Part of the statistical revolution means viewing sports as a science. However, there is a problem with this mindset: There is no laboratory in which we can work. In truth, statistical analysis in sports is more akin to a social science; we interpret the results the league gives us. Sometimes, fortune smiles on us and gives us a useful test case. This year, we have two teams, both in the same division, representing the opposite extremes of a question that has challenged basketball analysts--the importance of shot creation as compared to efficiency.
On the night of the 2006 NBA Draft, the Houston Rockets swapped the rights to their lottery pick, Rudy Gay, to the Memphis Grizzlies for Shane Battier. Three years later, those two contrasting players represent their teams. Gay is a scorer with undeniable talent who has struggled to convert his potential into success; Battier is the ultimate role player, a top-tier defender whose intelligence and shooting ability make him an asset on offense despite the difficulty he has creating for himself.
This summer, Memphis has built its roster around scorers, bringing in Allen Iverson as a free agent and dealing for Zach Randolph. Add Gay and second-year guard O.J. Mayo and the Grizzlies boast four of the league's top 50 players in usage rate last season. All four players used a higher percentage of their team's possessions than the leading healthy Rockets player, point guard Aaron Brooks. Of Houston's four best shot creators last season, Tracy McGrady and Yao Ming are currently sidelined by injury (Yao will miss the entire season, though McGrady is expected to return at some point), while Ron Artest and Von Wafer departed as free agents. What is left is a team full of role players.
The SCHOENE projection system allows us to quantify how extreme both teams are in terms of the ability to create shots amongst the talent on hand. To make sure the books balance, SCHOENE must adjust players' usage rates so the team as a whole averages a 20 percent usage rate (with five players on the floor, this must always be true, at least when usage rates are weighted by playing time). The biggest downward adjustment in the league, naturally, belongs to Memphis.
Team Usg Fudge
San Antonio .916
Oklahoma City .924
To be clear, what this number represents is what we multiply each player's usage rate by to bring the team down to an average usage rate of 20 percent. The difference between the Grizzlies and the Spurs, who rank second, is larger than the gap between San Antonio and fifth-ranked Miami. On the other end, as expected, Houston has the highest usage fudge in the NBA.
Team Usg Fudge
New Jersey 1.079
The Rockets are as extreme as Memphis at the other end of the spectrum. The question, then, is how the two rosters play out on the court, and what this tells us about both usage and efficiency. I believe strongly in the evidence supporting a strong relationship between usage and efficiency, one that sees players score less effectively when they are asked to handle a larger offensive load.
The best study of the issue I've seen was done last year by Eli Witus (fittingly, now a consultant to Houston's front office). Most past efforts were tripped up by the problem of looking at usage on a game-by-game basis. Naturally, players will use more possessions on nights where they have a more favorable matchup, so it is not surprising that these studies actually found that players' efficiency rose as their usage increased.
Witus was able to avoid this issue by looking at usage on a play-by-play level from the team perspective. He looked at the combined usage rates of every five-man unit from the first four months of the 2007-08 season as well as how the lineup's actual Offensive Rating compared to what would be expected based on the players' individual efficiency. Using this method, Witus found the negative relationship we would expect between usage and efficiency.
"In general, for every 1% that a lineup has to increase its usage, its efficiency decreases by 0.25 points per 100 possessions, and vice versa," Witus concluded, further adding, "For each 1% a player increases his usage, his efficiency drops by 1.25 points per 100 possessions."
Drawing on this research, SCHOENE makes a corresponding adjustment to players' efficiency when their usage rates are forced up or down by the team's usage. This adjustment is a big reason why the Grizzlies are projected to improve from 28th in the league in Offensive Rating to 15th this season, while the Rockets' projection falls (gently) from 15th to 19th.
Of course, things might not play out nearly that cleanly on the court, and that's why this is a learning opportunity. Memphis may be more interesting in this regard because, while Houston's hand was forced to some extent by injuries, the Grizzlies intentionally constructed an extreme lineup. Make no mistake, this certainly is an unusual group. Based on the numbers Witus reported, about two percent of all possessions in his data set involved lineups with a combined usage rate of greater than 120 percent (as the sum of individual usage rates; obviously when they were on the floor together, their combined usage rate was again precisely 100 percent). As the Straight Outta Vancouver blog considered last month, the combined usage for a possible Memphis lineup of Iverson, Mayo, Gay, Randolph and Marc Gasol was 123.8 percent last year, which is extraordinarily high.
What spreadsheets can never take into account is the egos involved. When the Grizzlies' four scorers are on the floor together, which player will defer? Beyond that, it's uncertain to what extent these players will be able to succeed off the ball, since much of their value is tied up in their ability to create. I am somewhat less concerned about Memphis' passing. For all the jokes about assists being few and far between, Gay is the only one of the four scorers who does not have a solid assist rate for his position. SCHOENE's projections have the Grizzlies' ratio of assists to made baskets on the low side, but hardly historically so.
For the Rockets, the more pressing question is who steps into a leading role on the offensive end. In the Houston chapter of Pro Basketball Prospectus 2009-10 alone, three different possibilities are put forward. Our Bradford Doolittle, who authored the chapter (available in PDF form as an excerpt from the book), suggests newcomer Trevor Ariza could put his athletic gifts to use as a scorer after playing off the ball in the Lakers' offense. Rockets.com's Jason Friedman offers Luis Scola, who has been a go-to guy for his Argentine National Team at times, including this summer's FIBA Americas Championship. Meanwhile, Houston GM Daryl Morey is high on the potential for Brooks to continue the momentum of his excellent postseason in his first full campaign as a starter.
By next April, we'll have the answers to these questions and a better idea of how well SCHOENE handles usage rates at their extremes. No matter what the Grizzlies and Rockets ultimately achieve this year, watching these two experiments play out on the court will be part of the fun of the NBA season for statistical analysts.
Pro Basketball Prospectus 2009-10 is now available on Amazon.com. See our PBP 09-10 page for more details and to purchase your copy in printed form or as a downloadable PDF.
Kevin Pelton is an author of Basketball Prospectus.
You can contact Kevin by clicking here or click here to see Kevin's other articles.