Craps loss chasing begins when the purpose of the next wager changes from playing the game to erasing the current deficit. The player increases stakes, adds faster or higher-edge bets, borrows from money set aside for later, or stays longer than planned because leaving while behind feels unacceptable.
Nothing in the dice process recognizes that recovery target. The next roll has the same probabilities it had before the loss. Chasing changes the amount exposed and the pressure surrounding the decision, not the chance that 6, 8, a hardway, or a point will arrive.
A normal loss can become a different kind of session
Imagine a player who brings $300 and begins with $10 on the Pass Line plus modest odds. Two short hands remove $80. The original plan was to play for an hour, but the player now decides that leaving down $80 is not an option.
The progression often looks like this:
- Increase the line bet from $10 to $25.
- Add $30 each on 6 and 8 to create more “ways to recover.”
- Add hardways or one-roll propositions after a point is established.
- Take money from the remaining session bankroll after another seven-out.
- Extend the visit because the new deficit is larger than the first one.
Each step feels like a response to the previous loss. Mathematically, each step is a new wager with its own price. The old loss is already settled.
The recovery target does not create an advantage
Let:
- (D) = current deficit;
- (b) = net profit per unit on the chosen wager when it wins;
- (S) = stake needed for one win to recover the deficit.
Ignoring other bets and the possibility of a loss, the minimum recovery stake is:
[ S=\frac{D}{b} ]
A $120 deficit could theoretically be recovered with an $11 street-style payout of about $10.91, but craps bets have different win probabilities and payout ratios. A one-roll 12 might offer a large payout, yet it loses on 35 of 36 combinations. A Place 6 pays only 7 to 6, so one win would require more than $102 staked to erase a $120 deficit immediately.
The formula is not a strategy. It reveals the trap: as the payout per unit becomes smaller, the stake required for instant recovery becomes larger. As the payout becomes larger, the chance of winning usually becomes smaller and the house edge may rise sharply.
Expected loss grows with action
For a wager amount (A) and house edge (h):
[ E(L)=A\times h ]
If several wagers are made, total expected loss is:
[ E(L)=\sum_{i=1}^{k}A_i h_i ]
Suppose a player normally risks $20 per decision on bets averaging a 1.5% edge. Expected loss per decision is:
[ 20\times0.015=$0.30 ]
After falling behind, the player raises total action to $80 and adds proposition bets, increasing the weighted average edge to 4%. Expected loss becomes:
[ 80\times0.04=$3.20 ]
The expected cost per decision is now more than ten times higher. This does not mean every next decision loses $3.20. It means the escalation has materially worsened the long-run price while also increasing short-term volatility.
The craps expected-loss guide shows how stake, decisions, and edge combine across a session. The house-edge page compares the price of individual wagers.
Why odds bets do not make chasing safe
Free odds behind a Pass or Come bet have zero house edge. That is mathematically favorable compared with most casino wagers, but it does not eliminate risk.
A player with $10 on the Pass Line and $50 odds has $60 exposed to one point resolution. The line portion carries the edge; the odds portion increases the amount that can swing. If the player adds odds only because the bankroll is down and “one hit gets me even,” the decision can still be loss chasing even though the added component is fairly priced.
A low edge answers how expensive the wager is per dollar over time. It does not answer whether the current bankroll can absorb the next result. The craps bankroll-risk guide addresses that separate question.
Why doubling systems reach a wall
A formal loss-chasing progression often doubles after each loss: 10, 20, 40, 80, 160, and so on. The apparent attraction is that one even-money win recovers all previous losses and earns the original unit. The hidden cost is exponential stake growth.
After (m) consecutive losses from an initial stake (u), cumulative loss is:
[ D_m=u(2^m-1) ]
With a $10 starting unit:
| Consecutive losses | Next required stake | Cumulative amount already lost |
|---|---|---|
| 1 | $20 | $10 |
| 2 | $40 | $30 |
| 3 | $80 | $70 |
| 4 | $160 | $150 |
| 5 | $320 | $310 |
| 6 | $640 | $630 |
A player who began with a harmless-looking $10 target needs $640 for the seventh decision and has already lost $630. Table limits, bankroll limits, or a further loss can stop the sequence before the recovery win arrives.
Craps makes the progression less clean than a textbook even-money example. Pass and Don’t Pass outcomes include pushes or different come-out treatment; point resolutions take varying numbers of rolls; many proposition bets do not pay even money; and multiple wagers may resolve together on 7. A doubling rule therefore adds complexity without removing the finite-bankroll problem.
The chance of a losing run is never zero. If a wager loses with probability (q), the probability of (m) consecutive losses is (q^m). A sequence may be uncommon, but the amount required grows much faster than the probability shrinks. Repeating the progression across many sessions also creates more opportunities to encounter the run that exceeds the available bankroll.
Common chasing patterns at a craps table
Loss chasing does not always look like a formal Martingale. It can appear as:
- pressing immediately after losses rather than after wins;
- moving from 6 and 8 to hardways, horn, or hop bets for a faster recovery;
- placing every box number because one of them “has to hit”;
- restoring a lost Come bet with a larger replacement;
- cancelling a planned break because the shooter changed;
- buying more chips without recalculating the total session loss;
- treating comps or cashback as justification for continued action;
- refusing to leave until the bankroll reaches an earlier high point.
The clearest test is not the bet name. Ask: Would I make this same wager, at this same size, if I were currently even? If the answer is no, the deficit is driving the decision.
Why the table environment can intensify it
Craps provides many simultaneous positions and frequent moments of hope. A point may be established while Place bets, Come bets, hardways, and bonus wagers remain active. A player can therefore feel “close” to a recovery even when several independent conditions still need to occur.
The social energy also matters. A long roll can make leaving feel premature; a quick seven-out can create a desire to repair the result with the next shooter. Chips make the changing deficit less visible than a bank balance. Fast settlement means another decision arrives before the previous disappointment has fully cooled.
None of those features alter the dice. They affect attention, time perception, and willingness to increase exposure.
A precommitted stop plan is different from predicting the dice
A useful stop rule does not claim to improve the game. It limits the amount of future exposure that can be added after an emotional trigger.
A practical plan can specify:
- the total amount available for the session;
- a maximum stake per decision;
- which bet types are allowed;
- a time or shooter limit;
- a loss amount that ends the session;
- a rule against ATM withdrawals or new credit;
- a mandatory break after a rapid drawdown.
For example: “I will bring $250, make only Pass/Come and 6/8 wagers, never exceed $40 total action on one resolution, and leave if $150 is lost.” The rule does not reduce the house edge, but it prevents an $80 disappointment from becoming an unplanned $500 exposure.
The strongest version is operational: carry only the planned cash, leave bank cards elsewhere, use account limits where available, and tell a companion the stopping point before play begins.
The break-even fallacy
Players sometimes say, “I am not trying to win; I only want to get back to even.” The casino does not offer a special probability for that objective. To return from (-D) to zero, the player must expose additional money to more negative-expectation decisions.
Even if recovery occurs, it can reinforce the behavior. A player who once turns a $200 deficit into a small profit may remember the successful escape and underweight the sessions where the same escalation produced a much larger loss. Recovery proves that variance can move in either direction; it does not prove that chasing was sound.
The broader principle is explained in why no betting system can change probability. Changing stakes changes the distribution of money won and lost, not the underlying future combination frequencies.
When chasing becomes a harm indicator
A single impulsive increase does not by itself establish a gambling problem. Concern rises when the pattern repeats, involves money needed elsewhere, defeats prior limits, produces secrecy or borrowing, or continues despite distress.
Regulators explicitly treat chasing losses and erratic betting as behavioral indicators that operators should consider in customer-interaction systems. The UK Gambling Commission’s premises-based customer-interaction guidance gives one current regulatory example. Requirements and intervention practices differ by jurisdiction, but the behavior itself is widely recognized as a warning sign.
If stopping feels impossible in the moment, the useful action is not to find a lower-edge recovery bet. Leave the table, create physical distance from access to money, and use the gambling-support or self-exclusion options available in the relevant jurisdiction.
A decision made before the deficit is usually clearer
Loss chasing turns a past result into a reason for taking more future risk. The most effective correction happens before the session: define the bankroll, bet range, allowed wagers, and stopping conditions while no recovery pressure exists.
Once the plan is breached, the session has changed. Leaving does not lock in a strategic failure; the loss was locked in when the losing wagers were settled. Leaving prevents a settled loss from dictating the size and duration of the next one.