The reason why a Trump victory was such a surprise was because of the overconfidence that spectators had in electoral forecasting models and polling data. For example, because of things like 538’s electoral forecasting “the market” was predicting almost a guaranteed Clinton victory. However, Taleb frequently put out there that there were serious flaws with these forecasting methodologies. Despite the fact that Nate Silver’s 538 had slightly more bearish probabilities than things like the New York Times there is no doubt that 538’s Bayesian approach was extremely fragile to new new data i.e. poor out of sample performance. Just to pick on 538 here, as Nate’s work gained the most publicity, the below diagram shows his estimated odds.
Now it is important to stress that Nate did give Trump a better chance of winning than a number of other forecasters (click link), however, his odds neverthless almost entirely inverted the moment it looked like Florida would go to Trump thereby demonstrating the fragility of this methodology in prediction.
Taleb goes on to say that as this estimator is it self stochastic due to feedback effects the ex-ante probabilities converge to 50/50 as you increase the variance of the estimator. Taleb’s then goes onto argue that people should not model the estimated odds like 538 and those that do are “Fooled by Randomness”. In light of this Taleb has published a paper on how one should correctly predict an election and has even uploaded a MOOC. Taleb’s methodology uses Option Pricing theory to determine the price of a binary Win/Lose option using a no arbitrage condition.
Taleb’s final “How to Predict an Election” paper is here:
He has also released a MOOC to outline his paper here:
- A nice summary introducing the debate: