Despite extensive research, computational creativity still lacks a formally grounded, practically relevant, and empirically testable definition that works across domains such as art, scientific research, and games. Existing approaches either offer informal conceptual descriptions without computational formalism, or present formal definitions without scrutinizing their assumptions or validating them in concrete applications. To fill this gap, the paper builds on psychological studies—particularly Boden’s three core criteria of newness, value, and surprise—and proposes a set of postulates that lead to a mathematical definition of creativity. Let the measurable quantities newness $N$, value $V$, and surprise $S$ be defined for a given artifact; then creativity $C$ can be expressed as
$$ C = \alpha N + \beta V + \gamma S, $$
where $\alpha, \beta, \gamma$ are domain‑specific weights. The formula is evaluated in two settings: assessing the creativity of individual chess moves and measuring the creativity of paintings by artists such as Picasso and Renoir. Results show that the unified metric $C$ captures creative patterns in both chess and visual art, supporting the view that a combination of newness, value, and surprise can characterize creativity across domains.
Review: The study demonstrates a principled way to operationalize psychological creativity criteria and validates them empirically, offering a versatile framework for cross‑domain creativity assessment.