Abstract
Research suggests that accuracy of professional sport draft selections, and talent identification accuracy more generally, is quite low (Koz et al., 2012). However, North American professional sport draft decision making, and accuracy research, has most frequently been conducted at a league aggregate level, and not at the team level (Teramoto et al., 2018). As such, the intra and inter-team variation in selection accuracy remains unknown. The main goal of this study was to explore the 30 National Basketball Association (NBA) teams and their draft selection accuracy over thirty-five seasons (1980-2015). This project carried out an analysis of entry draft position data and subsequent player career performance data (e.g. games played, and seasons played) using an overall team yearly draft accuracy metric, the coefficient of variation (CoV). The results suggest that draft picks made in the first round have become more accurate over time, where the league demonstrates a CoV of 0.837 from 1980-89, and 0.421 from 2010-2015. Second round draft picks have had the same relative accuracy across the league, ranging from 1.06 to 1.25. At the intra-team level, the Charlotte Hornets, New Orleans Pelicans, and Toronto Raptors have had the most accurate selections from 1980-2015. This research may allow us to better understand the impact of secular events (e.g., league expansion, draft rules/structure) on draft accuracy. In future applications, results can be used to further examine the correlation between draft accuracy at the organizational level to see what common dynamics exist among teams with different levels of accuracy.
