Combining the Opinions of Experts Who Partition Events Differently

This paper focuses on updating a client's beliefs about an event based on information about the different probabilities that various experts assess for that event. A substantial literature solves this problem when all experts assess their probabilities over the same partitioning of the possible outcomes of an event. But different experts often think about the same problem in quite different ways. This can lead to differences in how experts prefer to partition the possible outcomes of an event. Forcing the experts to use a common partition could lead to less informative probability assessments. Thus, this paper presents a new approach for combining probability assessments from different experts, which allows experts to assess their probability assessments across different partitionings.