When Game 6 of the World Series was rained out, manager Tony La Russa went to see the movie Moneyball. His impressions? "Good acting," he said, eliciting laughter from reporters. "I'm serious." Of course, La Russa, a former Oakland A's manager and a baseball traditionalist, did not stop there. He criticized the accuracy of the plot — that the 2002 A's won 103 games and the American League Western Division title largely because of sabermetrics.
It bothered La Russa that the movie brushed over the contributions of the league's most valuable player, Miguel Tejada; the Cy Young Award winner, Barry Zito; and standout hitters like Eric Chavez — all of whom the A's developed through traditional means — while emphasizing a newfound reliance on on-base percentage and the switch of Scott Hatteberg to first base from catcher. "I was offended because of what the book represented," La Russa said. "I knew a few of those guys as scouts. It strains the credibility a little bit."
Having also seen the movie and reflecting on LaRussa's comments, I wanted to draw an analogy to the emerging management practice of Predictive Business Analytics (PBA). Is this a situation where this management practice, as applied to developing a championship caliber baseball team, using an innovative and somewhat controversial analysis tool like sabermetrics can highlight a few important lessons that can be applied to business?
Let's set the stage and define what sabermetrics is. Sabermetrics is concerned both with determining the value of a baseball player or team in current or past seasons and with attempting to predict the future value of a player or team. Many areas of study are still in development, specifically in the area of performance measurement.
Sabermetricians frequently question traditional measures of baseball skill. For instance, they doubt that batting average is as useful as conventional wisdom says it is because team batting average provides a relatively poor fit for team runs scored. Sabermetric reasoning would say that runs win ballgames, and that a good measure of a player's worth is his/her ability to help their team score more runs than the opposing team. This may imply that the traditional RBI is an effective metric; however, sabermetricians also reject RBI, for a number of reasons. Rather, sabermetric measures are usually phrased in terms of either runs or team wins. For example, a player might be described as being worth 54 runs more than a replacement-level player at the same position over the course of a full season, as the sabermetric statistic Value over Replacement Player can indicate.
Now let's draw a few parallels to the world of business and describe how analytical tools such as regression analysis may be properly used to support PBA as an effective management practice:
- First, the accumulation of key cause and effect relationships to form a composite relationship is greater than any one single relationship which is a cornerstone to effectively using PBA. In effect, stop searching for the "Holy Grail" in your business. It likely doesn't exist and the effort to search subtracts from more constructive uses of the enterprises resources to discover relevant forward- looking measures that are one of the guiding principles of PBA;
- Second, it's worthwhile to challenge traditional measures for these may not always reflect current business models, competitive landscapes, and economic environments but also may mask or obscure better, more informative measures that exist but have not been used in the past;
- Third, attributing 100 percent credit to the practice likely overstates its true contribution, but more importantly it detracts from the contribution of other factors. In effect, PBA is a systemic management practice that leverages people, process, and technologies and in the hands of a skilled artisan can be a powerful enabler to anticipate/predict meaningful outcomes. Although some industries can use automated protocols, such as computer-based trading systems, most industries or lines of business benefit from the "artisan way" of using forward- looking measures and information; and
- Fourth, management judgments regarding which relationships to include and which to exclude or minimize is critical and the overall anticipated outcome(s) is not a calculation but should be used as a guide to seasoned decision-making.
Larry Maisel is the managing partner at consulting firm DecisionVu and a leader in the field of strategy and performance management. He is a regular contributor to Business Finance.