Distinguished Professor David Lazer and his colleagues analyzed global-scale databases of news events and found them wanting. Their recommendations for improvements would enable researchers to build models anticipating everything from the escalation of conflicts to the progression of epidemics.