Recently, a report sparked a media feed (not quite large enough to qualify as a frenzy). We heard that the murder rate in many cities had soared. As is often the case this was a gross misreading of the data. Two of the largest mistakes were ignoring the notion of statistical fluctuations and that the survey was not designed to show what the media said it showed. You can see all of the details at FiveThirtyEight which does the best statistical analysis in the news business (okay, maybe the only statistical analysis in the news business – but it is still fantastic).