Boxed In

George Box was a statistician and considered one of the greatest minds of the 20th century.  He was born in England, served in WWII, and eventually moved to America as an academic with such prestigious schools as the University of North Carolina, Princeton, and University of Wisconsin-Madison.  Mr. Box worked in many areas of statistics, including time-series analysis and Bayesian inference which are used still today to model the world of economics and finance.

Big things were happening nearly 50 years ago for models.  For one thing, George published a book with his co-author George C. Tiao titled Bayesian Inference in Statistical Analysis in 1973. In short, it is an approach to modeling which updates probabilities of a hypothesis as new information becomes available.  This happened to also be the year the Standard Model of particle physics was completed; this complex model is used to predict properties of elementary particles and forces with tremendous accuracy (i.e., correct to 14 decimal places). 

But this is 2022 and things might be changing.  Scientists from the Fermi National Accelerator Laboratory announced that the model did not accurately predict the mass of an elementary particle called a w boson.  For those interested, you can see an article from science.org here.  This could be a clue which will ultimately lead scientists closer to understanding reality. 

Economists are tweaking their models as well.  This century has perplexed many as consequences of some actions haven’t exactly played out like models suggested.  Take inflation for instance.  America’s economy has been through two major crises which resulted in the Federal Reserve printing gobs of money, yet the economic inflation which was thought to result didn’t exactly materialize.  Consider the inflation rate after the Great Financial Crisis.  Following the surge of new dollars appearing from thin air, inflation’s trend was below the previous one.  We are now experiencing inflation that hasn’t been seen in decades, yet the debate on its cause(s) remains open.  Some suggest the supply chain challenges are a bigger contributor than the covid-era money printing. 

George Box was pragmatic.  He understood that the map is not the terrain.  In other words, models are incapable of understanding the reality they seek to demonstrate fully.  Atlas likes to remain conscious of what is likely his most referenced insight: “all models are wrong, but some are useful.”