6/11/09
by: Carlos Chaloub
Permalink
 |
There have been many attempts to predict college players' stats in the NBA. None of them works all the time (after all, nobody could have predicted Gilbert Arenas exploding or Adam Morrison completely busting), since a lot more comes into play than simple statistics: basketball IQ, work ethic, character, athleticism, injuries, opportunity, and system. However, it is always a fun exercise to do these projections, at least just to compare to a fan's hunches on draft night.
One of the latest systems used to project NBA stats was devised by Jon Nichols, from basketball-statistics.com. While Jon doesn't divulge the formula he used to come up with his projections, we can at least check out the fruits of his calculations for some of the players in the upcoming draft.
There are two main issues with this system. First, the stats are not adjusted for strength of schedule at the college level. A player such as Stephen Curry lit up the opposition in a very weak conference, so his projections will come up higher than they should be. Second, the projections are calculated per 36 minutes. Not all players will end up being starters and playing 36 minutes or more. Unfortunately, there is no way of knowing who will end up being afforded such an opportunity. For example, Dell Curry, Stephen Curry's father, had a 19.4 career PPG per 36 minutes. However, his defensive deficiency prevented him from starting for most of his career (99 starts out of 1083 games).
Using the already provided stats, I set to fix at least fix the first problem by adjusting for strength of opposition. I normalized the strength of schedule for the players and adjusted their stats accordingly. Using Jon's projections, I concentrated solely on the Shooting Guard class, as well as some of the more prominent players linked with the Raptors on draft night: DeJuan Blair, Chase Budinger, Earl Clark, Stephen Curry, DeMar DeRozan, Tyreke Evans, Jonny Flynn, James Harden, Gerald Henderson, Jordan Hill, Jrue Holiday, Brandon Jennings, and Jeff Teague.
First, their basic stats per 36 minutes: Click to view table.
Next, I used their basic stats in the context of the Raptors’ last season performance to measure some of their advanced stats:
Using PER (Player Efficiency Rating), it seems that only the top three or four players have a chance to be players in the league, and from this list four of the top five will probably be gone by the time the Raptors are called to pick, and Blair has huge question marks with regards to his weight and knees. The rest are projected to be destined for mediocrity.
Word of caution: This is a fun exercise and not intended to be used as the be-all-and-end-all when comparing players in the upcoming draft. 
|
