Analytics
News, events, thought leadership and more.
In an entertaining keynote to kick off the Alteryx Inspire 2013 Conference in Phoenix today, Paul DePodesta of “Moneyball” fame, told several anecdotes from his experience as a baseball executive that also have significant relevance in the world of business analytics.
He spoke about having an analytic mindset - realizing (and redirecting) when emotion (rather than data) is driving a decision and avoiding “affirmation-confirmation” confusion, a tendency which drives people to seek out data that supports a view they have formed already or a decision they’ve already made.
Currently a Vice President for the New York Mets and best known by many for his time as assistant to Oakland Athletics General Manager Billy Beane (chronicled in the book and later movie “Moneyball”), DePodesta related examples of how using analytics has transformed the way baseball players are evaluated over the last decade to the similar challenges and trends he sees in businesses. He cautioned against habits that fly in the face of analytics such as creating false correlation and causation, which he illustrated with the example of former Oakland Athletic Miguel Tejada. Recounting the year after Tejada won the American League MVP award for the A’s, during which he struggled with his hitting early but later caught fire, DePodesta pointed out that no matter how Tejada was performing people wanted to attribute the cause to him being in the last year of his contract. In reality, there was no evidence that the contract year was causing either his poor performance early in the year or his outstanding performance down the stretch – in fact, Tejada performed that year almost exactly as he had the previous few years.
DePodesta also spoke about the dangers of gut-based decision making vs. data driven decisions, illustrated by a story of the Cleveland Indians meeting during which the team decided to trade Jeff Kent. According to DePodesta, that decision was made based solely on one scout’s comment : “That guy (Kent) has the weakest friggin’ hack I’ve ever seen.” Kent, DePodesta explained, later became National League MVP (for the San Francisco Giants) and the top hitting second baseman of a generation – cementing DePodesta’s view that far too much subjectivity (and not enough true analysis) was involved in evaluating baseball players.
DePodesta also called out several other best practices for achieving analytic excellence and avoiding some common pitfalls. Process excellence that drives outcomes via data analytics is essential according to DePodesta, as opposed to latching onto excellent outcomes that might be the result of luck more than process and thus will not be repeatable in the long term. Keeping a “Decision Diary” was one of the best practices used by the A’s, a process in which all decisions were documented along with the key criteria used in each decision as well as the level of confidence in the decision. This prevented the team from rewriting history and understanding when analytics were really driving successful decisions vs. when outcomes were really more the result of luck.
DePodesta’s talk seemed to resonate with the business analysts, Data Artisans and Line of Business executives attending Inspire 2013, as most face similar challenges of creating an analytic culture and analytic process excellence in their own organizations.
Rick Schultz
SVP of Marketing, Alteryx
