Skip to main navigationSkip to main content
The University of Southampton
Centre for Risk Research

Centre's Findings on Improving Probabilistic Forecasting are Published in Leading Journal

Published: 28 September 2015

Centre for Risk Research members Ming-Chien (Vanessa) Sung, David McDonald, and Johnnie Johnson have had an article accepted for publication in one of the world's leading operations research (OR) journals. Their article, titled 'Probabilistic forecasting with discrete choice models: Evaluating predictions with pseudo-coefficients of determination' is currently in press with the European Journal of Operational Research (EJOR).


Their research focuses on probabilistic forecasting, which involves a decision maker assigning probabilities to different possible future states of the world. Organizations are attaching growing importance to effective probabilistic forecasting, which is highlighted by the significant growth in the adoption of prediction markets to tap into the 'wisdom of the crowd'. Companies such as Hewlett Packard, Eli Lilly, General Electric, and Google have employed prediction markets to inform their strategy on the likely success of new products and meeting project deadlines. Probabilistic forecasting is also used to predict the outcomes of 'competitive events', such as elections, horseraces, and other sporting events. In the paper, the authors show, both theoretically and using real-world examples from horserace betting markets, that there are flaws in some widely-used methods for evaluating probabilistic forecasts, and suggest improvements to the methods to avoid these pitfalls.

EJOR "publishes high quality, original papers that contribute to the methodology of operational research and to the practice of decision making", and has an ABS rating of 4*.



Privacy Settings