Bloomberg’s World Cup Prediction Gaffe

Bloomberg recently ran this headline, “Germany Will Win the World Cup, UBS says after 10,000 Simulations.”

Spoiler alert: Germany went home after failing to get out of the group stage, and Bloomberg got some public shaming for misrepresenting the math with that headline.

This hits extra close to home because A. I didn’t have Germany winning on my bracket (eat that UBS calculators), and B. when those of us in the financial industry run several thousand simulations to test whether or not someone / something is able to meet some future obligation, we’re using the same mathematical modeling technique.

In this case, the math doesn’t deserve a bad rap. As the poet / philosopher Rakim once said, “don’t sweat the technique.”

Here’s the rub: UBS didn’t actually say Germany was going to win, they only gave them 24% odds of winning. That didn’t stop Bloomberg’s headline writers going to work for clicks however. Credit to poker player turned author Annie Duke and quantitative finance guru Cliff Asness for both standing up and making this point publicly in UBS’ defense.

Duke and Asness both said to remember that 24% odds of winning = 76% odds of losing. Germany’s early exit suddenly sounds a lot more probable, doesn’t it? UBS did clearly reference this reality in their report.

We always should invert our percentages to slow ourselves down and check our heads.

Probably does not mean definitely, with certainty, or that the predictor has magical clairvoyance. It means there is a chance of yes AND a chance of no.

Let Bloomberg’s errant arrogance serve as a reminder to us finance people that while we have the luxury of elegant methods and formulas, we don’t have the luxury of certainty. We can never let a headline suggest otherwise. We have to be better than that.

Leave a Reply

Your email address will not be published.