Big data are believed to become a driver of transformation in the future world. Data analysis can provide “evidences” or “trends” for supporting decisions and plan interventions at personal (e.g. health care) and political levels (e.g. consensus). Science behind collection, organization and representation of numerical data is a prerequisite to properly interpret the results of their elaboration. This is true also when dealing with homogeneous data, that is from a simple set of numbers resulting from different measures of the same variable. Often, results from data analysis are interpreted without checking nature (e.g. randomness) of the involved process, errors are not addressed, samples are not tested for their representatives of the populations, plots are shown with misleading scales for interpretations. In a future when computational capacity will increase and interpreting methodologies will be widely used, any “simple and robust” (e.g. statistical) analysis of some data and results could stimulate warnings in choosing options.