Close-call counterfactuals and belief-system defenses: I was not almost wrong but I was almost right.

Drawing on samples of professional observers of world politics, this article explores the interrelations among cognitive style, theoretical outlook, and reactions to close-call counterfactuals. Study 1 demon-strated that experts (especially high scorers on a composite measure of need for closure and simplic-ity) rejected close-call counterfactuals that redirected history when these counterfactuals undermined a preferred framework for understanding the past (the I-was-not-almost-wrong defense). Study 2 demonstrated that experts (especially high scorers on need for closure and simplicity) embraced close-call counterfactuals that redirected history when these counterfactuals protected conditional forecasts from refutation (the predicted outcome nearly occurred—so I was almost right). The article concludes by considering the radically different normative value spins that can be placed on willingness to entertain close-call counterfactuals, Many scholars have commented on a curious asymmetry: People often display far greater confidence in their explanations of the past than they do in their predictions of the future (cf. Fischhoff, 1975; Tetlock Belkin, 1996). One influential expla-nation holds that people engage in retrospective data fitting when