An Update on the Thunder Offense
How this season’s star-studded roster has succeeded
When Oklahoma City traded for Paul George and Carmelo Anthony, there was reason to believe that the Thunder offense would be among the best in the league. But, while former Thunder guard Victor Oladipo has shined with the Pacers and former big man Enes Kanter has had a productive season with the Knicks, the Thunder have failed to produce at the championship-contending level that fans expected. Fivethirtyeight’s Chris Herring focused on the Thunder early on in the season, highlighting the late-game offensive liability of Andre Roberson, the inability of the Thunder bench to implement head coach Billy Donovan’s defensive strategy, and the isolation-heavy tendencies of their trio of stars. With the season nearing an end, the Thunder stand at 40–29, fourth place in the West, with the 12th-ranked offense and 8th-ranked defense. How can the Thunder improve their offensive strategy to maximize performance heading into the playoffs?
To answer this question, we look at play type data from NBA.com. This data records how many plays of each type teams and players run, like post-ups, isolations, pick and rolls, etc. We focus on only points per play and frequency of each play type for this article. We identify comparable teams to the Thunder, and then look at play type data for the Thunder players to see which players are responsible for high and low efficiency plays. For more on the use of team comparisons using play type data see Nylon Calculus and FiveThirtyEight.
Using Principal Component Analysis, we identify teams for comparison purposes based on frequency and efficiency statistics for each play type. Among teams in the top twelve of offensive rating, the most comparable teams to the Thunder are the Timberwolves and the Raptors, ranked fourth and third respectively. The plots below show frequency and efficiency for play type by team. League efficiency rank is the percentage difference from the league average in that play type, expressed as a decimal. So a team whose points per possession on isolation plays is 15% above league average will be colored light orange, according to the scale to the right of the plots. Importance is measured as the correlation of points per possession in that play type with offensive rating. The idea is that more impactful play types will have a higher correlation with offensive rating.
There is not much noteworthy in the teams’ top three frequent plays, transition, pick and rolls, and spot-ups: some analysts have blamed inconsistent spot-up shooting for the lack of offensive dominance, and modern NBA teams thrive on transition opportunities and ball screens. These claims are illustrated by the importance, frequency, and efficiency of those plays in the graph below. However, there are three realistic areas of improvement for the Thunder that could help them match the efficiency of the Timberwolves and the Raptors: isolations, post-ups, and off-ball screens. The Thunder isolate the second-most in the league, behind only the Rockets. However, they are below league average in points per isolation. In addition, the Thunder do not exploit optimal post-up opportunities, using Anthony 15+ feet from the basket rather than Steven Adams closer to the basket. Finally, the Thunder are well above league average in off-ball screens, and yet it is their second-least utilized play.
As far as isolations, the Thunder have been relying heavily on their three superstars to produce points by themselves. Among the 87 players who have isolated at least 50 times this season, Anthony, George, and Westbrook rank 35th, 43rd, and 46th, among the likes of Ben Simmons, Marcus Morris, and Jeremy Lamb. With over 10% of the offense coming on isolations, it’s hard to imagine any sort of sustainable offensive success coming from scoring that inefficiently.
Speaking of isolations and Carmelo Anthony, Anthony has struggled with efficiency on post-up opportunities this season, averaging 0.83 points per play, and yet posts up the most out of anyone on the team, 167 times compared to Russell Westbrook’s 125 and Steven Adams’ 79.
In the first video, we can see that Anthony’s predictability in his midrange post-up game allows the help defender to slide over and take away the middle of the floor. Forced to either drive baseline or pull up, he pulls up for an inefficient, contested, midrange jumper. The second case is similar, as the help defender comes all the way over from they weak side to block Anthony’s shot. Anthony appears to look over his shoulder at one point but does not spot a potential opportunity for a wide-open corner three for a solid shooter, Terrance Ferguson. Whether a function of fitting into a new role or something else, Anthony’s production from the post is down from last season, when he averaged 0.92 points per play, ranked 23rd overall in 2017.
If the Thunder want to limit these inefficient post-ups, they should play more through Steven Adams in the post. An underrated post-up player, Adams is averaging 1.09 points per post-up opportunity, the best mark on the team and third in the league among players with at least 70 post-ups this season. He can also do this:
Adams gives the Thunder a different — and more importantly a more efficient — look on offense with his ability to score or pass out of the post.
If Anthony is going to cut down on the number of isolation/post-up plays, what should he be focusing on instead? Anthony has been subpar on spot-up shooting this season. Anthony is shooting 36.6% on spot-up opportunities this season while averaging 0.97 points per play, 76th in the league among 101 players with at least 150 spot-up opportunities. The reason for this low clip is that many of his spot-up threes are contested. Though the sample size is limited, he is actually shooting well coming off of screens. Anthony is 14 of 31 with an effective field goal percentage of 58% coming off screens. He averages 1.19 points per play, up from 1.00 in 2016 and 1.09 in 2017. The Thunder can better utilize Anthony by focusing on getting him open looks off screens rather than contested spot-ups on the wings.
By replacing some of Carmelo’s inefficient isolations and post-ups with more designed screen plays, the Thunder will boost their offensive efficiency. In addition, giving more post-ups to the big-bodied Steven Adams will improve their post efficiency, and allow Adams to create more scoring opportunities for the Thunder wings. Looking back at the above plots, you can see how fewer isolations, along with more efficiency in post-ups and spot-ups, will bring the Thunder closer to the Raptors and Timberwolves, and perhaps lead to more offensive success and a better offensive rating.
All stats per NBA.com (unless otherwise noted) as of March 12, 2018. Find up-to-date stats on NBA.com
Grant Carr is a junior mathematics major at Kenyon College.
