Chasing Losses: Why the Sunk-Cost Trap Empties Bankrolls
Every busted bankroll has the same autopsy. It is almost never one bad bet — it is the sequence after a bad run, where bets got bigger, games got worse and a manageable loss became a wipeout. That sequence is called chasing, and it runs on two of the most reliable bugs in human decision-making: the sunk-cost fallacy and loss aversion. Understanding the machinery is the first circuit breaker.
Play with a stop-loss →What chasing losses looks like
Chasing is any bet placed primarily to recover previous losses rather than because the bet itself is worth making. It has recognizable signatures: bet sizes that grow after losses, sessions that extend past their planned end 'just until I am even,' switching to higher-variance bets or unfamiliar games because they offer a faster road back, and topping up a session budget that was supposed to be final. The tell is the motivation. A player betting 2 SUI on dice because that is their unit is playing; a player betting 20 SUI because they are down 18 is chasing. Same interface, same game, completely different activity. The first is entertainment with a known cost. The second is an attempt to make the past not have happened — and no bet has ever done that.
Sunk cost: the money is already gone
The sunk-cost fallacy is the tendency to keep investing in something because of what you have already put in, rather than what the next unit of investment will return. It is why people finish bad movies, hold collapsing positions and — at a casino — keep betting because they are 'in too deep to stop.' The brutal, liberating fact: lost money is gone whether you keep playing or not. It does not sit in the game waiting to be reclaimed. The next bet's expected value is calculated fresh, from zero, every time. Dice on Llamabet pays 99 divided by your win chance with a 1% house edge — that number is identical whether you are up 100 SUI, down 100 SUI, or on your first roll. The game has no memory of your session. Only you do, and that memory is exactly the input a rational bettor would exclude.
Why chasing feels rational: loss aversion
Prospect theory — the framework from Daniel Kahneman and Amos Tversky that reshaped economics — measured something poker players always suspected: losses hurt roughly twice as much as equivalent gains feel good. Losing 50 SUI produces about twice the emotional intensity of winning 50 SUI. That asymmetry warps decisions when you are down. A risky bet that might erase the loss gets weighted by the enormous relief of getting back to zero, while its likely cost — being down even more — gets discounted. In that state, doubling your bet genuinely feels like the prudent move and stopping feels like 'locking in' the loss. But the loss is already locked in; it happened when the bet lost. Stopping just stops the bleeding. Recognizing that the urgency you feel is loss aversion talking, not analysis, is half the battle.
The escalation spiral
Chasing escalates in a predictable pattern. Stage one: bigger bets in the same game, so variance rises and the swings that triggered the chase get larger. Stage two: worse bets — long-shot roulette numbers, max-payout dice targets — because even a doubled bet no longer recovers the hole in one hit. Stage three: fresh deposits beyond the session budget, then money earmarked for other things, then in serious cases borrowed funds. Each stage feels like a small step from the last, which is what makes the spiral dangerous — nobody decides to bet rent money; they decide to bet slightly more than the previous bet, twenty times in a row. The house edge is constant throughout, so total expected loss scales directly with total amount wagered. Chasing does not change your odds. It just multiplies your exposure at the worst possible moments.
Chasing versus Martingale: the same error, formalized
The Martingale system — double your bet after every loss so the first win recovers everything — is chasing with a spreadsheet. It converts the impulse into an algorithm, which makes it feel disciplined, but the math is unchanged: a losing streak of modest length demands enormous bets (ten straight losses starting from 1 SUI requires a 1,024 SUI bet next), and either your bankroll or the table maximum ends the experiment with the whole line lost. Every negative-progression system shares this DNA: frequent small wins, rare catastrophic losses, expected value still negative because the house edge applies to every single wager. If a strategy's core promise is 'recover your losses,' it is a chase — dressed up. The honest framing is that no staking pattern can beat a negative-edge game; it can only rearrange when the losses arrive.
Circuit breakers that actually work
Rules made mid-tilt fail; rules made in advance hold. First, set a stop-loss before you play — a number in SUI, written down — and make hitting it the end of the session, not a decision point. Second, deposit only the session budget. Self-custody makes this practical: keep your bankroll in one wallet and move only tonight's stake to the wallet you play from, so continuing requires a deliberate, friction-filled transfer instead of one click. Third, use time-outs: after any loss that stings, a mandatory break — even ten minutes away from the screen — lets the loss-aversion surge subside before you act on it. Fourth, never play to fix a mood; gambling while upset about gambling is the spiral's ignition. And know the escalation flags that mean it is beyond a discipline problem: borrowing to play, hiding losses, chasing across days rather than within a session. Those are signs to reach for real support — GamCare (gamcare.org.uk) and the National Council on Problem Gambling (ncpgambling.org) both offer free, confidential help.
Frequently asked questions
Why is chasing losses so hard to stop?
Because it is powered by loss aversion — losses register roughly twice as intensely as gains, so the prospect of erasing a loss gets overweighted and the risk of deepening it gets ignored. In the moment, the chase bet genuinely feels like the responsible move. That is why effective defenses are rules set before playing, when you are calm, not willpower during the session.
Does the sunk cost fallacy apply to gambling?
Directly. Money already lost is a sunk cost — it is gone whether you continue or not, and it should have zero influence on the next bet. Every wager has the same expected value regardless of your session history, because the game has no memory. Betting bigger to 'win it back' just wagers more money against the same house edge.
What is a good stop-loss for a gambling session?
A common rule is 20% to 50% of the session bankroll you brought, decided and written down before the first bet. The exact number matters less than two properties: it is set in advance, and hitting it ends the session automatically rather than opening a debate. Depositing only that amount makes the rule self-enforcing.