The intelligence community (IC) is asked to predict outcomes that may often be inherently unpredictable--and is blamed for the inevitable forecasting failures, be they false positives or false negatives. To move beyond blame games of accountability ping-pong that incentivize bureaucratic symbolism over substantive reform, it is necessary to reach bipartisan agreements on performance indicators that are transparent enough to reassure clashing elites (to whom the IC must answer) that estimates have not been politicized. Establishing such transideological credibility requires (a) developing accuracy metrics for decoupling probability and value judgments; (b) using the resulting metrics as criterion variables in validity tests of the IC's selection, training, and incentive systems; and (c) institutionalizing adversarial collaborations that conduct level-playing-field tests of clashing perspectives. (Contains 3 figures and 1 footnote.)