# Forecasting Beginnings and Endings By:: [[Brian Heath]] 2023-05-02 Mathematical forecasting models struggle with beginnings and endings. Beginnings have no data to rely upon. Endings are often so frantic that no time is given to getting them right. An outside observer could collect and categorize all beginnings and endings to form an aggregate understanding of how things begin and end, but these only represent the distribution of what could be. Reducing the uncertainty to a precise level for beginnings and endings is most irrelevant or impossible because of the reasons brought up before - beginnings have no data and endings have no one to collect the data. The nature of forecasting is that the best predictor of the future is the past. This is why middles are where forecasters live. Middles have pasts and futures. Beginnings and endings are where gamblers and risk-takers live. If you try to apply middle ideas to the beginning and end, you'll quickly fall into the pit of unclassifiable uncertainty. Rarely are analytical professional gamblers, which probably explains why we have so few techniques for the beginning and end. Alternatively, maybe the chicken came before the egg. There are no analytical techniques for beginning and endings, so analysts always try to live in the middle. The hammer has been found, so where are all those nails? #### Related Items [[Analytics]] [[Forecasting]] [[Systems Thinking]] [[Tools]] [[Business]]