KEY TAKEAWAYS
CONTENTS
The sunk cost fallacy in sports betting is the habit of letting money, time or effort you have already spent decide what you do next. A bettor holds a futures ticket long after the team's chances have collapsed, adds to a losing position to "average down," or turns down a fair cash-out because accepting it would turn a paper loss into a real one. In each case the original stake is treated as a reason to keep going, when it is actually a cost that stays exactly the same no matter what is done now. This matters for decision quality because every new decision should be judged by its own expected value: the price available today set against an honest estimate of the probability. This article defines the fallacy, shows how it appears in futures, parlays and live tickets, works through two hypothetical examples, and explains a forward-looking way to evaluate a position. It is not a method for recovering losses. Understanding the bias does not remove variance or create an edge, and no past result changes the odds of the next bet.
Sunk Cost Fallacy Definition: Past Stakes vs. Future Value
A sunk cost is a cost that has already been paid and cannot be recovered by any choice available now. The sunk cost fallacy is the tendency to let that cost influence a decision anyway. Behavioral economists describe it as continuing a behavior or endeavor because of previously invested resources such as time, money or effort. The principle is simple: only future costs and future benefits can change with your choice, so only they belong in a rational comparison.
In betting, the stake on a placed ticket is the clearest sunk cost. Once a $100 bet is accepted, that $100 is committed. If the bet wins, you collect a payout; if it loses, the money is gone. What remains open is whether to hold, cash out, hedge or add to the position. Each of those choices has its own price and its own probabilities, and the original stake appears in none of those calculations except as an amount that is already spent in every scenario.
Research suggests the pull is real. In a well-known 1985 study by Arkes and Blumer, discussed in later work by Arkes and Ayton, theater patrons who paid the full price for season tickets attended more plays over the following six months than patrons who paid a discounted price, even though every ticket allowed the same attendance. The more people had paid, the more motivated they were to use the ticket, which is the fallacy in miniature. The pattern is also called the Concorde fallacy, after the supersonic airliner project that the British and French governments kept funding after the economic case had faded.
How the Sunk Cost Fallacy Shows Up in Betting
The fallacy rarely announces itself as a bias. It arrives as a reasonable-sounding sentence in the bettor's head. Four patterns account for most cases, and each one can be tested with the same forward-looking question.
Holding hopeless futures and parlays
A futures ticket bought in the preseason can look unrealistic by midseason. The thought "I've held it this long" treats time already waited as a reason to keep waiting. Time held is not evidence of value. A parlay with one leg left works the same way: the legs that already won shaped the ticket's current worth, but they do not raise the probability of the last leg. Whether to hold depends on the ticket's current value, such as a cash-out offer, against the chance it still pays.
Doubling down on a losing position
Adding to a losing position to "average down" is the second pattern. The bettor reasons that a second stake at a better price improves the overall position. The second stake is a brand-new bet and has to justify itself on its own price and probability. If it would be a poor bet for someone with no earlier ticket, it is a poor bet for you, and the earlier stake cannot repair it. When the added money is chosen to get back to even, it crosses into chasing losses.
Refusing to cash out or hedge
A third pattern is refusing to sell a ticket for less than was paid. A cash-out offer below the stake feels like admitting a loss, so the bettor holds. But the shortfall against the stake exists whether or not you accept the offer. The real comparison is the offer against the ticket's expected value. The same logic applies to hedging a bet: a hedge should be judged by how it changes risk and expected return from here, not by whether it feels like conceding.
Treating time and research as a reason to bet
Sunk costs are not only money. Hours of research on a pick can make a bettor reluctant to abandon it when news changes, or to pass on a game after a long analysis. Effort spent is not a reason to bet. A thoroughly researched game that ends with "no value at this price" is a good result of the process, and the correct bet is no bet.
Sunk Cost Fallacy Examples: Two Worked Scenarios
Both examples use hypothetical odds and probability estimates to illustrate the arithmetic. They are not current market prices, and the estimates could be wrong.
Example 1, adding to a futures position. Earlier in the season a bettor staked $200 on Team X to win the title at +800, which would have paid $1,600 in profit and $1,800 in total payout. The implied probability at +800 was 100 / (800 + 100) = 11.11%. Team X has since struggled, and the same futures market now lists +1500. A new $100 stake at +1500 would win $1,500 in profit and return $1,600 in payout, so the break-even probability is 100 / (1500 + 100) = 6.25%. The bettor honestly estimates Team X's title chance at 5%, so the expected value of the new $100 is 0.05 × $1,500 − 0.95 × $100 = $75 − $95 = −$20.00. The $200 already staked appears nowhere in that calculation. If the estimate were instead 8%, the expected value would be 0.08 × $1,500 − 0.92 × $100 = +$28.00, which would be a reason to bet, but a reason that came from the estimate and not from the earlier ticket.
Example 2, a cash-out offer. A bettor staked $100 on a longshot at +1000, which pays $1,000 in profit and $1,100 in total payout. After a bad turn of events the sportsbook offers a $60 cash-out, and the bettor feels that taking it means "losing $40." But the $100 is spent either way. The real choice is $60 now against an expected return of $1,100 × p from holding, where p is the bettor's estimated win probability. The break-even probability for holding is $60 / $1,100 = 5.45%. The table shows how the decision depends on that estimate.
| Estimated win probability | Expected return from holding | Cash-out offer | Higher expected return |
|---|---|---|---|
| 4% | $44.00 | $60 | Cash out |
| 5.45% | $60.00 | $60 | Equal |
| 6% | $66.00 | $60 | Hold |
The table shows that the decision flips on the probability estimate, not on the $100 stake. At 4%, holding is expected to return $44, which is $16 less than the offer. At 6%, holding is expected to return $66, which is $6 more. Cash-out prices include the sportsbook's margin, and a probability estimate for a longshot is uncertain, so treat this as a range of possibilities rather than a verdict. Neither choice removes the risk of ending with less than the original stake.
How to Evaluate a Bet on Forward-Looking Value Only
The core fix is to change the question. Instead of "how do I get back what I put in?", ask "would I make this bet at today's price if I had no position?" If the answer is no, then holding, adding or refusing to sell is being driven by the past. This single reframe turns a feeling about what you spent into a comparison of current price and current probability.
In practice, a forward-looking review has three inputs: the current price, your honest probability estimate and the risk you are willing to carry. The current price means the odds still available on the same outcome, or the cash-out or hedge offer on the ticket you hold. The estimate should be written down before you compare it with the price. The risk question is about bankroll, never about recovery. Then compare expected value from here under each available choice, including doing nothing.
Two supporting habits help. The first is a fresh-eyes test: imagine the ticket was handed to you today for free, or that a friend is asking your advice about the identical position. The second is a pre-commitment note written at purchase, stating what would make you sell or hold, such as a price level or a news event. Pre-commitment moves the decision to a calmer moment, though it must be revisited when real information changes. Updating because a key player is injured is forward-looking; refusing to update because of what you paid is not.
Uncertainty matters here. A probability estimate for a longshot can easily be off by several percentage points, as the table showed, and a small estimate error can flip the decision. That is a reason to keep decisions modest and to avoid treating any output as certain, not a reason to default to whatever protects the earlier stake.
Common Mistakes and Misconceptions
Believing that selling is what makes a loss real. The loss against the original stake exists either way; holding is simply a new bet that uses the ticket's current value as its stake. Seeing it that way removes the false comfort of a position that is "not a loss until it is sold."
Overcorrecting into "always sell" or "never hold." Ignoring sunk costs does not mean a ticket should be sold. Sometimes the forward-looking answer is to hold, when the current win probability exceeds the break-even implied by the offer. What matters is that the reason is the current value and not the history.
Treating the idea as permission to bet again. A bettor who learns that past losses are irrelevant may conclude that a fresh bet is a clean slate for winning the money back. No future bet is obligated to repay a past loss, and an earlier result does not change the next bet's odds or expected value. Looking for a bet to recover money is the problem, whether or not the earlier stake is called sunk.
Assuming the bias only involves money. Time, research and emotional attachment to a team or a story all count, and non-monetary investment is harder to notice because it never appears in a bankroll record.
Sunk Cost vs. Gambler's Fallacy, Loss Aversion and Chasing Losses
Several biases overlap with this one, and separating them helps choose the right fix. The gambler's fallacy is a belief about probability: that past independent outcomes make a different result "due." The sunk cost fallacy is about value and commitment, not a belief that the odds changed; a bettor can know a longshot is unlikely and still hold it because of what it cost.
Loss aversion explains part of the pull: losses feel larger than equal gains, and the reference point is often the stake. Sunk cost thinking is what happens when that feeling is allowed to steer the next choice, and chasing losses is the escalation that follows when the bettor raises stakes or frequency to get back to even. The first is a distortion in evaluating a position; the second is a behavior that increases risk.
In daily practice the fallacy appears in three places: the live ticket screen with a cash-out button, the futures page with an old ticket, and the research notebook with hours of work behind a pick. The decision process is the same in each: write the current price, write the estimate, and decide as if starting fresh. If the urge is to "make it back," step away. If gambling is affecting your finances or wellbeing, use your sportsbook's limit tools or a support service.
Related Concepts and Next Steps
A sensible next step is to review your own record. Look at the last ten tickets you held or added to after they went against you, and check whether you could have written a forward-looking reason at the time. A written reason is the difference between a decision and a reaction. Readers building a full picture of bettor psychology can follow the learning sequence from the overview of common biases, to loss aversion, to chasing losses, and then back to this forward-looking test, which applies to every one of them.
Frequently Asked Questions
