St Petersburg Paradox

1. Introduction

The St Petersburg Paradox is a classic puzzle at the intersection of probability theory, economics, and philosophy. It concerns a highly stylized lottery—the St Petersburg lottery—that has an infinite expected monetary value, yet appears to justify only a modest willingness to pay for most agents, including those who understand the mathematics involved. This apparent gap between formal expectation and intuitive valuation has been used to challenge simple versions of the principle that rational agents should maximize expected monetary gain.

The paradox is often presented as a reductio ad absurdum of naïve expected value reasoning. When the standard formula for expectation is applied to the lottery, it yields a divergent series. However, when individuals are asked how much they would actually pay to play such a game once, their answers tend to be finite and comparatively small. The tension arises when both of the following are taken seriously: the mathematical result that the game is “infinitely valuable” in monetary terms, and the intuition (and observed behavior) that paying arbitrarily large amounts to play would be irrational.

Subsequent sections examine how this puzzle has influenced theories of utility, risk aversion, and decision under uncertainty, and how it has motivated alternative norms for choice in cases involving infinite expectations. The paradox’s enduring importance lies less in the particular coin-toss game and more in the conceptual issues it raises about how to model preferences, value extremely unlikely outcomes, and constrain theoretical principles so that they align with both mathematical coherence and plausible rational behavior.

2. Origin and Attribution

The origin of the St Petersburg Paradox lies in early 18th‑century exchanges among mathematicians working on the emerging theory of probability. The game itself, and the problem it generates, are generally attributed to members of the Bernoulli family, especially Nicolas Bernoulli and Daniel Bernoulli.

2.1 Nicolas Bernoulli’s Formulation

The earliest known formulation appears in a 1713 letter from Nicolas Bernoulli to the French mathematician Pierre Rémond de Montmort. In this correspondence, Nicolas described a coin-tossing game whose payoffs doubled with each additional tail and asked how much a “reasonable man” should pay to participate. This letter is taken by historians as the first clear statement of the problem.

"

“I have proposed a certain problem … in which the expectation appears infinite, yet no one would stake a large sum on it.”

— paraphrasing Nicolas Bernoulli’s 1713 letter to de Montmort

Nicolas did not fully develop a resolution, but he highlighted the mismatch between infinite expectation and finite valuation as an objection to simple expectation-based decision rules.

2.2 Daniel Bernoulli’s Analysis

Daniel Bernoulli, a cousin of Nicolas, is credited with giving the paradox its canonical treatment. In his 1738 paper:

"

“Specimen theoriae novae de mensura sortis”
(Commentarii Academiae Scientiarum Imperialis Petropolitanae)

he analyzed the lottery in detail and proposed a utility-based response. This work, read to the St Petersburg Academy in 1731, gave the problem its enduring name and associated it with logarithmic utility.

2.3 Naming and Canonization

The label “St Petersburg Paradox” or “St Petersburg Game” reflects Daniel Bernoulli’s affiliation with the Imperial Academy of Sciences in St Petersburg at the time of his analysis, rather than the location of the game’s invention. Over time, textbooks and later commentators came to attribute the paradox primarily to Daniel, while more specialized historical scholarship has emphasized Nicolas’s priority in posing the problem.

3. Historical Context

The St Petersburg Paradox emerged during a formative period for probability theory and economic thought in early modern Europe. It reflects intellectual currents in the early 18th century surrounding games of chance, insurance, and rational decision‑making under risk.

3.1 Probability, Games, and the Bernoulli Tradition

By the time Nicolas Bernoulli proposed the game (1713), foundational work by Pascal, Fermat, Huygens, and Jacob Bernoulli had already established combinatorial methods for calculating probabilities. Jacob Bernoulli’s Ars Conjectandi (published posthumously in 1713) framed probability as a guide to “reasonable expectation,” linking games of chance with broader questions about rational belief and fair pricing.

Within this context, Nicolas Bernoulli’s puzzle can be viewed as testing the limits of the idea that expected value could serve as the general measure of a gamble’s worth. The St Petersburg game was part of a broader use of idealized lotteries to probe the norms of rational behavior.

3.2 Early Economic Concerns: Insurance and Utility

The early 18th century also saw growing interest in insurance, annuities, and life contingencies, where decisions involved uncertain future payments. Mathematicians such as Abraham de Moivre were applying probability to these financial questions, and there was increasing recognition that the same amount of money could have different significance to people in different circumstances.

Daniel Bernoulli’s later analysis of the paradox is closely tied to this milieu. His paper was explicitly motivated by practical issues in maritime insurance and investment, rather than by gambling per se. The paradox thus sits at the junction where mathematical probability met nascent ideas about utility and risk attitudes.

3.3 Intellectual Networks and the St Petersburg Academy

The paradox was also shaped by institutional and geographical factors. Correspondence networks among mathematicians—linking Basel, Paris, London, and St Petersburg—served as the medium through which problems and solutions circulated. Daniel Bernoulli’s membership in the Imperial Academy of Sciences in St Petersburg provided the setting for his 1731 presentation, which helped canonize the paradox as a “St Petersburg” problem.

Within this historical environment, the St Petersburg Paradox became a test case for how far emerging probabilistic and economic theories could be extended without revision.

4. The St Petersburg Lottery Described

The St Petersburg lottery is a stylized gambling game defined by a simple rule involving repeated coin tosses and exponentially increasing payoffs. Its structure is crucial for generating an infinite expected monetary value.

4.1 Rules of the Game

The standard version proceeds as follows:

1. A fair coin is tossed repeatedly until it lands heads for the first time.
2. Let n be the toss number on which the first heads appears (n = 1, 2, 3, …).
3. The player’s payoff is 2ⁿ monetary units (e.g., ducats).

Thus:

The defining feature is the doubling of the payoff with each additional tail before the first heads.

4.2 Entry Fee and One‑Shot Play

Players are asked to name a maximum entry fee they are willing to pay for a single play of this game. There is no repetition or opportunity to average outcomes over many plays within the standard paradoxical setup; the decision concerns a one‑time participation.

Formally, if c denotes the cost of entry, then:

• If the player does not play, their wealth remains w₀.
• If the player plays, their final wealth is w₀ − c + 2ⁿ, where n is determined by the coin toss sequence.

The paradox arises when this simple lottery structure is combined with the classical rule for computing expectation and a normative principle about how rational agents should set the fee c. Those issues are addressed in subsequent sections; here, the essential ingredients are the geometric decay of probabilities and the geometric growth of payoffs that, together, create the distinctive payoff profile of the St Petersburg game.

5. Formal Statement of the Paradox

The paradox is typically formulated by juxtaposing two claims: a mathematical result about the game’s expected monetary value and a behavioral or normative claim about what rational agents ought to be willing to pay.

5.1 Mathematical Setup

Let:

• X be the random variable representing the monetary payoff of the St Petersburg lottery.
• n ∈ {1, 2, 3, …} denote the toss on which the first heads occurs.
• P(n) = (1/2)ⁿ (for a fair coin).
• Payoff(xₙ) = 2ⁿ.

Then:

[ X = 2^n \text{ with probability } P(n) = \left(\frac{1}{2}\right)^n. ]

The expected monetary value (EMV) of this game is:

[ \mathbb{E}[X] = \sum_{n=1}^{\infty} P(n) \cdot 2^n = \sum_{n=1}^{\infty} \left(\frac{1}{2}\right)^n \cdot 2^n = \sum_{n=1}^{\infty} 1, ]

which diverges to infinity.

5.2 Normative Principle Involved

A common classical principle of rational choice—especially before expected utility theory—states that:

"

A rational agent should be willing to pay up to a gamble’s expected monetary value for a fair ticket.

Applied here, this suggests that a rational agent should, in principle, accept any finite entry fee to play the St Petersburg game once, because the EMV is infinite.

5.3 The Tension

The paradox arises from the contrast between:

Formally, one obtains the conditional:

• If rational agents should maximize expected monetary value,
• and if the St Petersburg game has infinite EMV,
• then rational agents should be willing to pay arbitrarily large finite sums to play.

Yet this implication conflicts with typical judgments about what it is reasonable to pay, yielding the paradoxical situation that is explored and interpreted in the subsequent sections.

6. Logical Structure and Argument Form

The St Petersburg Paradox is generally understood as a reductio ad absurdum targeting a specific decision rule—maximization of expected monetary value (EMV). Its logical structure highlights how unqualified application of that rule to the St Petersburg lottery produces an apparently unacceptable conclusion.

6.1 Core Argument Pattern

The argument may be schematically reconstructed as follows:

1. Setup of the lottery: The St Petersburg game is defined so that payoff on the n‑th toss is 2ⁿ with probability (1/2)ⁿ.
2. Calculation step: Using the standard expectation formula, EMV is the series
   (\sum_{n=1}^{\infty} (1/2)^n \cdot 2^n), which diverges to infinity.
3. Normative premise: A rational agent should be prepared to pay up to a gamble’s EMV for one play of that gamble.
4. Derivation: Therefore, a rational agent ought to be willing to pay any finite amount to play the St Petersburg game once.
5. Conflict with intuition/behavior: In fact, typical reflective agents regard paying very large sums as irrational; empirically, people state only modest willingness to pay.
6. Paradoxical conclusion: At least one of the premises (2–4) must be revised or restricted.

6.2 Reductio Character

The reasoning thus fits the reductio ad absurdum pattern:

• Assume the unqualified EMV‑maximization principle.
• Combine it with straightforward probabilistic reasoning about the game.
• Derive the conclusion that rationality requires arbitrarily high willingness to pay.
• Treat this as absurd or at least highly implausible.
• Conclude that the starting principle (or some background assumption) must be rejected or amended.

Different authors locate the “fault” in different places: some challenge the EMV principle itself; others question how the lottery is modeled; still others dispute the claim that the conclusion is genuinely irrational. But the structure of the paradox is generally framed as: from plausible assumptions, an implausible requirement is derived, so some part of the system must be reconsidered.

7. Expected Monetary Value and Infinite Expectations

The St Petersburg Paradox places the concept of expected monetary value (EMV) under scrutiny by exploiting a case with an infinite expectation. This section focuses on the mathematical and conceptual issues surrounding such expectations.

7.1 Classical EMV Calculation

In classical probability theory, the EMV of a discrete gamble with outcomes xᵢ and probabilities pᵢ is:

[ \mathbb{E}[X] = \sum_i p_i x_i. ]

For typical lotteries with bounded payoffs, this sum converges to a finite number that can be interpreted as a “fair price” under certain idealized conditions (such as risk neutrality and linear utility of money).

In the St Petersburg game, however, EMV is:

[ \mathbb{E}[X] = \sum_{n=1}^\infty \left(\frac{1}{2}\right)^n 2^n = \sum_{n=1}^\infty 1 = \infty. ]

This is an example of an infinite expectation: the series does not converge to a finite real number.

7.2 Conceptual Status of Infinite Expectations

Infinite expectations raise questions about how EMV should be interpreted:

• Some readings treat EMV as a fair price: but an “infinite fair price” is not actionable, since an agent cannot pay more than any given finite amount.
• Others see EMV as an average gain per play in the long run: yet for infinite expectation, the usual law-of-large-numbers interpretations become delicate, and the practical meaning of “average” is less clear.

In response, commentators have distinguished between:

The paradox thus serves as a test case for whether EMV can serve as a universal decision criterion, especially in the presence of lotteries with unbounded or heavy-tailed payoffs. Later developments, such as expected utility theory, emerged partly to refine or replace EMV as the primary normative metric in such contexts.

8. Bernoulli’s Utility-Based Resolution

Daniel Bernoulli’s 1738 analysis proposed a now‑canonical way to address the paradox by shifting from expected monetary value to expected utility. His key idea was that the subjective value—or utility—of money does not increase linearly with its amount.

8.1 Logarithmic Utility of Wealth

Bernoulli suggested that the utility of wealth w could be represented by a logarithmic function, U(w) = ln(w). This captures the idea that a gain of 100 ducats matters more to a poor person than to a rich one, reflecting diminishing marginal utility.

In his own words:

"

“The utility resulting from any small increase in wealth will be inversely proportional to the quantity of goods previously possessed.”

— Daniel Bernoulli, Specimen theoriae novae de mensura sortis

Applied to the St Petersburg lottery, the expected utility of playing (starting from wealth w₀ and paying cost c) is:

[ \mathbb{E}[U] = \sum_{n=1}^\infty \left(\frac{1}{2}\right)^n \ln\left(w_0 - c + 2^n\right). ]

This sum converges for reasonable values of w₀ and c, yielding a finite expected utility.

8.2 Finite Rational Willingness to Pay

The rational entry fee c*, on Bernoulli’s view, is determined by solving:

[ \mathbb{E}[U(w_0 - c^* + X)] = U(w_0), ]

i.e., the fee at which the expected utility of playing equals the utility of not playing. With U(w) = ln(w), Bernoulli’s calculations produced a finite c*, consistent with people’s modest willingness to pay.

Thus, on this approach:

8.3 Conceptual Shift

Bernoulli’s proposal introduced two important shifts:

1. From money to utility: Rational choice concerns maximizing expected utility, not money.
2. From linear to concave valuation: Utility is concave in wealth, embodying risk aversion.

Proponents argue that, once these shifts are accepted, the St Petersburg Paradox dissolves: the apparent conflict arose from an inadequate assumption about the value of money, rather than from a flaw in using expectations per se. Subsequent sections discuss both the generalization of this idea and criticisms, including claims that concavity alone cannot address all St Petersburg‑type constructions.

9. Concavity, Risk Aversion, and Diminishing Marginal Utility

The St Petersburg Paradox has been a central example for articulating how concave utility functions, risk aversion, and diminishing marginal utility interact in decision theory.

9.1 Concavity and Diminishing Marginal Utility

A utility function U(w) is concave if, for any wealth levels w₁, w₂ and any λ ∈ [0, 1]:

[ U(\lambda w_1 + (1-\lambda) w_2) \ge \lambda U(w_1) + (1-\lambda) U(w_2). ]

Concavity mathematically encodes diminishing marginal utility of wealth: each additional unit of wealth increases utility by less than the previous unit. Logarithmic utility, as proposed by Daniel Bernoulli, is a canonical example.

In economic terms:

This pattern helps explain why extremely large but unlikely gains may not dramatically increase expected utility.

9.2 Risk Aversion and Lottery Evaluation

In modern expected utility theory, risk aversion is characterized by concave U. A risk‑averse agent prefers the certainty equivalent of a lottery—some sure amount—over the lottery itself when they have the same expected monetary value.

In the St Petersburg context, concavity implies that:

• The utility gain from very large payoffs (e.g., 2¹⁰, 2²⁰, …) grows slowly.
• The probability of those payoffs shrinks rapidly.
• The product of probability and utility increment can sum to a finite number, even when the monetary expectation is infinite.

Thus, the paradox illustrates how infinite EMV need not translate into infinite expected utility for risk‑averse agents.

9.3 Interpretive Roles

Different authors assign different explanatory roles to these concepts:

• Some treat diminishing marginal utility as a psychological or economic regularity about human preferences.
• Others view concavity primarily as a formal property within an axiomatic theory of rational choice.
• Risk aversion may be seen as either a descriptive tendency or, more controversially, a normative attitude.

In all cases, the St Petersburg game serves as a vivid example where concavity and risk aversion jointly constrain the impact of rare, extreme outcomes, thereby reshaping how expectations are evaluated without denying the underlying probability calculus.

10. Objections and Strengthened St Petersburg Games

While Bernoulli’s utility-based approach resolves the original lottery under a logarithmic (or generally concave) utility, later authors have argued that this solution is incomplete. They propose strengthened St Petersburg games designed to produce paradoxes even under concave, unbounded utility functions.

10.1 Menger’s Critique

In 1934, Karl Menger argued that Bernoulli’s resolution fails in general because one can construct variants of the game where the expected utility remains infinite despite concavity. Menger’s strategy was to choose payoffs that grow faster than Bernoulli’s logarithmic utility can “tame.”

For example, with a utility function U(w) that is unbounded above (even if concave), one can define payoffs aₙ such that:

[ \sum_{n=1}^{\infty} P(n) \cdot U(w_0 - c + a_n) = \infty. ]

This leads to strengthened St Petersburg games, where:

• Probabilities often remain geometric (e.g., (1/2)ⁿ),
• but payoffs increase more rapidly than in the original game.

10.2 Structure of Strengthened Games

In strengthened versions, the mapping from toss number n to payoff is adjusted so that, for a given concave U:

[ U(a_n) \approx k \cdot 2^n \quad \text{or faster}, ]

for some k > 0. Then:

[ \mathbb{E}[U] \approx \sum_{n=1}^\infty (1/2)^n \cdot k 2^n = \infty. ]

Thus, even when agents evaluate outcomes with concave utility, the expected utility of such a game diverges, recreating the paradoxic tension between infinite expectation and intuitively finite willingness to pay.

10.3 Responses to Strengthened Paradoxes

Reactions to Menger-style constructions vary:

Strengthened St Petersburg games thus serve to test whether concavity alone suffices to address all paradoxes of this kind, or whether additional constraints—on preferences, utilities, or gambles—are needed.

11. Bounded Utility and Alternative Decision Rules

In response to strengthened St Petersburg‑type paradoxes, some theorists have proposed that rational preferences must be represented by bounded utility functions, while others advocate alternative decision rules for cases involving infinite or undefined expectations.

11.1 Bounded Utility Hypothesis

The bounded utility hypothesis holds that there exists some finite upper bound Umax such that, for all wealth levels w, U(w) ≤ Umax. Under this assumption, no gamble can have infinite expected utility, because:

[ \mathbb{E}[U] \le U_{\text{max}} \sum_{n} P(n) = U_{\text{max}}, ]

for any probability distribution {P(n)}. Proponents argue that this blocks both the original and strengthened paradoxes.

Critics raise several concerns:

• Boundedness may seem psychologically or behaviorally implausible for very large stakes.
• It constrains the space of admissible preference orderings in ways some find arbitrary.
• It may conflict with other axiomatic desiderata in decision theory.

11.2 Modified or Supplementary Decision Rules

An alternative strategy is to adjust the norms of rational choice without bounding utility. Proposals include:

• Lexicographic or multi‑stage criteria: For instance, first avoid sure losses, then maximize expected utility, and only then consider extremely low‑probability, high‑payoff tails if they meet further criteria.
• Truncation or tail‑insensitivity rules: Ignore or heavily discount outcomes beyond a certain low probability threshold when they render expectations infinite or unstable.
• Dominance‑based approaches: Prioritize options that dominate others across a wide range of realistic states, even when expectations are infinite.

These approaches often preserve standard expected utility maximization for ordinary, bounded cases, while explicitly restricting or supplementing it where infinite expectations arise.

11.3 Comparative Overview

The choice among these strategies involves trade‑offs between mathematical generality, behavioral realism, and normative appeal, and remains a topic of ongoing debate.

12. Behavioral and Psychological Perspectives

Beyond formal utility theory, the St Petersburg Paradox has been interpreted through the lens of behavioral economics and psychology, emphasizing how real people perceive probabilities, outcomes, and risk.

12.1 Probability Weighting and Prospect Theory

Prospect theory, developed by Daniel Kahneman and Amos Tversky, proposes that people transform objective probabilities via a probability weighting function: very small probabilities may be either overweighted or underweighted relative to their actual values.

For St Petersburg‑style games:

• Some models suggest that extremely low probabilities (e.g., of very high payoffs) are effectively ignored or heavily discounted.
• Under such weighting, the “tail” of the distribution contributes little to subjective value, producing a finite decision weight despite infinite EMV.

Within cumulative prospect theory, the valuation of the lottery depends on both the value function (often concave for gains) and the weighting function, which together yield a finite overall evaluation compatible with limited willingness to pay.

12.2 Heuristics and Cognitive Limitations

Behavioral researchers emphasize that individuals often use heuristics rather than full expected value calculations:

• Scope neglect: People may be insensitive to the sheer magnitude of extremely large payoffs.
• Mental models: Intuitive reasoning about coin tosses may focus on short sequences and typical outcomes, downplaying remote possibilities.
• Complexity aversion: The infinite structure of the game may be mentally simplified to a small number of salient outcomes.

On this view, the paradox illustrates the gap between mathematically idealized reasoning and how human cognition actually operates under uncertainty.

12.3 Attitudes Toward Ambiguity and “Too Good to Be True” Offers

Some psychologists note that a St Petersburg–type offer may trigger distrust or ambiguity aversion:

• Individuals may doubt the feasibility of arbitrarily large payments (even if told otherwise).
• They may suspect hidden conditions or counterparty default, effectively truncating the lottery in their minds.

These attitudes lead to behavior similar to that predicted by finite resource constraints without explicitly modeling those constraints in formal terms.

Overall, behavioral and psychological perspectives tend to treat the paradox less as a challenge to probability theory itself and more as a revealing case of how human agents process extreme risk, rare events, and abstract mathematical structures when making real decisions.

13. Applications in Economics and Risk Analysis

Although the St Petersburg lottery is highly idealized, the paradox has influenced concrete applications in economics and risk analysis, particularly in contexts involving rare but extreme outcomes.

13.1 Insurance and Risk Sharing

In insurance theory, the paradox historically reinforced the idea that individuals may rationally pay a premium exceeding expected loss due to risk aversion and diminishing marginal utility. The St Petersburg case dramatizes how:

• Very unlikely but very large losses or gains can distort EMV.
• Concave utility provides a framework in which taking insurance (or diversifying risk) is rational even when the actuarial expectation appears unfavorable.

This insight has informed models of optimal insurance coverage, coinsurance, and reinsurance, where agents balance small certain costs against low‑probability, high‑impact events.

13.2 Investment and Portfolio Choice

In finance, the paradox has been cited in discussions of:

• Gambles with fat‑tailed returns (e.g., certain speculative strategies).
• Kelly‑type criteria and log‑utility models for optimal betting or leveraging.

Some analysts use St Petersburg‑like examples to illustrate how focusing solely on expected returns can be misleading when outcomes are highly skewed and dominated by rare, massive payoffs. Expected utility and measures of risk (variance, Value at Risk, etc.) become central to evaluating such strategies.

13.3 Catastrophic Risks and Risk Regulation

In risk analysis and policy, analogues of the paradox arise in discussions of:

• Catastrophic environmental or technological risks with very low probabilities but enormous potential damages.
• Debates over how much society should invest in prevention or mitigation.

Here, St Petersburg‑style reasoning highlights tensions between:

The paradox thereby serves as a thought experiment for testing risk‑management principles in domains where tail events play a significant role.

13.4 Pedagogical Use in Applied Fields

In teaching economics, finance, and risk engineering, the St Petersburg Paradox is often used to:

• Motivate the move from EMV to expected utility.
• Illustrate risk attitudes and the limitations of linear valuation.
• Prompt discussion about model validity when real‑world constraints (finite wealth, legal limits, institutional credibility) are present.

Its primary practical impact is thus indirect: by shaping how theorists and practitioners think about the valuation of uncertain prospects, especially those involving extreme but unlikely outcomes.

14. Impact on Modern Decision Theory

The St Petersburg Paradox has had a lasting influence on the development of modern decision theory, helping to shape both its axiomatic foundations and its treatment of risk.

14.1 Stimulus for Expected Utility Theory

Bernoulli’s move from expected money to expected utility prefigured later formalizations by von Neumann and Morgenstern and Savage. The paradox is frequently cited as a motivating example for:

• Representing preferences with a utility function rather than raw payoffs.
• Allowing that utility may be concave, capturing risk aversion.
• Interpreting rational choice as maximizing expected utility (EU).

Although the modern axiomatizations do not rely on the St Petersburg game specifically, they inherit the insight that EMV is insufficient as a general decision rule.

14.2 Infinite Utilities and Axiomatic Constraints

The paradox and its strengthened variants have influenced debates about the admissibility of unbounded utility functions within expected utility theory. Some theorists argue that:

• Allowing unbounded utility leads to paradoxes of infinite expectation.
• Imposing boundedness or other constraints secures more robust theorems (such as existence of optimal acts in certain frameworks).

Others contend that such restrictions may be too strong or ad hoc. The St Petersburg family of cases thus serves as a test of how far axiomatic systems can be extended while retaining intuitive plausibility.

14.3 Alternative Decision Frameworks

The paradox has also informed the development of:

• Non‑expected utility theories (e.g., rank‑dependent utility, cumulative prospect theory) that adjust how probabilities and outcomes are combined.
• Lexicographic and multi‑criteria choice models that treat some considerations (e.g., avoiding certain ruin) as overriding others.
• Imprecise probability and robust decision‑making approaches that hesitate to assign precise expectations in the presence of extreme tails.

In many of these frameworks, St Petersburg‑type lotteries are used illustratively to show where classical EU reasoning yields implausible prescriptions.

14.4 Pedagogical and Conceptual Role

Within contemporary decision theory, the paradox is commonly regarded as a pedagogical rather than an active research problem. It is used to:

Even so, discussions about infinite expectations, unbounded utilities, and extreme risks continue to draw on the conceptual template set by the St Petersburg case.

15. Comparisons with Related Paradoxes of Infinite Value

The St Petersburg Paradox belongs to a broader family of puzzles involving infinite expectations or infinite values. Comparing it with related paradoxes helps clarify what is distinctive about its structure and what is shared across cases.

15.1 Pascal’s Wager

Pascal’s Wager argues that, given a nonzero probability of infinite reward (eternal salvation), the expected utility of belief in God is infinite, allegedly making belief rational on prudential grounds.

Comparison:

Both raise questions about how agents should respond to infinite expectations and whether standard decision rules extend straightforwardly to such cases.

15.2 Pasadena and Altadena Games

In contemporary philosophy of probability, Pasadena‑type games (introduced by Nover and Hájek) involve series whose expected values are conditionally convergent or undefined, rather than straightforwardly infinite. They challenge the assumption that every well‑specified lottery has a determinate expectation.

Compared with St Petersburg:

• St Petersburg’s EMV clearly diverges to infinity.
• Pasadena‑type games render EMV or EU undefined or sensitive to summation order.
• Both illustrate limits of simple expectation‑maximization principles.

15.3 Long‑Run versus One‑Shot Paradoxes

Some paradoxes of infinite value concern long‑run sequences or repeated play (e.g., strategies in certain infinite‑horizon Markov decision problems), whereas the St Petersburg game is framed as a one‑shot decision. This difference affects how appeals to law of large numbers or average payoffs over time may or may not be applicable.

15.4 Shared Themes

Across these related paradoxes, common themes include:

• The adequacy of expectation as a sole decision criterion.
• The role of boundedness and regularity conditions on utilities and payoffs.
• The handling of rare extreme outcomes and infinite stakes.

The St Petersburg Paradox provides one of the simplest mathematical settings in which these issues can be explored, making it a natural reference point for analyzing more complex infinite‑value puzzles.

16. Legacy and Historical Significance

The St Petersburg Paradox has left a substantial legacy across several disciplines, shaping both technical developments and conceptual discussions.

16.1 Influence on Economics and Utility Theory

Historically, the paradox is closely associated with the emergence of utility theory in economics. Daniel Bernoulli’s response introduced:

• The distinction between objective payoffs and subjective utility.
• The idea of diminishing marginal utility of wealth.
• The notion that rational choice should maximize expected utility, not expected money.

These ideas were later taken up and formalized by economists and game theorists, becoming foundational in microeconomics and finance.

16.2 Role in Probability and Philosophy

For probability theory and the philosophy of probability, the paradox has served as:

• A classic illustration of how infinite expectations challenge naïve interpretations of expectation as a universal guide to fair pricing.
• A case study in the relationship between mathematical formalism and rational decision norms.
• A stimulus for debates about the interpretation of probability (frequentist, Bayesian, etc.) in contexts involving extreme payoffs.

Philosophers have used it to probe the coherence of various decision‑theoretic frameworks and to compare them with alternative approaches in ethics and epistemology.

16.3 Continuing Pedagogical Importance

In contemporary teaching, the paradox remains a staple example in courses on:

• Decision theory and rational choice
• Microeconomic theory and risk
• Probability and statistics

It is valued for its simplicity and its capacity to highlight issues about risk attitudes, utility curves, and the limitations of EMV with a minimum of technical apparatus.

16.4 Broader Cultural and Intellectual Impact

Beyond technical circles, the St Petersburg Paradox has occasionally been invoked in discussions of:

• The rationality of lottery play and speculative investments.
• Attitudes toward low‑probability, high‑impact events in public policy.
• The tension between formal models and common‑sense judgments.

Overall, its historical significance lies not in any direct practical application of the St Petersburg lottery itself, but in the profound influence it has exerted on how theorists conceptualize value, risk, and rational choice under uncertainty.