The talk first explains the concepts behind X-ray data quality indicators, which are still not appreciated by many practising crystallographers, are frightening for novice crystallographers, and lead to unfortunate decisions in data collection and processing as well as refinement if not understood correctly. The second part of the talk will report new results that highlight the relation between accuracy and precision, by analysing simulated data that differ in mosaicity, cell parameter variation, wavelength dispersion and beam divergence. These simulated data were used for refinement using a close-to-optimal model as starting point. The results are surprising and lead to a re-evaluation of current understanding of model and data error. Furthermore, simple guidelines for measurements with pink beam will be given.