| Abstract: | In the past two decades, two theoretical models of GRBs and their afterglows, the fireball (FB) model and the cannonball (CB) model, have been used extensively to interpret the mounting observational data on GRBs and their afterglows. But, despite their similar names, the two models were, and still are, quite different in their basic assumptions and predictions. Their claimed ability to describe well GRBs and their afterglows despite their complexity and diversity, may reflect the fact that their posteriori data fitting for each GRB involves multiple choices and many free adjustable parameters. Scientific theories, however, must be falsifiable. Hence, only confrontations between accurate observational data and key predictions of the GRB models, which do not depend on free adjustable parameters, can serve as critical tests of their physical validity. In my talk I will shortly review the FB and CB models and confront their key falsifiable predictions with accurate observational data rather than prejudices. These data include the observed supernova-GRB-XRF connection, the polarization of the prompt gamma rays, correlations between various prompt emission properties and the prompt GRB energy relative to the afterglow energy. Additional critical tests provided by the GRB afterglows, include their canonical behavior, a jet break, closure relations between their temporal and spectral behavior, before and after the break, and correlations between afterglow properties. |