Share this post
FaceBook  Twitter  

In the early 1980s, two of the authors of Principles of Business Forecasting, Robert Fildes and Keith Ord, were leaders in the movement to use experiments to determine which principles led to more accurate forecasts. The experiments show that forecasting methods today are capable of providing much more accurate forecasts that they were 50 years ago. This movement was a golden age for progress in forecasting. However, this knowledge has been not been widely adopted in business and government.

In recent years, Fildes and Ord have been concerned with ways of implementing new approaches. To my knowledge, Principles of Business Forecasting is the first textbook to include some of the new methods. It is aimed both at beginning forecasters and practitioners , so it also covers the basics. Despite the improvements in methodology, forecasting accuracy in business and government has been shown to be getting poorer. Much of the forecasting has been done by people who have no awareness of scientific forecasting methods. Principles of Business Forecasting is directed at that market.

Research by Kesten Green and I has found that forecasters have turned to "analytics" using "big data." That approach ignores all cumulative knowledge about forecasting methods. It also violates the Golden Rule of Forecasting and Occam's razor. In his 1972 paper "Alchemy in the Behavioral Sciences," Hillel Einhorn warned of the danger of relying on the computer. Hopefully, Ord, Fildes and Kourentzes's book will bring forecasting back to the basics and convince forecasters to use more accurate methods and principles. There is profit to be made.

J. Scott Armstrong, The Wharton School, University of Pennsylvania.