Decoding the MPN principle: Clarifying the logical structure of most probable number estimation

The Most Probable Number (MPN) method is widely used for estimating viable microbial concentrations from dilution-based presence-absence data obtained from inoculated culture units. Although commonly presented as a statistical procedure derived from the Poisson distribution, its logical structure includes an initial deterministic interpretive component. The observed pattern of positive and negative growth outcomes at defined inoculum volumes constrains the range of concentrations compatible with the data. Only after this interval has been established does the Poisson model assign relative likelihoods within it and yield a most probable estimate. By distinguishing between deterministic interval inference and subsequent statistical refinement, this article clarifies the inferential structure of the MPN method. It shows that statistical modelling operates within logical constraints already defined by the sampling design and outcomes, while remaining fully probabilistic in its estimation. This reinterpretation provides a more transparent conceptual framework for understanding both the strengths and the limitations of dilution-based enumeration.