Risk Preferences and Prospect Theory: Neural, Evolutionary, and Behavioral Insights

Risk Aversion for Gains

People tend to prefer a guaranteed outcome over a risky one when both have the same expected value. For example, most people would rather take $50 for certain than a 50/50 shot at $100. This preference flips when facing losses.

Risk Seeking for Losses

When facing potential losses, people tend to prefer the risky option over a guaranteed loss of the same expected value. Someone would rather gamble on a 50/50 chance of losing $100 than accept a certain $50 loss. This asymmetry between gains and losses is a core feature of Prospect Theory.

Lakshminarayanan et al.

Researchers gave capuchin monkeys choices between safe and risky options framed as either gains or losses. The monkeys showed risk aversion when outcomes were framed as gains and risk seeking when framed as losses — the same pattern humans show. This supports the idea that loss aversion and framing effects may have deep evolutionary roots.

Rabin's Calibration Theorem

A mathematical proof showing that if someone turns down a small, even-odds gamble out of risk aversion, then by the same logic they should turn down absurdly large gambles too — which no rational person would do. It reveals that expected utility theory cannot explain everyday risk aversion without producing ridiculous predictions at higher stakes. This suggests we need a fundamentally different model, like Prospect Theory, to explain human behavior.

Tom et al.

Participants made risky choices while brain activity was recorded using fMRI. Activity in the striatum was more responsive to potential losses than to equivalent gains, and this asymmetry in brain activity predicted how loss averse each individual was in their choices. This provides direct neural evidence for loss aversion.

Hsu et al.

Researchers similarly looked at striatum activity during risky decision-making and found that its response to probabilities followed an inverse-S shape. This means the brain is oversensitive to small and large probabilities, and undersensitive to middle-range ones. This mirrors the probability weighting function described in Prospect Theory.

Apicella et al.

Researchers found that men with higher testosterone levels were more willing to accept a gamble offering a large gain with a 50% chance and a smaller loss with a 50% chance. This suggests biological factors like hormones play a role in financial risk-taking. It supports the broader idea that risk preferences are not purely rational calculations but are shaped by physiology.

Weber-Fechner Law

The psychological response to a stimulus grows with the logarithm of the stimulus, not its absolute size. In plain terms, the difference between $10 and $20 feels much bigger than the difference between $1,010 and $1,020, even though both are a $10 gap. This underlies the concept of diminishing sensitivity in Prospect Theory and appears across nearly every domain of human perception.

Diminishing Sensitivity

The idea that each additional unit of gain or loss has less psychological impact than the one before it. Winning $100 feels great, but winning $101 barely registers on top of that. This produces the characteristic S-shaped value function in Prospect Theory.

Extinction Probability

An evolutionary argument for why loss aversion may exist. If an organism faces a 50/50 gamble of gaining or losing one offspring, repeated losses could wipe out the lineage entirely, while gains have a natural ceiling. This asymmetry in survival consequences would have selected for organisms that weight losses more heavily than equivalent gains.

Probability Heuristic

Most people don't think carefully about exact probabilities and instead use a rough mental shortcut that treats any gamble as simply "risky" versus "safe." Rather than distinguishing between a 5% and 10% chance, they essentially lump all non-certain outcomes into one mental bucket. This helps explain why small probabilities are often overweighted in decision-making.

Affect and Probability Judgment

Emotional state colors how likely people think bad or good events are. When in a positive mood, people tend to think good things are more probable; when in a negative mood, they inflate the perceived likelihood of bad outcomes. This is related to the broader inverse-S probability weighting pattern seen in Prospect Theory.

Johnson and Tversky

Participants were put into positive, negative, or neutral moods by reading emotionally charged stories, then asked to estimate the probabilities of various real-world events. Those in a negative mood rated bad events as significantly more likely, and those in a positive mood shifted in the opposite direction. This supports the idea that affect directly distorts probability judgment rather than just emotional response to outcomes.

Haisley et al.

Researchers tested whether offering employees a lottery ticket was more motivating than a gift card of equal expected value for completing a health risk assessment, in a study of nearly 1,300 employees. The lottery turned out to be at least as effective — and for some groups more so — despite having the same or lower expected monetary value. This supports the idea that probability overweighting of small chances makes lotteries feel more attractive than their actuarial value suggests.

Overround

The built-in profit margin that bookmakers add by setting odds so that all implied probabilities across a set of outcomes add up to more than 100%. For example, in a two-team game, a fair market would have both sides at 50% (totaling 100%), but a bookmaker might price each at 52%, guaranteeing profit over time regardless of the outcome. It is the mechanism by which gambling markets systematically favor the house.

Mellers et al.

Participants were given individual gambles, saw the outcomes played out, and then rated how they felt on a scale from very negative to very positive. Emotional responses depended not just on what actually happened, but also on what could have happened — a $50 gain felt better when the alternative was winning nothing than when it was winning $100. This demonstrates counterfactual thinking and supports the role of regret and elation in decision-making under risk.

Schultz et al.

Researchers recorded dopamine neuron activity in monkeys across three situations: receiving an unexpected reward, receiving an expected reward, and not receiving a reward that was expected. Dopamine fired strongly for unexpected rewards, showed no change for expected ones, and dipped below baseline when an expected reward failed to arrive. This is the neural basis of prediction error, supporting the idea that the brain tracks the gap between expectations and outcomes rather than outcomes themselves.

Hertwig et al. (Description vs. Experience)

When participants learned about gambles through direct experience rather than being told the probabilities upfront, their behavior flipped relative to standard Prospect Theory predictions. Instead of overweighting small probabilities, they underweighted them, and instead of underweighting large probabilities, they overweighted them. This suggests the classic probability weighting function may be specific to described gambles, and that how we learn about risk matters enormously.