Revisiting the Predictability of Social Events

Social predictions do not passively describe the future; they actively shape it. They inform actions and change individual expectations in ways that influence the likelihood of the predicted outcome. Given these dynamics, to what extent can social events be predicted? If predictions shape the data we see, what does it even mean to make a good “prediction”?

In this talk, I'll provide an overview of the history behind these classical questions and share new insights derived from revisiting them using modern tools from performative prediction [PZMH20’] and (online) multicalibration / outcome indistinguishability [HKRR18’, DKRRY21’].