Now that the 2019 playoffs have wrapped up, the league's focus has turned to this year's draft. The NBA draft class of 2019 boasts a handful of elite prospects, two of whom played for the Duke University Blue Devils.
Zion Williamson, the Naismith award-winning lovable leaper from South Carolina, and RJ Barrett, the quiet Canadian wing and reigning FIBA U19 MVP, have been 2 of the top 5 prospects for more than a year. In this post, we will be analyzing how RJ and Zion helped Duke win games and how they compare to each other. Most of the data we will be using can found here. Some of the metrics will be found only in the production version of the application.
On/Off Metrics
We can start with analyzing the On/Off impact metrics for each of these players to gain an initial understand of how they impact the team. The On/Off metrics are just one tool that we will use to try and ascertain which player may have been more valuable on a per minute or per possession basis.
After checking the On Court numbers, we can determine that each possession with Zion was slightly better than each one with RJ. However, the # of possessions played matters as well. This impact is more clearly reflected in the Off Court numbers.
The On Court - Off Court differential numbers show that despite being slightly less effective when On the court, the team misses RJ significantly when he is Off. The offense completely falls apart and the difference between the On court and Off court numbers are greater than Zion's. Leading us to believe that RJ is actually more important/less replaceable than Zion. There is an obvious answer to this conundrum though, which we will discuss in the following section.
Zion's Injury
The impact of "shoe-gate" went beyond Nike's ticker price. Zion missed, for all intents and purposes, six full games at the end of the season. This is clearly the reason for the gap in # of possessions but also a driving reason behind the gap in both the On court and Off court differences between himself and RJ. During those 6 games, RJ played in his normal role with the starters. So almost all of the Off court time for RJ does not include "normal starter minutes", as he played all 38 games for Duke. The inverse is true for Zion, as a majority of his Off court time does include "normal starter minutes". So his relative On/Off impacts will be skewed. Additionally, Zion's injury would have negatively impacted RJ's On court impact. We can measure the synergy between these two players and these minutes are extremely impactful.
Teammate Synergies
As demonstrated in the above graphic, the combination of RJ and Zion is a potent one. This combination(+25.1) is an improvement on each of their individual net ratings(+19.3 and +23.2) and therefore has a positive synergy. Synergy for our purposes is defined as the average difference between the individual net ratings of two players and their joint net rating. The minutes missed by Zion would have negatively impacted RJ's On court numbers, resulting in a larger gap than normal. However, it is likely a gap still would have existed. We can analyze all of the two player combinations for RJ and Zion, as well as all other rotation players, in this tool.
The weighted average synergy(synergy score weighted by # of possessions) gives a single number to understand the positive or negative impact a player has on his teammates. Zion's impact is well of 2x that of RJ's, which can not only be explained by missing a few games. Additionally, we can now see a basic distribution of how these players affect each other as well as the team.
It is not a "hot take" to think that Zion is a better prospect for the 2019 NBA Draft and frankly that's not our job either. But what we can illuminate is how well he played at Duke, what his contribution meant, in terms of team impact, and how these are relate/compare to those of RJ Barrett. Additionally, we can demonstrate how the data we are generating and visualizing can be used to analyze a player's impact on his or her team, as well as understand how this player compares to other members of the team.
Continue reading... Part 2: Analyzing lineups over a season, picking a fifth starter and building the best rotation