Unpredictable Analytics and the World Series

Michael Vizard

There's no joy at Information Builders Inc. as the mighty predictive analytics application has struck out.

Prior to the start of the baseball playoffs, one of IBI's leading lights in the area of predictive analytics used its software to come up with a prediction that the Los Angeles Dodgers would most likely beat the Los Angeles Angels in the World Series.

As we all know, the New York Yankees, who IBI said were the third most likely to win the Series, are playing the Philadelphia Phillies starting tomorrow in this year's rendition of the Fall Classic. Now in this case, the faulty prediction may have more to do with the person doing the analyzing than the actual software. Turns out that Kevin Quinn, vice president of product marketing at IBI, is an ardent New York Mets fan. It's hard to say how much influence this might have played on the model that Quinn set up, given the heated rivalry between the Mets and the Phillies and the general envy that Mets fans feel about anything to do with the Yankees.

That all said, Quinn notes that most companies would not use predictive analytics on something with so many random factors as baseball. The model Quinn used, for instance, said that the Dodgers essentially had a one-in-three chance of winning the World Series. In a business application, probability ratings would have to be above 60 percent before anyone would consider a prediction to be anywhere close to actionable.

But even then, no matter how good the software is, there is no such thing as a sure thing.

Add Comment      Leave a comment on this blog post

Post a comment





(Maximum characters: 1200). You have 1200 characters left.




Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.