I’m continuing to work my way through the papers in the International Journal of Forecasting special issue on the M4 Competition. One of the early articles is “Forecasting in social settings: The state of the art” by Makridakis, Hyndman, and Petropoulos. The paper gives a nice review of the science of forecasting over recent years, including comparisons among the M4, M3, and M1 competitions. Although the focus is mainly on time series forecasting and automated methods there is some discussion of judgmental forecasting and judgmental adjustment of statistical forecasts. This overlaps a little with forecasting tournaments such as those conducted by Good Judgment Open and others. In the latter the forecasting questions regard geopolitical events which may be categorical instead of quantifiable. For example, forecasting the outcome of a foreign election typically involves selecting one of a set of candidates. That is, there may be many social settings where individual judgments are all that one has to make a forecast. Understanding which technique to apply to which question (statistical, judgmental, judgmental adjustments to statistical, statistical adjustments to judgmental, judgmental/statistical fusion) across a broad portfolio of forecasting questions remains a challenging problem.