Hollinger went and took some of the wind out of my sails this afternoon. It's impressive and interesting work which appears (the formula he uses isn't published with the article) to have all the benefits and drawbacks of one, true number formulas.
To take one example, I have a hard time seeing how Wilson Chandler's statistical profile scores slightly better than Julian Wright's.
Their 06-07 seasons:
And their freshman seasons:
Going down Hollinger's six factors indicative of pro success
1. Age: Wright is 12 days younge than Chandler. I'd guess that makes them even.
2. Steals: Wright averaged 3 steals/100 possessions as a sophomore and 2.9 S/100 as a freshman. Chandler averaged 1.3 S/100 as a sophomore down from 1.7 S/100 as a freshman. That's a significant, consistent advantage to Wright.
3. Blocks: Wright averaged 2.7 blocks/100 possessions as a sophomore and 3.4 BS/100 as a freshman. Chandler averaged 2.7 BS/100 as a sophomore and 3.3 BS/100 as a freshman. I would assume this one is even also.
4. Rebounds: Quoting Hollinger, "Boards, especially offensive boards, are a good indicator of future pro success as well." Wright got 159% as many offensive boards per opportunity as did Chandler last year and 115% as many offensive boards per opportunity as did Chandler in their freshman seasons. Chandler did have significant defensive rebounding edge in their freshman seasons (144%) but Wright was the better defensive rebounder last year (109%). Wright is a significantly better offensive rebounder. That the formula likely doesn't take into account Wright's improvement as a rebounder between his freshman and sophomore seasons is a limitation.
5. 3-Pointers: The formula would be really helpful here as this could be where Chandler must make up a lot of ground on Wright. Chandler has made 41 more three-pointers than Julian Wright in their college careers. Of course, Chandler needed 137 attempts to make 41 threes. Julian Wright can't make three-pointers. This is a fact. At least he has the good sense not to attempt them.
6. Pure Point Ratio: Whatever ground Chandler doesn't make up with his three-point shooting is made up here. Pure point ratio = (((A*2/3)-TO)*100)/Min. Wright and Chandler both played 1048 minutes last year. Chandler had 48 assists and 54 turnovers. Wright had 82 assists and 91 turnovers. Chandler's PPR is -2.1. Wright's is -3.5.
I would have guessed that Wright's extra assists and slightly better A:TO would be an advantage. I do not, however, write basketball analysis for a living. This is something I obviously need to study.
That I find it difficult to imagine a world where Wilson Chandler has roughly equal potential to Julian Wright as basketball players doesn't diminish the usefulness of Hollinger's formula. (Again, I'd like to know what the formula is but I understand why might not want to publish it.)
I don't think that Thaddeus Young, Nick Fazekas, Josh McRoberts, and Jared Dudley have more potential than Joakim Noah (Noah, Julian Wright, and Brewer really look bad in this formula. They are also three of the best defensive players in the draft in that they were good defensive players in college and they figure to be physically able to guard at least one position well in the NBA.) but I'm also the guy who thought that PJ Tucker deserved to be taken in the lottery last year. I'm used to being wrong some of the time and I like learning new things.
Here's hoping Hollinger continues to discuss the particulars of his formula and that he's wrong and I'm right about some of these players.