Wednesday, January 15, 2020

Prophetic Probabilities

In Milton Friedman's 1953 paper "The Methodology of Positive Economics," he proposed that a model shouldn't be evaluated on the reasonableness of its assumptions, but on its predictive power. If we don't know why a model works, but it works, that's better than a model that makes all the sense in the world and always ends up being wrong.

Lots of people have a problem with this. I don't think I'm one of them. Friedman is allowing for the discovery of a causal relationship's existence before we understand the causation, which is the first step to exploring causation, right? Before we knew what caused the seasons we had a lot of mythological answers, but they were good because they accurately predicted the seasons, which was what people needed.

I thought of this Friedman argument when I saw this attempt to forecast the future leaders of the Church of Jesus Christ of Latter-day Saints. Now, the blogger notes the weak predictive power of his model, both in that post and in an earlier post here. But one thing I'd like to do if I had time* would be to create graphs for the history of the Quorum of the Twelve Apostles, showing the discontinuities that take place each time there is a change in quorum membership. Right now, the graph only shows the certainty that President Nelson is the current president, but it would be interesting to see who was likely to be president compared to who was, and to see how unlikely some presidencies were before they happened.

In 1997, Time magazine had a cover story about the church and its administration. I remember one line that said something was likely to continue "if the church's hard-line Number-3 man, Boyd Packer, someday becomes president, as seems likely." Of course, Elder Packer did not end up being President of the Church. But the historical predictions would be great for some perspective before we start going around thinking we know who the next four presidents will be.

* = I read today an apocryphal statement attributed to Warren Buffett, that goes something like, "Make a list of your top 25 goals, and put the top five on your to-do list, and the other 20 on your avoid-at-all-costs list." This project would definitely be one of those 20.

No comments: