The Importance of Catastrophe Theory and Nonlinear Modeling Applied to the Collapse of Afghanistan in the Late Summer of 2021 |Biomedgrid

Was the fall of Afghanistan predictable? What theory might have been overlooked in interpreting relevant data? What sorts of modeling errors may have occurred? Here we looked into such questions and present evidence that the fall of Afghanistan could have been better predicted with catastrophe theory and nonlinear models, rather than by using only linear modeling and other theories. In particular, using data, from April to August of 2021, from the Long War Journal's website on the number of provinces in Afghanistan controlled by the Taliban or the government, it appears that the fall of the government or the rise of the Taliban were both better predicted by quadratic models than either linear or exponential models. While none of the models would have predicted the fall of the government in mid-August 2021, linear models predicted substantially later dates than did the nonlinear models. Even as early as May or June of 2021, nonlinear models were predicting an earlier collapse of the government than were linear models, ruling out “hindsight” as the best explanation of our findings. For didactic purposes, our results serve as learning points for students to avoid exclusive reliance on both linear modeling and status quo theories that would suggest that sudden change is either unlikely or impossible, despite numerous historical counter-examples.