• Drafting Errors and Decision Making Theory in the NBA Draft

      Sailofsky, Daniel; Applied Health Sciences Program
      Even with the recent influx of available data with regards to draft-eligible players and NBA teams investing more resources into scouting than ever before, NBA decision makers still struggle to consistently evaluate talent and select productive players (Berri et al., 2010) in the draft. In this paper, I examine the NCAA statistics and pre-draft player factors that predict both draft position and NBA performance for all NCAA players drafted to the NBA between 2006-2013. Following this analysis, I determine what errors NBA teams are making and how these errors relate to general decision making theory. To compare the predictors of draft position and NBA performance, linear regression models are specified for both draft position and NBA performance. The NBA performance model sample necessarily excludes players whose production cannot be assessed due to not playing a minimum (>=500) amount of NBA minutes, and therefore a Heckman (1971) sample selection correction is applied to the performance model to correct for this non-randomly selected sample. Both models are specified for the entire dataset as well as for subsets for position (Bigs, Wings, Point Guards) and conference size (Big Conference, Small Conference). The findings of this paper demonstrate that NBA decision makers continue to base their draft selections on factors that do not actually predict future NBA success, such as scoring, size, and college conference. Many of the decisions made by NBA decision makers relate to Heath and Tversky’s (1991) competency hypothesis, as front offices forego the use of reliable distributive data and select players according to their perceived knowledge. NBA decision makers also display risk averse behaviour (Kahneman and Tversky, 1973) and an insistence on sticking with the status quo (Samuelson and Zeckhauser, 1988) in their decisions. More specifically, this study also points to ball control and offensive efficiency as predictors of individual player success. These findings can not only affect NBA decision makers in the factors that they emphasize in player evaluations, but can also be used to change the way that sport executives think about general decision making and their own innate decision making biases.