The responses to a questionnaire which was administered to forecasters actively involved in probability forecasting are summarized in Murphy and Winkler (1971). These responses and subsequent discussions with forecasters reveal the presence of a number of “problems” concerning probability forecasting. In this paper, we identify several of the more important problems, describe their nature, indicate some approaches and results which clarify certain aspects of the problems, and make some recommendations related to research studies and operational practices in probability forecasting. In particular, we discuss the formulation of judgments and the assessment process, the interpretation of probability forecasts, the occurrence of “hedging” by forecasters, and the evaluation of probability forecasts and forecasters.
1 Supported in part by the National Science Foundation (Atmospheric Sciences Section) under Grant GA-1707.
2 Contribution No. 170 from the Department of Meteorology and Oceanography, University of Michigan.
3 Presently on leave and visiting the Graduate School of Business Administration, University of Washington, Seattle.