Algebraic geometers often work with surfaces that have been "simplified" to their most essential form, called minimal surfaces. A minimal surface is one that contains no special curves that can be collapsed to a point to produce a simpler, equivalent surface. These collapsible curves are called (-1)-curves, named after a numerical property they possess. Determining whether a given surface is already minimal, or finding where these (-1)-curves hide, is an important but often technically demanding task.
The paper introduces a numerical criterion, meaning a test based on computable numbers associated with a surface, that can tell you whether a surface of general type is minimal. "General type" refers to the most complex and varied class of algebraic surfaces, making this a broadly applicable result. The criterion essentially gives geometers a practical shortcut: instead of hunting geometrically for (-1)-curves, you can check certain numerical conditions to confirm they do not exist.
The authors then apply this criterion to a specific and well-studied family called product-quotient surfaces. These are constructed by taking the product of two curves and dividing out by the action of a finite group of symmetries. They have a natural geometric structure with two fibrations, meaning two different ways of slicing the surface into families of curves. The refined criterion tells you not just whether (-1)-curves exist, but also that any such curves must lie entirely within a fiber of one of these two fibrations. This structural insight makes the problem of minimizing product-quotient surfaces much more systematic, since you know exactly where to look for the curves that need to be collapsed.