Many analysts believe that the 2013-14 draft class has been the weakest crop of rookies in years–the “bust before the boom,” if you will. Michael Carter-Williams, the 11th pick by the Philadelphia 76ers, is a front-runner for Rookie of the Year honors. Meanwhile, the Cavs are still waiting for Anthony Bennett, the 1st pick, to show up. Now that we are approaching All-Star Weekend 2014, it is time to understand each rookie and his style.
Before you read any further, judge the 2013 draft class for yourself! Here is a list of the top 15 picks from last summer–how many have you heard of?
Given all the chatter on this topic, I wanted to get a holistic understanding of our current crop of NBA rookies. Assuming the current class is weak in aggregate, within-class comparisons are not that meaningful. Instead, I decided to look at the top 15 picks from each rookie class since 2003 (LeBron’s class) and compare their pre-All Star rookie performances to get a better perspective on how the current class is faring. Since almost half of the 2013 Draft’s top picks were guards, I will first start with them.
After removing players who did not provide enough signal (those who played fewer than 20 games in the pre-All Star period), 60 guards remained. I then aggregated key box score stats for each guard during essentially the first half of his debut NBA season. Using these stats, I “clustered,” or put guards into groups based on how similar they are to one another across all dimensions measured. I used the k-means clustering technique to distribute each player into four possible groups. (I determined the number of appropriate clusters to use for this analysis by plotting the total within-groups sum of squares against the number of clusters in a k-means solution and picking the one that resulted in a bend in the graph.) For those interested in the statistical nuances of this methodology, please click here.
The summary table below (click image to enlarge) highlights the average stats for guards in each of the four clusters across all dimensions considered in this analysis. Please note, in order to make the clusters more relatable, I describe each qualitative in the “General Description” column:
So you are probably wondering in which of these clusters this year’s rookies fall and what other players are they grouped with. To help answer your burning questions, I made the following visualizations–omitting some of the lesser known players (many of whom are no longer active in the league):
Clearly there are some inconsistencies between the pre-All Star stats and current player performance for more experienced players. The notable exceptions are in Cluster 3, or the “Unremarkable” group. Obviously, Paul George and DeMar DeRozan are very strong players. However, when you go back and look at their box score stats from the pre-All Star period of their rookie season, it is not surprising that they were put into this bucket. George averaged just 8.0 points on 30.6% shooting from three point range and 1.0 assists. Similarly, DeRozan score 8.3 points on 28.6% shooting from long range and averaged less than 1 assist per game. Interestingly, both players had relatively high overall field goal percentages. However, both are All-Stars this year, so they are definitely not “unremarkable” players!
With the perspective that this player grouping is a guess based on limited information early in a player’s career, let’s see where each of the rookie guards fall…
Ben McLemore, of the Sacramento Kings, falls in the “unremarkable” group. Though he opened the season fairly strong, receiving the honor of Western Conference Rookie of the Month for November, he has been struggling since December. He has been replaced in the Kings starting lineup and was not invited to participate in the Rising Stars Challenge (during All-Star Weekend). McLemore has the potential to be a solid guard in the league, based on his level of play earlier this season. We will just have to wait and see how he closes out the year.
Kentavious Caldwell-Pope, of the Detroit Pistons, joins McLemore in the “unremarkable” group. Like McLemore, Caldwell-Pope was also not invited to the upcoming Rising Stars Challenge.
Giannis Antetokounmpo (I’ll call him A-to-Z for short), of the Milwaukee Bucks, is part of the “Efficient & Strong Defensively” cluster. Though he did not start the season strong, A-to-Z had a strong December and is showing a lot of promise. It helps that he is still just 19 and has even grown an inch since draft day!
Michael Carter-Williams, of the Philladelphia 76ers, and Victor Oladipo, of the Orlando Magic, are both in the “High Volume” cluster. MCW has been a scoring machine so far this season, averaging 17.2 points and 6.6 assists per game. He is currently the leading favorite for Rookie of the Year, and his main area for improvement is his shooting efficiency. Oladipo’s stats are not as impressive, as he is adjusting to the point guard role (versus his natural preference to be a shooting guard). His main area for improvement is turnovers.
Finally, we have Trey Burke, of the Utah Jazz, in the “Strong Shooting” group. Burke was named NBA’s Western Conference Rookie of the Month for both December and January. He has been playing really well, leading all rookies in assists (at 6.8 per game) during the month of January. Though he struggled from beyond the arc last month, Burke was averaging 37% from downtown in December. He is also very precise from the free throw line.
We have just passed the midway point of the season, so there is still a lot of basketball left to play. This analysis serves to provide a more comprehensive perspective of individual rookie performances based on a simplification of the data, but it is not meant to directly compare one player with another. After all, what happens on the court, stays on the court…
Is Anthony Bennett a huge, regrettable bust? Is Cody Zeller underrated? Stay tuned for a similar deep dive on forwards and centers!