in , placed the guessers in the Eighth Circle of Hell, with their heads twisted, facing their backs, with tears wetting their buttocks, so that they cannot look forward. This is the punishment for claiming to know the future that would be within the reach of only God.
“Prediction is very difficult. Especially if it is about the future”, the phrase is from one of the creators of quantum physics, . Estimating a model based on a set of past data/variables is one thing. It’s like predicting the present. But predicting future results (“out of sample”) is much more complex, requiring assumptions about the role of still unknown factors. Many predictions are very complex; others simply impossible.
Predictions by political scientists and economists have something in common: they are anchored in expectations about expectations. The metaphor of a hypothetical contest whose participants should choose —based on photographs of hundreds of candidates— the most attractive, and be rewarded for correctly predicting the winner is classic; it allows a distinction between the analyst’s preference and that of others, the aggregation of which will determine the final result. Betting on what we want to happen is a permanent source of self-deception. But it is the most common. In the current context of polarization, predictions are confused with partisan and politically tribal fans.
The metaphor captures another distinction: it is not just the pure and simple opinion of the participants, but their perception of the evaluation that the average person will make. The prize is for whoever guesses the winner, and everyone will make the calculation based on the expectations of the group of participants. The analogy here is with the stock market: what an individual thinks about the value of a share does not matter, but what others think about it; or more importantly, what predictions they make about the market’s average valuation of the stock.
The same goes for voting. Both turnout and the vote itself are a function of expectations about the electoral results.
Phenomena of enormous political importance, such as , the fall of and , were not anticipated by any analyst. Many of these rare events (in jargon, the “long tail”) are characterized by processes that accelerate dramatically and erupt suddenly. Timur Kuran identified “informational cascades” that trigger herd effects and that are not observable by analysts because the . Suddenly millions of supporters of a cause appear out of nowhere. Or they refuse to respond to opinion polls, introducing bias.
Making predictions oscillates between hubris and charlatanism, but they are inevitable and should be seen as educated guesses.
In the Eighth Circle, the Diviners have for company Seducers, Flatterers, Simoniacs, Corrupt, Hypocrites, Thieves of the sacred, Bad advisors, Sowers of discord, Alchemists. Many of them can be identified by name: they are real characters from Florentine politics and history. I think about how long the list of characters in our divine everyday comedy would be.
LINK PRESENT: Did you like this text? Subscribers can access seven free accesses from any link per day. Just click the blue F below.
