum bet is printed in one corner of the layout. That number is only the admission price. The real hourly cost comes from how much action is repeatedly exposed to house-edge bets, how often those wagers resolve, and how many extra bets are added around the line bet.

That is why two players at the same $15 table can have completely different expected losses. One may make a Pass Line bet and take odds. Another may keep the same line bet but add Place 6 and 8, Field bets, hardways, and one-roll propositions. Their table minimum is identical. Their mathematical spend rate is not.

Start with action, not the table minimum

Expected loss is the long-run average cost of wagering. For a single bet type, the basic relationship is:

Expected Loss = Amount Wagered × House Edge

For an hour of play:

Expected Loss per Hour = Hourly Action × House Edge

If several different bets are being made, calculate them separately and add them:

Total Hourly Expected Loss
= Pass/Come Expected Loss
+ Place-Bet Expected Loss
+ Field Expected Loss
+ Proposition Expected Loss
+ Other House-Edge Bet Expected Loss

The phrase hourly action is the part that requires care. It means the amount actually wagered under the house-edge convention being used. It does not automatically mean “bet size multiplied by every dice roll.”

A Pass Line bet does not become a new wager on every roll

Suppose a player puts $15 on the Pass Line. On the come-out roll, the bet may win, lose, or establish a point. If a point is established, that same $15 line wager remains alive until the point or a 7 resolves it.

The familiar Pass Line house edge of about 1.41% is normally quoted per initial line wager, not as 1.41% of the $15 again on every roll while the point is unresolved.

So if the player starts 30 new $15 Pass Line decisions in an hour:

Pass Line Action = 30 × $15 = $450
Expected Loss ≈ $450 × 0.01414 = $6.36

That is a useful estimate. Multiplying $15 by every physical dice roll and then applying the same 1.41% would overstate the cost because it mixes two different measurement bases.

This distinction matters throughout craps because some bets resolve immediately while others can remain working across several rolls.

Place bets require a different counting habit

A Place 6 or Place 8 wager can stay on the layout through neutral rolls. The standard house-edge percentage of about 1.52% is usually expressed per wager resolved, not per roll that the bet sits on the table.

For a $6 Place 6 bet, one roll has three broad possibilities:

• a 6 appears and the bet wins $7;
• a 7 appears and the $6 loses;
• another total appears and the bet remains unresolved.

The expected change on one roll is:

(5/36 × $7) - (6/36 × $6) = -$1/36 ≈ -$0.0278

That is about 0.463% of the $6 stake per roll of exposure. If instead you wait until either 6 or 7 resolves the wager, the familiar edge is about 1.52% per resolved Place 6 decision.

Both descriptions are valid. The mistake is using the per-resolved-bet percentage while also counting the same live bet as a fresh wager on every neutral roll.

For a practical player estimate, the easiest method is usually to count how often a wager is actually made or reset, then use the matching published edge. If you are building a roll-by-roll model, use per-roll expectation consistently instead.

Odds bets change volatility without adding house edge

Free odds behind a Pass or Come bet pay at true odds. Their house edge is 0% when the payoff is correct.

That does not mean odds are irrelevant to the session. They increase the amount of money at risk when a point is active, so they increase swings and bankroll requirements. They can make a winning or losing session much larger even though they do not add expected loss by themselves.

Example:

• $15 Pass Line
• $30 odds after a point is established

The $30 odds bet adds no theoretical house cost, but a seven-out now removes $45 instead of $15. Expected value and volatility are different questions.

Read Craps Odds for that distinction before treating “0% edge” as “no risk.”

Extra one-roll bets can dominate the hourly cost

The fastest way to make a low-edge craps session expensive is to layer high-edge action on top of the main wager.

Imagine this approximate hour:

The exact numbers depend on the pay schedule and how the bets are made, but the pattern is clear. A player who says, “I only risk $15 on the line,” may actually be creating several thousand dollars of total action per hour.

The house edge on the main bet can be small while the weighted edge of the whole betting pattern is much larger.

Calculate the weighted cost of the whole rack of bets

A useful way to summarize a mixed strategy is to calculate the total expected loss and divide by total house-edge action.

Suppose the hour contains:

• $450 Pass Line action at 1.414%;
• $900 Place 6/8 action at 1.515%;
• $360 Field action at 2.778%;
• $180 proposition action at an average 10% edge.

Expected loss is approximately:

Pass Line: $450 × 0.01414 = $6.36
Place 6/8: $900 × 0.01515 = $13.64
Field:      $360 × 0.02778 = $10.00
Props:      $180 × 0.10    = $18.00
-----------------------------------
Total Expected Loss        = $48.00

House-edge action excluding free odds is $1,890.

Weighted House Edge = $48 / $1,890 ≈ 2.54%

That weighted figure describes the average mathematical cost of the bet mix, not any single printed wager.

Table speed changes cost only when it changes decisions and action

A crowded table may produce fewer completed betting decisions per hour than a fast table with experienced players and a quick crew. More decisions generally mean more wagering opportunities.

But “rolls per hour” alone is not enough. What matters is how your bets respond to those rolls.

A player who makes only a line bet creates a new house-edge wager when the line decision resets. A player making a fresh Field bet every roll creates house-edge action on every roll. A player pressing Place bets creates a third pattern.

So the useful question is not:

How many rolls happened?

It is:

How much qualifying action did my betting pattern create during those rolls?

The Craps Roll Speed and Total Action page goes deeper into this operational side.

A $25 table can be cheaper than a $10 table

Minimums do not rank games by hourly cost.

Consider two simplified players:

Player A — $25 table

• $25 Pass Line only
• 28 new line decisions per hour
• hourly line action: $700
• expected loss: about $9.90

Player B — $10 table

• $10 Pass Line
• $12 Place 6 and $12 Place 8
• repeated $5 Field bets
• occasional $5 hardways

Even with smaller individual chips, Player B may create well over $1,500 of house-edge action in the same hour. The expected cost can easily exceed Player A’s.

The lower table minimum creates a lower starting point, not a guarantee of lower total exposure.

Session length multiplies the expected cost, not the certainty of losing

If a betting pattern has an estimated expected loss of $35 per hour, then a four-hour theoretical cost is:

$35 × 4 = $140 expected loss

That does not mean the player will be down exactly $140. Craps has substantial short-term variance. The player can finish ahead, lose much more than $140, or land near the expectation.

Expected loss is a pricing model for repeated action. It is not a prediction of tonight’s cash-out.

This is also why doubling session length is more meaningful than asking whether a particular shooter is “hot.” More time usually means more betting decisions, and more decisions give the house edge more total action to work on.

Bankroll questions need variance as well as expectation

Two strategies can have the same expected loss and very different risk of short-term ruin.

A player who keeps $500 in a session bankroll and bets $10 at a time may survive ordinary swings comfortably. A player with the same expected hourly loss but much larger odds bets or volatile propositions may hit the same long-run cost with much rougher short-term movement.

For bankroll planning, track at least:

• expected loss per hour;
• average amount simultaneously exposed;
• largest likely single-decision loss;
• volatility of the bet mix;
• session duration;
• stop point that does not depend on “getting even.”

Use the bankroll risk calculator for risk-of-ruin style thinking and the expected loss calculator for the average-cost side.

Do not compare house-edge percentages without their pay schedule

Craps proposition payoffs can vary. A Field bet may pay differently on 2 or 12. Hardway payouts and special side bets can differ. Electronic and stadium versions may use approved variants that are not identical to the felt you played yesterday.

Nevada’s Gaming Lab publishes current rules of play for approved games. That is a useful reminder that the actual rules and payout schedule define the wager; the label alone is not enough.

Before putting a percentage into an hourly-cost model, confirm the exact bet and payoff in front of you.

The practical worksheet

For a real session, use a simple table:

Then total the last column.

If you cannot estimate the number of bets, start with a 15-minute observation of your own play and multiply by four. It will still be an approximation, but it is usually better than pretending the table minimum equals hourly cost.

The useful answer to “what does craps cost per hour?”

There is no single craps hourly-loss number. The meaningful answer is personal to the betting pattern.

A disciplined estimate follows four rules:

1. count every house-edge bet, not only the line bet;
2. match the action count to the way the quoted edge is defined;
3. keep free odds separate from expected loss because they add variance, not house edge;
4. multiply the resulting hourly cost by realistic time, while remembering that actual results can swing far from expectation.

For the underlying percentages, continue to Craps House Edge and Craps RTP. For a full beginner path, use the Craps guide. The arithmetic becomes useful only after the betting behavior is described honestly.