🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
st_petersburg_paradox
The St. Petersburg paradox describes a gamble with theoretically infinite expected value that virtually no rational person would pay much to play. This demonstrates that expected value alone cannot fully capture rational decision-making under risk. It reveals the necessity of utility functions, risk aversion, and the diminishing marginal value of wealth in modeling real decisions.
A casino offers: flip a fair coin until you get tails. If tails on first flip, win $2; second flip, $4; third flip, $8; etc. Expected payout = infinite. Yet most people would pay less than $20 to play this game, because the utility of the nth dollar is far less than the first.
A startup pitches investors with a deal: invest $1, and if their product goes viral in week 1 you get $2, week 2 you get $4, week 3 you get $8, doubling each week indefinitely. The mathematical expected return is infinite, yet every seasoned investor in the room is willing to put in no more than a few hundred dollars — because the astronomically large payouts require astronomically unlikely streaks.
An online lottery advertises: roll a die repeatedly until you roll a six; win $6 for rolling it on the first try, $36 on the second, $216 on the third, multiplying by 6 each round. The expected value is technically infinite, but when surveyed, participants say they'd pay an average of only $15 to enter — because the chance of surviving enough rounds to collect life-changing money feels vanishingly small.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Is expected monetary value used as the sole criterion for evaluating a risky decision?
Type: binaryDoes the decision involve outcomes with extreme variance or heavy tails that have low probability but enormous magnitude?
Type: binaryIs diminishing marginal utility of wealth or risk aversion being ignored?
Type: binaryIs the claim that a strategy is 'optimal' based solely on maximizing expected value?
Type: binaryThe St. Petersburg paradox describes a gamble with theoretically infinite expected value that virtually no rational person would pay much to play. This demonstrates that expected value alone cannot fully capture rational decision-making under risk. It reveals the necessity of utility functions, risk aversion, and the diminishing marginal value of wealth in modeling real decisions.
The mathematical expectation is infinite, but the probability of large payoffs decreases exponentially. The psychological and economic value of the nth dollar is far less than the first, making the theoretically infinite expectation practically irrelevant.
Use expected utility rather than expected value for risk decisions. Consider variance, skewness, and tail probabilities in addition to means. Recognize that most people are risk-averse and that this is rational, not irrational.
Critiques of extreme tail-risk insurance and catastrophe bond pricing often invoke St. Petersburg-type reasoning. It also underlies critiques of risk-neutral pricing in financial derivatives.
Use these tools to detect, analyze, or train this aspect.