🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
ellsberg_paradox
The Ellsberg paradox reveals that people systematically prefer bets with known probabilities over bets with unknown probabilities (ambiguity aversion), even when expected values are identical. This violates subjective expected utility theory. Ambiguity aversion is distinct from risk aversion: it is a preference for known risk over unknown risk.
An urn contains 30 red balls and 60 balls that are either black or yellow in unknown proportion. Most people prefer to bet on red (known probability 1/3) over black (unknown probability), and also prefer to bet on black-or-yellow over red-or-yellow — a pattern inconsistent with any subjective probability assignment.
A fund manager is offered two investments: one with a clearly documented 40% historical success rate, and one in an emerging market where the success rate is completely unknown. Even if the unknown investment might have a higher success rate, the manager consistently chooses the known-probability option — paying a premium simply to avoid uncertainty.
In a game show, a contestant can draw from Urn A (50 red, 50 blue balls) or Urn B (100 balls, unknown mix of red and blue). Most contestants bet on red from Urn A rather than Urn B, and also prefer to bet on blue from Urn A rather than Urn B — even though Urn B's unknown composition could theoretically be more favorable.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Does the decision involve options with known versus unknown probability distributions?
Type: binaryDoes the person show a consistent preference for known-probability bets over unknown-probability bets, even at equivalent expected values?
Type: binaryCould the preference be explained by risk aversion alone, or does it require ambiguity aversion?
Type: binaryIs subjective expected utility theory being used without accounting for Knightian uncertainty?
Type: binaryThe Ellsberg paradox reveals that people systematically prefer bets with known probabilities over bets with unknown probabilities (ambiguity aversion), even when expected values are identical. This violates subjective expected utility theory. Ambiguity aversion is distinct from risk aversion: it is a preference for known risk over unknown risk.
Humans are averse to not knowing the odds. Unknown probabilities feel more threatening than known risks, even when the expected values are the same. This is sometimes called Knightian uncertainty aversion.
Distinguish between risk (known probability distribution) and uncertainty (unknown distribution). Use maxmin expected utility or other ambiguity-tolerant frameworks when decisions involve unknown probabilities.
Ellsberg-type effects explain investor home bias (preference for domestic stocks with known return distributions), reluctance to enter unfamiliar markets, and excessive insurance purchasing against uncertain events.
Use these tools to detect, analyze, or train this aspect.