Methods, systems, and computer program products are provided for decision making under uncertainty. A causality framework is provided to define relationships between an alternative (Ai) of a set of alternatives and at least one performance indicator of a set of performance indicators. A k-concurrent lottery (Li) associated with the alternative (Ai) is conducted, the alternative (Ai) having consequences over k ones of the set of performance indicators. The k-concurrent lottery (Li) is homogenized over the set of performance indicators if the k ones is less than a number of the set of performance indicators. An expected utility of the k-concurrent lottery (Li) for each one of the set of alternatives is computed. One alternative is selected from the set of alternatives, the selection occurring in accordance with a selection criteria based on the expected utility.