When uncertainty is high and the situation is complex and not well understood, it is hard to beat a simple no-change forecast. Applying that principle to problem of long term forecasting of global average temperatures, Scott Armstrong challenged former Vice President Al Gore to bet his dramatic forecasts of dangerous manmade global warming against a no-change benchmark in 2007. Mr Gore did not accept the bet, but what if he had? We now have two years of monthly data since the challenge was issued, and Mr Gore's prediction (assumed to be an increase of 0.03C per year) was less accurate than assuming the temperature would not change from the 2007 average for both years. Mr Gore's forecast was more accurate in only four of the 24 months. To read more, visit The Global Warming Challenge site.
The Program Committee of the International Symposium on Forecasting (ISF) 2010 invites the submission of abstracts related to the theory and practice of forecasting. The deadline for abstract submission is 1 March 2010. The 30th ISF will take place in San Diego, CA, U.S.A. on June 20-23, 2010. For more information visit the symposium site.
The original plan was to make the chapters available for sale. This plan is still underway but is moving slowly. While waiting, Scott Armstrong is posting the nine papers that he was involved in so they are available for teachers and others. This plan will change when the publisher is able to establish a system to sell the chapters.
The full text papers are available at Principles of Forecasting in the left-hand menu of the ForPrin site.