Study: “The average prognosticator is no better than a coin flip in terms of the accuracy of his or her predictions.”

May 3rd, 2011

Individuals are bad at making predictions, especially about the future, said Yogi Berra. Now, a study from Hamilton College, titled “Are Talking Heads Blowing Hot Air? An Analysis of the Accuracy of Forecasts in the Political Media,” looking at the accuracy of predictions made by so-called “pundits,” proves Yogi’s aphorism.

From the Executive Summary:

We sampled the predictions of 26 individuals who wrote columns in major newspapers and/or appeared on the three major Sunday television news shows (Face the Nation, Meet the Press, and This Week) over a 16 month period from September 2007 to December 2008. Collectively, we called these pundits and politicians “prognosticators.” We evaluated each of the 472 predictions we recorded, testing its accuracy.

What were the results?

The most accurate prognosticator was Paul Krugman of The New York Times and Princeton University, followed by Maureen Dowd, another columnist for the Times, and former Pennsylvania governor Ed Rendell.  The worst prognosticator was Cal Thomas, a syndicated columnist.  South Carolina Senator Lindsay Graham was next worst, and Michigan Senator Carl Levin was third worst.

We discovered that a few factors impacted a prediction’s accuracy.  The first is whether or not the prediction is a conditional; conditional predictions were more likely to not come true.  The second was partisanship; liberals were more likely than conservatives to predict correctly.  The final significant factor in a prediction’s outcome was having a law degree; lawyers predicted incorrectly more often. Partisanship had an impact on predictions even when removing political predictions about the Presidential, Vice Presidential, House, and Senate elections.

What were the implications?

We have discovered a number of implications from our regressions and analysis of the data.  First, we have discovered that 9 of the analyzed prognosticators are significantly better than a coin flip.   Two are significantly worse, and the remainder are not significantly different, in a statistical sense, than a coin flip.   This means that the average prognosticator is no better than a coin flip  in terms of the accuracy of his or her predictions.

Our 60-70% FantasySCOTUS prediction rate looks much more attractive when compared with the accuracy of so-called “pundits.” Simply put,individuals cannot predict. This is why I don’t watch talking heads. They are usually wrong.