This morning we heard on the radio that the chance of rain is very low – 6% – but last night the TV weather forecast said it would be 35% and the app shows us 50%. So… how do we dress for the day? How do we dress the kids? Will we take an umbrella?
We all know these small decisions in everyday life, but also bigger ones: Do you have a surgery when one of the doctors presents a high chance of spontaneous recovery while another doctor thinks the surgery is essential? Invest in the oil market when one expert says that it is worthwhile, since the alternative energy has exhausted its way, and the other says the opposite? Should one go to Sinai when security experts claim it is dangerous while tourism experts expect peace and quiet?
Many of the decisions we make in our personal and professional lives are made under conditions of uncertainty and rely on projections from experts in the field. Decision makers in large organizations – board members, government ministers, etc. – rely on expert evaluations. As long as the experts agree, the decision is easy, but in many cases they disagree and their predictions are contradictory.
Dr. Itai Arieli, Dr. Yakov Babichenko and Prof. Rann Smorodinske of the Davidson Faculty of Industrial Engineering and Management have developed a new model for weighing expert forecasts and presenting an operative solution to this problem for some situations. The article was published in the Journal of the American Academy of Sciences -PNAS.
According to Dr. Arieli, “We offer a mathematical model that allows us to objectively examine the quality of decision making and build a prediction based on several expert forecasts. This is based on the assessment of the aggregate forecast quality in the worst situation. For some descriptors (scenarios) we also offer a concrete solution (mathematical formula) to the problem, which provides a very good aggregate forecast. This outlook is similar to that of a person exposed to all the knowledge and data held by all the experts. ”
Dr. Arieli emphasizes that this is a preliminary study, and therefore it is simple descriptions. In another article born from the same study, we deal with multi-expert descriptors, as opposed to the current article that focuses on descriptors where there are only two experts. “