»Unique competencies in all relevant methods for sizing of requirements (function and quality)«
The reliability of expert judgment depends very much on knowledge, capabilities, and experience and decreases with increasing project size, agility, complexity, and automation degree.
Expert judgments are person-dependent and therefore not objective, hardly reasonable, unprecise, usually not reproducible, and thus not reliable.
Studies prove, that 80 % of all expert judgments have a deviation by a factor 2 - 9 between estimates and actuals.
Facts about success of software and IT projects: Numbers remain »constant« for 20 years
Heuristic (top-down, bottom-up) and parametric sizing methods are used to objectively determine the size of a project. Parametric methods meant functional sizing (FS) methods, "hard-facts" or, according to the ISO standard, FSM (Functional Size Measurement).
FS methods have a different number of specific metrics. With these metrics, functional & qualitative requirements are transferred to unitless numbers (size), usually called points. All FS methods return as result just a size, but not effort, duration, or cost.
Typical examples of FSM methods are, e.g. Use Case Points, Story Points, Feature Points, Function Points, COSMIC Points, SNAP Points, MKII Points, etc. Further, there are specific FS methods, e.g. RICEW Points (for sizing of ERP applications), Mainframe Points, BPM Points, BI Points, Service Points, etc.