data-disasters

How metric definitions, ambiguous calculations, sample sizes, and domain knowledge make calculating a humble average a formidable and thought-deserving task

A personal encounter with 'intelligent' data products gone wrong