Introduction
For the Stat 3021 project, I was interested in determining how NBA draft position impacted a player’s performance. More specifically, I aimed to determine “Do players drafted 1st overall tend to have a higher average win share per 48 minutes (WS/48 — see https://www.basketball-reference.com/about/ws.html for information on the calculation of this stat) than players drafted 2nd overall in their rookie seasons?” I hold a longstanding passion for basketball, and I especially am interested in front office decision making. This project allowed me to explore the difference in value between the 1st and 2nd overall picks and better understand how management teams see these picks.
The first pick is understandably considered to be more valuable than the second pick, as the first pick allows the team which controls it to have choice of any player. Because players that are chosen first are considered more desirable than players chosen second, I believe that players picked first would have a higher rookie WS/48 than players chosen second. Statistically, this would mean that if μ1 is the mean WS/48 of players chosen first overall, and μ2 is the mean WS/48 of players chosen second overall, my hypothesis is that
H0: μ1 = μ2
Ha: μ1 > μ2
The above hypothesis …show more content…
The year of the draft has an impact on the players’ performance, due to overall player quality in a draft, (for instance, a player might be picked #2 in a strong draft, but #1 in a weak draft) league performance averages, (some years the average WS/48 is higher than other years) and other factors, such as lockouts. Since the performance of each player is dependent on the year, pairing the players based on year drafted minimizes the effect of year-to-year variability. For this test, the significance value will be set to α =