Skip to main content
Category: Paradoxes
Type: Decision Theory Paradox
Origin: Introduced by Nicolas Bernoulli in the early 18th century and analyzed by Daniel Bernoulli in 1738
Also known as: St. Petersburg Lottery, Bernoulli’s Paradox, Paradox of Infinite Expectation
Quick Answer — The St. Petersburg Paradox is a betting game in which the expected monetary value is infinite, yet almost nobody would pay a very high price to play. A fair coin is tossed until the first heads appears; the pot doubles each time tails appears. Mathematically, the average payout diverges to infinity, but real people treat the game as worth only a limited amount—forcing decision theory to distinguish money from subjective utility and to confront how we actually feel about risk.

​
What is the St. Petersburg Paradox?

The St. Petersburg Paradox is a famous puzzle at the intersection of probability, economics, and psychology. It starts with a simple lottery: a casino flips a fair coin until it lands heads for the first time. If heads appears on the first toss, you win 2 units of money. If it appears on the second toss, you win 4; on the third, 8; on the fourth, 16; and so on, doubling each time. From a purely mathematical perspective, the expected payoff of this game is unbounded. The chance of heads on the first toss is 1/2 and pays 2, contributing 1 to the expected value. The chance of heads on the second toss is 1/4 and pays 4, contributing another 1. Each additional possible outcome contributes 1 as well, leading to an infinite sum. On paper, the game seems infinitely valuable. Yet when you ask real people how much they would pay for a single ticket, answers are modest—often equivalent to a normal night out or less. Even very wealthy, risk-tolerant individuals rarely offer enormous sums. This gap between infinite expected value and finite willingness to pay is the heart of the St. Petersburg Paradox. It reveals that “expected money” and “experienced value” are not the same thing, and that our intuitions about risk demand a richer model than naive expected value.
“The St. Petersburg Paradox shows that we do not maximize expected money; we care about the usefulness of money, our starting wealth, and the pain of large risks. It pushed decision theory toward utility, diminishing returns, and more realistic models of human behavior.”

​
St. Petersburg Paradox in 3 Depths

  • Beginner: Imagine a game where you could, in theory, win an enormous amount of money, but the chance is tiny. Even if math says the “average” outcome is infinite, your gut tells you not to risk your life savings for a ticket. The paradox captures this clash between the math of unlikely jackpots and human common sense.
  • Practitioner: In investing, product pricing, or insurance, you constantly face low-probability, high-impact events. The St. Petersburg Paradox explains why decision-makers cap their exposure, use insurance, or impose position limits, even when raw expected value looks attractive: they care about volatility, downside risk, and how additional money changes their life.
  • Advanced: Daniel Bernoulli’s solution introduces a concave utility function—classically logarithmic—to model diminishing marginal utility of wealth, turning the infinite expected monetary value into a finite expected utility. Later work shows that simple utility curves alone do not resolve every variant of the paradox, motivating more nuanced treatments of risk, time, and bounded rationality.

​
Origin

The paradox originated in correspondence among the Bernoulli family in the early 18th century. Nicolas Bernoulli first proposed the lottery that would come to be known as the St. Petersburg game, named after the city where many of these ideas were discussed and published. He framed it as a challenge to conventional notions of fair price and rational gambling behavior. Daniel Bernoulli, Nicolas’s cousin, provided the most famous analysis in a 1738 paper on risk and utility. Observing that people clearly did not treat expected monetary value as the sole decision criterion, he argued that the psychological value—or utility—of money grows more slowly than the amount itself. Doubling wealth does not double happiness, especially for already wealthy individuals. By modeling utility as a logarithmic function of wealth and proposing that rational agents maximize expected utility rather than expected money, Daniel Bernoulli offered a principled explanation of why people might value the St. Petersburg game at a finite level. His work became a cornerstone of modern decision theory and economics, influencing everything from portfolio choice to insurance pricing and the study of risk aversion.

​
Key Points

Several structural insights make the St. Petersburg Paradox enduringly useful rather than a mere curiosity.
1

Infinite Expectation vs. Human Intuition

The game’s formal expected payoff is infinite, but human willingness to pay is finite and usually modest. This tension shows that expected monetary value alone is a poor guide to real-world decisions where ruin and large fluctuations matter.
2

Diminishing Marginal Utility of Wealth

As wealth increases, the extra “felt value” of each additional unit of money shrinks. Losing half your wealth hurts more than gaining an extra half feels good. Concave utility functions—like the logarithm Bernoulli proposed—capture this pattern and produce finite expected utility for the game.
3

Risk Aversion and Position Sizing

The paradox highlights that rational agents may rationally reject favorable gambles when they are too large relative to their wealth or risk tolerance. Sound decision-making requires sizing bets and limiting exposure, not just chasing high expected value.
4

Limits of Simple Models

Later critiques show that even utility-based solutions have edge cases and variants of the paradox they cannot fully tame. This pushes decision theory toward richer models incorporating reference points, loss aversion, and bounded rationality, as seen in behavioral economics.

​
Applications

The ideas behind the St. Petersburg Paradox inform many real-world choices about risk and reward.

Investment and Portfolio Management

Professional investors use position limits, diversification, and risk budgets to avoid “St. Petersburg–style” exposure—small chances of huge losses or gains that could destabilize long-term plans. Utility-like thinking underlies concepts such as Kelly sizing and risk-adjusted return metrics.

Insurance and Risk Transfer

People buy insurance even when the expected monetary value is negative, because smoothing outcomes and avoiding catastrophic losses has high utility. The paradox illustrates why paying a predictable premium can be rational despite lower expected money.

Pricing Lotteries and Extreme Bets

Lotteries, speculative options, and meme assets often offer extremely skewed payoff distributions. The St. Petersburg framework helps explain why some people overpay for tiny chances at windfalls, while others avoid such bets entirely, and why regulators worry about exposure to “tail events.”

Personal Finance and Career Choices

Individuals routinely trade maximum expected income for stability and well-being—choosing a steady job over a highly uncertain startup path, or diversifying skills instead of making one all-or-nothing bet. The paradox gives a crisp mathematical illustration of why this can be rational.

​
Case Study

Suppose an investor with moderate wealth is offered the St. Petersburg game by a trusted counterparty. She can buy a ticket for any price she chooses. A naive expected-value calculation suggests she should be willing to pay any finite amount, since the game’s expected payoff is infinite. Instead, she models her preferences using a utility function that reflects her risk tolerance—for example, a logarithmic utility over total wealth. She then computes the expected utility of playing the game at various ticket prices and compares it to the utility of keeping her current wealth. For low prices, the added “lottery upside” increases her expected utility; for high prices, the possibility of losing a substantial fraction of her wealth outweighs the benefit of a very unlikely enormous jackpot. Numerically, this analysis yields a finite “indifference price”: a ticket cost above which the expected utility of playing is lower than that of not playing at all. In practice, this price is often modest relative to her wealth. The case shows that a rational, forward-looking agent, armed with a realistic utility function, can justifiably refuse to pay large sums for a game with infinite expected value—capturing the spirit of Bernoulli’s original solution.

​
Boundaries and Failure Modes

The St. Petersburg Paradox comes with important caveats that prevent overextending its lessons.
  1. Utility Models Are Approximations, Not Psychological Truths: Concave utility functions are tools for modeling risk preferences, not literal descriptions of happiness. Over-interpreting a specific functional form can mislead, especially when extrapolated far beyond observed wealth levels.
  2. Real-World Games Have Implicit Caps: In practice, no casino can pay arbitrarily large sums, and players have limited time and capital. Once you impose realistic caps on maximum payout or number of coin tosses, the expected monetary value becomes finite, and the paradox softens.
  3. Misuse: Justifying Any Risk Aversion as ‘Rational’: It is tempting to invoke the paradox to defend extreme conservatism in all decisions. But excessively concave utility can itself lead to dominated choices, such as rejecting modest, repeated favorable bets that would almost surely improve long-run outcomes.

​
Common Misconceptions

Over time, the St. Petersburg Paradox has attracted some persistent misunderstandings.
Reality: Expected value remains crucial in many domains, especially when payoffs are bounded and stakes are small relative to wealth. The paradox shows that for extreme, highly skewed gambles, expected value must be combined with risk preferences and context.
Reality: While expected utility theory offers a powerful response, certain variants of the game and empirical behavior highlight its limits. The paradox is best seen as a gateway into a richer understanding of risk, not as a one-time puzzle that utility theory closes forever.
Reality: Some tail risks are worth taking—especially when they are small, repeated, and have limited downside. The key is to size positions prudently and consider the whole distribution of outcomes, not to avoid low-probability events altogether.
The St. Petersburg Paradox is tightly linked to modern ideas about risk and value.

Expected Value

The simple average payoff of a gamble, weighting each outcome by its probability. The paradox shows its limitations for unbounded or highly skewed distributions.

Expected Utility Theory

The framework, inspired by Bernoulli, in which agents maximize expected utility rather than expected money. It remains the backbone of much of modern decision theory.

Risk Aversion

A preference for more certain outcomes over riskier ones with the same expected value. The paradox provides a vivid illustration of why risk aversion can be rational.

Kelly Criterion

A formula for sizing bets to maximize long-term growth while respecting risk. It embodies the idea that both expectation and volatility matter, echoing St. Petersburg–style concerns.

Black Swan Events

Rare, high-impact events in markets and systems. The paradox prepares us to think carefully about distributions with heavy tails and extreme payoffs.

Expected Value in Models

Concepts like /models/expected-value and related decision models in this atlas highlight where naive expectation works well and where it needs to be supplemented with richer tools.

​
One-Line Takeaway

The St. Petersburg Paradox reminds us that good decisions depend not just on mathematical expectation but on how outcomes feel relative to our wealth, goals, and tolerance for risk—pushing us from “maximize expected money” toward richer, utility-aware thinking.