The Signal and the Noise (reviewed last time) is one of many books that cite’s Philip Tetlock’s work on expert predictions, and so I thought there might be value in summarizing that work here. Tetlock is a professor of psychology and political science, and in the 80s, he started collecting predictions from experts across academia and government on a variety of topics. What he found was that the experts were about as good at predicting the future as random chance. Experts were better than undergrads, but still absolutely rubbish, and worse than just assuming the status quo would persist.
That’s the headline result, but there are also some extensions. The most commonly cited of these is the idea of hedgehogs and foxes, from the title of an essay by Isaiah Berlin discussing Tolstoy. He, in turn, took the title from a Greek poet; “The fox knows many things, but the hedgehog knows one big thing.”
In the context of predictions, what Tetlock found was that Hedgehogs (experts who focused on a single big idea, and specialized heavily) were significantly worse at forecasting than Foxes (experts who have a plethora of little ideas, and are often multidisciplinary). Unfortunately, hedgehogs are the ones who go on tv – why ask experts who think there are two sides to an issue when you can have panelists who think everything looks like a nail? Big, bold predictions may be usually wrong, but they’re exciting, and that’s what matters for a public intellectual.
Takeaway: experts don’t have much to tell us in predicting the future (though have significant value in other ways, one hopes), but adaptable experts with breadth have at least a little more to tell us than heavily specialized ones do. In the end, humility about what we can predict and what is predictable may be our only option.