The concept of truth has long been a subject of philosophical debate. This paper proposes a probabilistic approach to defining truth by extending the concept of truth value from a binary to a continuous variable. Truth value is defined as the degree of correspondence between a statement and what it refers to. A simple mathematical formulation is offered and compared to a Bayesian approach.

This paper illustrates the use of this definition using some simple examples such as the shape of the Earth. A brief discussion on how this definition can impact different fields is provided, along with potential applications.