Scott Armstrong presented a commentary on the findings of the M4-Competition at the M4 Conference in New York City on December 10, 2018. The paper, by Scott Armstrong and Kesten Green, concludes that data models should not be used for forecasting, describing six reasons why. The reasons are not new—for example Einhorn pointed out the lack of theoretical support in "Alchemy in the behavioral sciences" in 1972—but they remain true.

The conference paper slides are available from ResearchGate, here.

Accurate predictions and the correct assessment of uncertainty are indispensable for all types of future oriented decisions: from determining appropriate inventory levels and predicting the amount of sales of companies to forecasting revenues and costs and the formulation of a long-term strategy. The purpose of the Makridakis, or M, Competitions has been to provide empirical evidence to guide organizations on how to improve the accuracy of their forecasts and how to assess future uncertainty as realistically as possible.

Scott Armstrong presented a paper with Kesten Green at the International Symposium on Forecasting in Boulder, CO, on 19 June titled "Do Forecasters of Dangerous Manmade Global Warming Follow the Science?". A pdf copy of the slides is available from ResearchGate, here.

With the proliferation of taxes, subsidies, and regulations being justified by appealing to predictions of dangerous manmade global warming, Scott Armstrong and Kesten Green believe that climate change is currently the most important forecasting problem in the world. With that in mind, Scott will present their latest research on the topic at the International Symposium on Forecasting in Boulder, Colorado, on June 19.

Following the New York Times injunction to "follow the science", the talk will be titled "Do forecasters of manmade global warming follow the science?". For some background on the topic, see Scott's essay describing their research has been posted under the title of "Is the Earth becoming dangerously warmer?" on WUWT.

The authors of "Forecasting methods and principles: Evidence-based checklists"—Scott Armstrong and Kesten Green—are pleased to announce that they have been given permission to post the published version of this paper on The paper addresses the problem of "How to help practitioners, academics, and decision makers use experimental research findings to substantially reduce forecast errors for all types of forecasting problems." A Chinese translation is included in the paper.