Boy or Girl Paradox Calculator

The boy or girl paradox demonstrates that a probability question cannot be answered correctly until the information-gathering process is clear. In its classic form, a family has two children and at least one is known to be a boy. Under a simplified model of independent, equally likely births, the possible ordered outcomes are boy-boy, boy-girl, girl-boy, and girl-girl. The condition removes girl-girl, leaving three equally likely possibilities. Only boy-boy has two boys, so the conditional probability is one third. That result changes when the statement identifies a particular child. If the older child is known to be a boy, only boy-boy and boy-girl remain, and the probability that both are boys is one half. The difference is not caused by biology changing. It arises because “at least one” filters the sample space differently from “this specified child.” Conditional probability always depends on both the event of interest and the exact evidence supplied. The calculator generalizes the reasoning to families of two through ten children. There are 2^n ordered gender sequences for n children. If the evidence is that at least one child is a boy, the all-girl sequence is excluded, leaving 2^n - 1 possibilities. Exactly one sequence consists entirely of boys, so the requested probability is 1/(2^n - 1). If a particular child's gender is fixed as boy, the remaining n - 1 children have 2^(n-1) possible sequences, and only one makes every child a boy. This is an idealized probability model. It treats boy and girl as equally likely binary outcomes, assumes births are independent, and does not model biological sex ratios, twins, nonbinary gender identity, or family-planning behavior. Those assumptions make the sample-space lesson easy to see, but they should not be mistaken for a complete demographic model. The historical paradox uses gendered wording; mathematically it is simply a conditional probability exercise with two outcome categories. Real versions of the puzzle can become more complicated when information is volunteered or selected. Hearing “at least one child is a boy” from a randomly chosen parent may have a different likelihood than asking whether the family contains a boy, because parents may choose which child to mention. Variants involving a birth day, such as “at least one boy born on Tuesday,” similarly depend on how the fact was obtained. Use this calculator to compare two precisely defined conditions and to see why seemingly minor wording changes can alter a Bayesian sample space.