Everyone reasons probabilistically more or less constantly, figures Nickerson (experimental psychology, Tufts U.), and he is interested in how well they do so. He explores its origins, its status, its use in various contexts, what recent research has revealed about it, and other facets. His history, not of gambling but of the serious study of it, begins with the seminal correspondence between Pascal and Fermat in 1654. He assumes no specialized knowledge, but the ability to learn and follow sometimes complicated notation.