If you watched a lot of NCAA Tournament games last year, you most likely heard the story of Butler graduate manager Drew Cannon.
In short, Cannon was put on Brad Stevens' staff as an advanced statistics expert, going through numbers to help Butler put its best lineup on the floor at all times.
Before Cannon was snatched up by one of the game's smartest coaches, though, he was guest-writing for KenPom.com, submitting blog posts last summer that gave statistical projections for the nation's top players.
He used different formulas for returners (taking into account past stats, basic demographics, team stats, high school rankings, mock draft projections and awards) and freshmen (basic statistical information along with a few other secret ingredients added in), but in essence, he gave us a baseline on what to expect statistically from individual players.
I want to use this blog post to evaluate KU freshman Perry Ellis' performance last year to see if he went above or below what was expected of him, but first, let's look at just how close Cannon was to projecting KU center Jeff Withey's actual performance.
No wonder Cannon is on Stevens' staff; it's almost like he had the answers to the test given to him beforehand. Cannon's projections almost are the exact replica of Withey's production at KU his senior year. Perhaps the only exceptions are that Withey fouled a touch less than expected and made a few more twos.
Still, it's amazing to think that a formula could so closely predict human performance.
Switching over to Ellis, let's take a look at his projections compared to his performance a year ago.
Though Ellis' minutes were limited (he averaged 13.6 minutes per game), we can see from the numbers that he greatly exceeded Cannon's projections, especially on the offensive end. He averaged 1.14 points per possessions used while taking on a larger offensive load (ending 21.9 percent of KU's offensive possessions when he was on the floor) than expected.
Here are a few other positives from Ellis' numbers:
• Turnovers: This is where Ellis overachieved most. He gave the ball away on just 10.7 percent of his ended possessions, and though it's often overlooked, that kind of ball security from a big man greatly enhances his offensive value. Ellis had just 20 turnovers in 503 minutes last season.
• Free throws: Ellis also helped his offensive output by getting to the free-throw line often (a 52.5 free-throw rate is a solid number) and making those shots once he was there (his 73.8-percent accuracy was nearly 10 percentage points better than his projection).
• Defensive rebounding: Though KU coach Bill Self often pushed Ellis to be more aggressive, his defensive rebounding percentage ended up well above average. Ellis' 19.9-percent defensive rebounding percentage was second on the team behind Withey and much higher than his projected total of 16 percent.
If we're using the projections as a guide, here are two areas of improvement for Ellis in the offseason:
• Overall defense: Ellis wasn't a disruptive defensive player last season. His block percentage was barely half of his projection (2.1 percent), while his steal percentage also wasn't as high as you'd expect for a player with his quickness (1.8 percent).
• Two-point shooting: Ellis finished with below-projection two-point numbers despite an impressive end to the season. In his final seven games, Ellis was 30-for-46 on twos (65.2 percent), which lets you know just how much he struggled early. As an undersized 4, Ellis will have to continue his evolution offensively, learning how to create space and also avoid blocks against taller competition than he faced in high school.
The numbers above indicate that Ellis is ready to take the next step for KU in 2013-14. He's already a good offensive player — thanks to his low turnover count and ability to create and make free throws — and with some improvements defensively and on two-point jumpers, he could quickly turn into an all-conference-type player with an increase in playing time next season.