In my previous post Sisyphus didn’t need a fitness tracker, I recommended that you only collect, measure and analyze big data if it helps you make a better decision or change your actions.

Unfortunately, it’s difficult to know ahead of time which data will meet that criteria. We often, therefore, collect, measure and analyze any data that might help us achieve our goals. Along the way we sometimes make dubious connections between the results we are seeing and the data we are analyzing, leading us afoul of smart thinking.

The Chicken Man

I am currently reading You Are Now Less Dumb: How to Conquer Mob Mentality, How to Buy Happiness, and All the Other Ways to Outsmart Yourself, the new book by David McRaney. In the chapter about the post hoc fallacy, McRaney recounted the superstitious rituals of Wade Boggs, the third baseman for my boyhood Boston Red Sox, and one of the best baseball players of all time who was inducted into the Baseball Hall of Fame in 2005.

Boggs was known as the Chicken Man, McRaney explained, “because he insisted on eating chicken before every game. He was also obsessed with the number seventeen, and began practice in the batting cage at exactly 5:17, and then ran sprints at exactly 7:17. Once, while in a slump, the announcer forgot to mention Boggs’s number when he called out his name to the crowd. Boggs’s slump ended with that game, and from then on he asked the announcer not to mention his number before play. One biographer wrote that Boggs’s entire life consisted of these routines. He was a clockwork man, a person who ritualized everything in order to keep track of his output. By remaining consistent and mechanical, Boggs saw his performance become measurable, comparable.”

Consistent, mechanical (i.e., repeatable often via automation), measurable, and comparable are often lauded as characteristics of sound metrics. While it’s easy to chuckle at the metrics the Chicken Man used, if you took a brutally honest look at your organization’s metrics, you might find more than a few of them are for the birds.

Statistical Neurosis

Boggs is an extreme example, but he’s not a statistical outlier. “Sports can do that to people,” McRaney explained, “make players and fans into statistical neurotics more compulsive than any Dungeons and Dragons master could hope to be. It is this devotion to a quantified lifestyle that causes so many athletes to adopt magical beliefs. If they look at the numbers and see an improvement, everything that preceded that bump is suspect. Everything that comes before a positive outcome is lumped into the mixture of rituals and behaviors worth repeating.”

The statistical neurosis that sports engender fuels the frenzied world of sports analytics, where professionals and amateurs alike dissect every minute statistical detail. For the uninitiated, spend an hour listening to sports radio or reading sports blogs the morning after a local team loses a game (it doesn't matter what sport it is or whether it is a professional or amateur game), and you will begin to understand how data can drive people off a cliff.

Sports analytics is a cautionary tale for big data analytics. As you collect, measure, and analyze big data to help your organization change its business processes and make better business decisions, make sure that you don’t lose sight of the thin superstitious line that separates the Data Scientist from the Chicken Man.