The NBA preseason – as noted yesterday – is not a very good sample. Teams play no more than eight games. Many of the players in these games will be playing in some other Association this year or just enjoying really good seats at NBA games. And these games do not mean much to the NBA regulars who participate. Still, for the NBA rookies it is our first glimpse of what these players can do against mostly NBA players.
Like the analysis of the entire NBA offered yesterday, I began the analysis of the rookies with data from Doug Steele’s NBA and MLB Stats Home Page. Once again I calculated Win Score per-minute for each rookie, and because what matters is how a player performs per-minute relative to his position, I adjusted for position played by subtracting off the position averages (noted HERE).
Not only did I look at what these rookies did in the preseason, I also calculated each rookie’s college Win Score per-minute – adjusted for position played – for each player’s last year of NCAA basketball. I then ranked each rookie in terms of both adjusted preseason performance and adjusted college performance.
Before we get to the players, two issues need to be noted. First, players play much better in college relative to the NBA. And that is to be expected. In college these rookies were star players going up against inferior talent. Now these players are trying to fit in on NBA teams while playing against the best players in the world. The second issue is that the correlation coefficient between adjusted preseason and college performance is 49%. So there is some correlation, although college performance does not provide much predictive power for those looking to forecast preseason performance (as if you would want to do this). I will note that that college performance does a better job of forecasting NBA regular season performance, and that will be discussed in more detail in the future.
With the basics explained, HERE is the list of the 24 drafted players who both played 100 minutes in the preseason and also played in college last year. At the top of this list is Paul Millsap. And this is of course no surprise.
Okay, I am kidding. Millsap was a second round draft choice of the Utah Jazz and hardly a household name (even in Utah). Still, Millsap did star at Louisiana Tech where he averaged 19.6 points and 13.3 rebounds per game. Although Millsap led the 2006 draft class in rebounds per game, coming from a relatively small school downgraded his draft status.
In 138 preseason minutes it looks like Millsap does indeed have some ability to rebound the basketball. An average power forward captures one rebound every four minutes played. In the preseason Millsap grabbed one rebound every 2.33 minutes played. This would translate into 20.5 boards per 48 minutes played.
Can Millsap do this in the regular season? We do have two samples for Millsap and in both college and the preseason he has shown the ability to rebound. Unfortunately he is a forward, and this is precisely where Utah has talent in abundance. Two of the best forwards in the game – Andrei Kirilenko and Carlos Boozer – play forward for the Jazz. So we are probably not going to see much of Millsap this year to see if he can be productive when the games count.
Next on the list is Steve Novak, the player the Houston Rockets took with the 32nd pick. Novak’s strength is shooting. In college last season he hit 47% of his three point shots. In the preseason he took 42 shots and made 20, for a conversion rate of 48%. Will this continue in the regular season? Like Millsap, Novak happens to play the same position as Tracy McGrady. If Novak does play major minutes for the Rockets, that will be a very bad sign for this team.
If we are looking for rookies who might play this year we probably should look at the first round choices. Of these, five players were above average in the preseason. This list includes Rajon Rondo, Tyrus Thomas, Marcus Williams, Jordan Farmar, and Shawnee Williams. Of these, only Rondo and Thomas were ranked in the top half of the listed players in terms of college performance. So we now have two data points that suggest Rondo and Thomas might be good pro players. And we definitely have conflicting data points for a player like Farmar, who was both turnover prone and an inefficient scorer in college. It is hard to believe that a player who had such problems in college could play better in the NBA.
It should be emphasized again that preseason data is not a very good sample. Consequently fans of Adam Morrison and Brandon Roy should not be too dismayed. In fact, I still think Roy should be the favorite to win the Rookie of the Year.
Two more points need to be made about the preseason data. First, none of the international players – who did not play college basketball – were above average in the preseason. Again, not sure that means much for fans of Andrea Bargnani.
The second, and final point of this post, is that Renaldo Balkman – the much maligned choice of Isiah Thomas – was above average in 95 preseason minutes. And if we look at Balkman’s college performance we can see why Thomas spent a first round choice to acquire his talents. In the future I think a post on Balkman might be a good idea.
– DJ
Our research on the NBA was summarized HERE.
Wins Produced and Win Score are Discussed in the Following Posts
Simple Models of Player Performance
What Wins Produced Says and What It Does Not Say
Wages of Wins Journal 
Charles Follymacher
November 6, 2006
In my fantasy pool, I reached down and got Millsap, a move greeted with a rousing chorus of “whooooooo!?”s
I wish he’d get some more minutes, but I guess coach is breakin him in real slow like. Guy ahead of him doesn’t suck (if he’s injury free).
dberri
November 6, 2006
Charles,
I can’t believe Millsap will pay off in fantasy basketball this year. You are basically betting that Boozer and Kirilenko are going to get hurt. Even if that happens, I am not sure Millsap plays. And remember, although he did rebound in college and the preseason, we do not know for sure he can do this consistently in games that actually count.