The 2006 squad was the greatest of my Duke career. Though it bowed out to LSU in the Sweet 16, making Tyrus Thomas a rich man in the process, it was the #1 team for a good while during the regular season. The biggest reason for Duke’s success, most thought, was J.J. Redick. He won the Naismith Award, given to the finest college basketball player of the year, and if a writer thought he was the best player in the country, he certainly would have agreed he was the best at Duke University. My favorite player, though, was Shelden Williams, who seemed to always be there to bail us out in the paint. Earlier this summer, I studied the numbers and realized the sportswriters should have asked me before they voted.
Last week, my friend Shawn Kwatra posted my analysis of the 2007 Duke basketball team on his fine sports blog. He did a good job of explaining my methodology, but if you missed it, here it is again: I based my numbers on a system created by economists at the Wages of Wins blog, and I am in their debt not only for their system but also for their way of thinking about basketball. They base their model on the basic box score statistics, and it has a 95% correlation between predicted wins and actual wins, so it’s very solid. Their metric for performance is called Win Score, and it’s calculated thusly:
Win Score = Points + Rebounds + Steals + ½Assists + ½Blocked Shots – Field Goal Attempts – Turnovers – ½Free Throw Attempts – ½Personal Fouls
The value of points is obvious, but we unfortunately tend to gloss over other statistics which are just as integral to a team’s success. Rebounds and steals end the opposition’s possessions and give one’s team more chances to score. Blocked shots create rebound opportunities and also indicate how much a player helps on defense. Assists help quantify how much a player facilitates the offense. Field goal attempts, turnovers, and half of free throw attempts (the 2nd shot) end possessions for your team, so they subtract from the score. Personal fouls both end possessions and give free throw opportunities to the other team.
I have one more acronym to explain: PAWS (Position Adjusted Win Score). This stat compares a player’s win score to that of the average player at his position. (Is it strange that centers are more productive than guards? Not really: note how much teams obsess over finding good big men.)
I also added Win Score/min and PAWS/min to the table so we can more easily compare players with unequal minutes played.
Enough of that, here are the numbers (click on the thumbnail to expand):
A couple passing observations:
1. The numbers indicate neither Josh, Greg, nor Marty changed much between ’06 and ’07.
2. Jamal Boykin outperforms expectations again, albeit in garbage time. Should he have played more?
3. This team was clearly better than ’07: the Win Score differential is 1463.5-1019.5, although the ’06 team’s faster pace may explain a bit of that.
Moving on, here are how the seven regular rotation players fare:
Shelden Williams: 0.432 WS/Min (0.207 PAWS/min)
Josh McRoberts: 0.271 (0.056)
J.J. Redick: 0.185 (0.057)
Lee Melchionni: 0.150 (-.002)
Sean Dockery: 0.145 (.013)
DeMarcus Nelson: 0.133 (-0.019)
Greg Paulus: 0.080 (-0.051)
So Josh and J.J. were very good; Lee, Sean, and DeMarcus were decent; Greg wasn’t very good, and Shelden was ridiculous. No one from either ’06 or ’07 comes close to the Landlord’s year, either in absolute numbers or in comparison to average players at their own positions.
Why is Shelden ranked so much higher than J.J.? Here’s the tale of the tape:
1. Blocks and rebounds. I loved Shelden because he always seemed to bail us out when we missed shots or when the interior defense collapsed, and the numbers bear that out. He collected 12.8 rebounds and 4.6 blocks per 40 minutes. In 2006, he had 137 blocks. In 2007, the entire Duke team had 157 blocks.
J.J. averaged 2.1 rebounds per 40 minutes, and despite playing nearly every minute of every game, he blocked only two shots all year. These are awful numbers even for a guard, and they’re a big reason he can’t get off the bench in the NBA, where everyone is even bigger and even faster.
So far, Shelden is way ahead.
2. Assists. J.J. had 2.8 per 40 minutes, Shelden 1.3. Both are below average for their respective positions, but regardless: advantage, J.J.
3. Steals. Shelden had 60, J.J. 52. So Shelden has a slight edge in an area that should be an advantage for the guard.
4. Turnovers. Shelden had 89, J.J. 90. It’s a wash.
5. Fouls. Shelden had 114, J.J. 54. Advantage, J.J.
6. Shooting efficiency. This was J.J.’s greatest attribute, the one reason he was an excellent college player. Yet the numbers indicate that Shelden was even more efficient. J.J. had 1.5 points per field goal attempt, which is fantastic…but Shelden had 1.65.
How? Well, our opinions are largely shaped by our memories, but our memories favor the most vivid events. A three-pointer is easier to remember than a 5-foot jump hook. Thanks to the degree of difficulty, we also tend to remember missed 5-footers more than missed three-pointers. So great-shooting big men are sometimes forgotten because they’re just doing what they’re supposed to do.
Everyone remembers the nights J.J. went 9-13 and scored 30 points, but on the year, he shot 47%, including 42% on 3’s. These are fantastic numbers, but Shelden shot 58%. Even when you add the extra points from long-distance shots, Shelden was more efficient.
J.J. made 221 free throws on 86% shooting, Shelden 201 on 74% shooting. That’s an advantage for the former, but it’s not enough to move the table.
One could argue that Shelden would not have been as efficient a scorer without J.J. there to punish double-teams. That’s probably true. Shelden’s excellence, however, must have benefited J.J. as well, the way a healthy Shaq creates shots for his own smaller teammates.
Was J.J. superior to Shelden in heart, teamwork, and so forth? J.J. was a more vocal leader, but if I remember correctly, Shelden was no mercenary himself.
Most importantly, Shelden’s contributions in the other aspects of the game still far outweigh J.J.’s. (So do Josh’s, hence his own higher ranking.) Unfortunately, sportswriters are infatuated with scoring and forget about how teams get the opportunity to score in the first place. Note that the 2nd place finisher for Naismith with Adam Morrison, another prolific scorer who is struggling in the NBA because he is below average in every other category. J.J., I love you, but I can rest easier knowing your underrated counterpart also has his jersey retired in Cameron.