--- name: prospect-theory description: Use when asked to understand or predict how people actually judge risky choices — evaluating outcomes as gains and losses relative to a reference point, feeling losses more than equal gains, overweighting small probabilities, and being swayed by framing — in order to debug one's own decisions, design offers, or explain surprising behavior; distinct from [[expected-value-analysis]] (the rational benchmark it departs from), [[game-theory]] (strategic interaction), and [[decision-analysis]] (the structured process it can inform). --- # Prospect theory Prospect theory, developed by Daniel Kahneman and Amos Tversky, describes how people really choose under risk, as opposed to how expected-value theory says they should. Its core findings: people judge outcomes as gains or losses from a *reference point*, not as final wealth; losses hurt roughly twice as much as equal gains please (*loss aversion*); sensitivity diminishes as amounts grow; small probabilities are overweighted and moderate-to-large ones underweighted; and the same choice framed differently produces different decisions. Using it means spotting these patterns in a decision, whether yours or someone else's, and correcting or designing for them. ## Input required - **The choice** and the options, with outcomes and probabilities. - **The reference point** the decision-maker is using (status quo, expectation, sunk cost, a peer's result). - **How the options are framed** — as gains or losses, in what wording. - **Who is deciding and for whom** — oneself, a team, customers. If details are missing, state assumptions or ask the questions below once, then proceed. ## Questions to ask first Ask in **one batch of at most five questions**, skipping any already answered, then proceed: 1. What are the options, with outcomes and probabilities? 2. What does the decider treat as "normal" or "break-even" right now? 3. Is any option described as avoiding a loss versus gaining a benefit? 4. Is there a sunk cost, a deadline, or a recent loss influencing the choice? 5. Is the aim to improve your own decision or to design an offer or message for others? ## The core ideas - **Reference dependence** — outcomes are coded as gains or losses relative to a reference point; moving the reference point changes the choice. - **Loss aversion** — a loss looms larger than an equal gain, by about 2:1 in many studies. - **Diminishing sensitivity** — the difference between $0 and $100 feels larger than between $1,000 and $1,100. - **Risk attitudes flip** — typically *risk-averse for gains* (take the sure $500 over a 50% shot at $1,000) and *risk-seeking for losses* (gamble to avoid a sure loss). - **Probability weighting** — tiny chances are overweighted (lottery tickets, insurance for rare events); near-certainty is underweighted; the *certainty effect* makes a sure thing special. - **Framing** — "90% survival" versus "10% mortality" gives the same facts but different choices. - **Related effects** — endowment effect (valuing what one owns more), sunk-cost behavior (throwing good money after bad), status quo bias, and mental accounting. ## Method 1. **Restate the decision in neutral terms**: final outcomes and probabilities, no gain/loss language. 2. **Identify the reference point** being used and which outcomes feel like gains and which like losses. 3. **Find the distortions.** Check for loss aversion keeping the status quo, risk-seeking to dodge a sure loss, overweighted rare events, and sunk costs steering the choice. 4. **Re-frame several ways.** Evaluate the options with different reference points (current state, goal, outside view) and as final totals. If the preference flips, the framing is driving it. 5. **Compare with the rational benchmark.** Compute the [[expected-value-analysis]] and ask whether the preferred option differs, and whether the difference is justified by real risk limits, not just feeling. 6. **Decide on purpose.** Keep the intuition if it reflects real values (a ruinous loss deserves extra weight); correct it if it reflects framing. 7. **For designing offers or communications**, apply the findings honestly: present information clearly, make defaults sensible, and avoid exploiting biases against the audience's interests. ## Output format - **Neutral restatement** of the choice. - **Reference point and frame** currently in play. - **Distortions detected**, each with how it shows up. - **Reframed comparison** — does the preference change? - **Benchmark** — expected value and risk, versus the intuitive choice. - **Recommendation** — what to do, and the reasoning that survives reframing. ## Ground rules - **Describe, don't scold.** These patterns are normal human psychology, not stupidity. - **Not every deviation is an error.** Loss aversion can be rational when a loss threatens survival. - **Effects vary.** The 2:1 ratio is an average from experiments; individuals, stakes, and cultures differ. - **Don't manipulate.** Use the theory to inform and to remove distortions, not to push people into choices against their interests. ## Self-check before answering - Was the choice restated without gain/loss framing? - Were reference point, loss aversion, probability weighting, and sunk costs each considered? - Was the preference tested under at least one alternative frame? - Is the advice separate from the description of the bias? ## Common pitfalls - **Treating the status quo as risk-free.** Doing nothing has costs. - **Sunk-cost chasing** — continuing because of past investment. - **Overweighting dramatic, rare events** — and neglecting common ones. - **Gambling to avoid a sure loss** — turning a small loss into a large one. - **Framing blindness** — believing the wording did not matter. - **Overgeneralizing the research** — applying lab averages to a specific person as if they were laws. ## Example prompt > We've spent $400k on a project that's behind schedule. Management > says "we can't stop now." Use prospect theory to explain what biases > may be at work, re-frame the choice, and tell me what the decision > looks like with sunk costs ignored. ## Learn more - [[expected-value-analysis]] for the rational benchmark. - [[decision-analysis]] for the end-to-end decision process. - [[game-theory]] for strategic choices among interacting parties. - [[regret-minimization]] for decisions shaped by anticipated regret. - [[decision-journal]] for catching your own patterns over time. - [[premortem]] for countering optimism and sunk-cost commitment.