Thunder Numbers: The more things change
(Editor’s note: These numbers were compiled before Saturday night’s NBA action, so keep that in mind.)
I think we can all agree that the Thunder have looked a little different than we remember them being last spring. Without much change in the roster, the Thunder were expected to pick up right where they left off last season. A few blowout losses and a couple close games against not so stellar teams have raised concerns, despite a pretty reasonable 5-3 record. Last night’s win was reassuring, but something still feels off. For this week’s column, I’ll looked at the Thunder’s stats over the first eight games of the season to identify where the Thunder are deviating most from last year, and I’ll throw in some speculation as to whether these trends are likely to continue.
First, a caveat: Whenever anyone quotes any stats this early into the season, sample size problems should absolutely be considered and everything should be taken with a grain of salt. Right now, a cold stretch or a big game is amplified to such a degree it is often hard to tell what’s really going on. To avoid having to mention this after practically every sentence in this article, I’m mentioning it now.
We should begin by looking at the big picture. I believe the best way to do that is to look at offensive and defensive efficiency ratings, and Dean Oliver’s “Four Factors of Basketball Success”. The offensive and defensive efficiency simply refer to points scored and points allowed per 100 possessions. The Four Factors, effective field goal percentage (eFG%), turnover percentage (TOV%), offensive rebounding rate (ORR), and free throw rate (FTR), were identified by Dean Oliver as the prevailing, mostly uncorrelated, factors that contribute to winning in basketball. Looking at these factors for a team and their opponent accounts for how efficiently they score (eFG% and FTR) and how many extra scoring chances they have (TOV% and ORR). Below I’ve compiled the change in these factors and scoring efficiencies for NBA teams from last season to this season.
A number of troubling things can be seen from this table. Based on scoring efficiencies, the Thunder have declined on both the offensive and defensive side of the ball (a higher defensive efficiency, more opponent points/ possession, is bad). League scoring has skewed 1.5 points per 100 possessions lower on average (this is not atypical at the beginning of a season) so the offensive decline isn’t as troubling, but defensively OKC has given up 6.1 more points per 100 possessions, which is 7.8 (!) points per possession higher than expected. To put this in context, the Thunder have had the largest decline in defensive efficiency of any team in the NBA, and the second largest decline in efficiency differential. In a lot of ways the team the Thunder are most similar to in the above chart is Cleveland, only OKC didn’t just lose the back-to-back MVP of the league. This doesn’t bode well for the rest of the season because efficiency differential has been shown to be the best predictor of future success, better than W-L record or point differential.
So what’s behind this loss in efficiency, especially on defense? Looking at the Four Factors, the Thunder’s eFG% and opponent’s eFG% have deviated vastly from last season. In fact, OKC’s offensive eFG% has changed about as much as their defensive eFG%, but seems to be buoyed by a huge free throw rate improvement. This improvement in free throw rate is third highest in the league despite the fact that the Thunder were already stellar in this category last year. At the current rate the Thunder have more free throws attempts per game than any team since the 1998 Lakers and the most made free throws made per game since 1985. This can’t be expected to continue at this rate, but it does bode well in general. The changes in turnover rate and offensive rebound rate are small in comparison to the huge changes in the other two factors, and probably aren’t contributing much one way or another.
To look at what’s causing our shooting percentages to differ so much from last season with largely the same personnel, we can look at how frequently and how well the Thunder and their opponents are shooting at different locations on the floor.
First the defense:
Defensively, it seems like OKC is having problems across the board, as the FG% for Thunder opponents has increase at each location on the floor. The surprise is where the biggest problems are popping up. The areas with the biggest increase in FG% for Thunder opponents is on the perimeter, with the FG% for 3-pointers 10% higher and for long 2’s over 6% higher. This is a big surprise because generally OKC’s length on the perimeter is their strong suit and no personnel changes have occurred for the guards or wings. Three point FG% was indeed one of the Thunder’s specialties last year, with the Thunder finishing third in the league, but this year they are fourth worst. I am at a loss to explain this, but let me throw out some ideas wildly without much evidence. It is possible that the Thunder are conscious of their mismatches inside, and when they double team they are slow to close out on the shooter. It is also possible that their experimentations with small ball lineups are hurting them, giving them more of a need to pack it inside on D, and less length for the close-outs. I’m not sure I believe any of these options, but certainly something is up. Some have been holding onto the hope that when Nick Collison returns to the lineup the Thunder defense will improve, but given that a lot of the problem seems to be on the perimeter, I don’t think we can count on that to solve everything.
Now to the offense:
Looking at the Thunder’s offensive shot locations, one thing stands out, and that’s three-point shooting. OKC has an eFG% that is 16.5% below what it was at last year. That is ridiculous. The Thunder would be scoring 2.7 more points per game if they were hitting 3s at the last season’s rate. Even if the Thunder were shooting at the worst rate of any team last year, they would be scoring 2.1 more points per game. OKC is unlikely to be a “great” three-point shooting team without a change in personnel, but given that we can probably rule out “historically awful”, some improvement should be expected. Something to note is that the Thunder are actually shooting better this season in 3 of the 5 shot distances, so besides the three-pointer, things are generally good. Despite the fact that the Thunder are often criticized for needing a scoring low post presence the Thunder are improved and above the league average in FG% at the rim and less than 10 feet. The lack of ball movement and assists has been a frequent point of criticism for the Thunder since the start of the season, and it is shown clearly here. The Thunder have dropped significantly in percentage of field goals assisted in every shot location. The 3pt AST% is remarkably low, suggesting a decrease in shots from kick-outs and a large increase in shots from pull-ups, which may explain some of our three point shooting problems.
Individually, who is causing our problems offensively? Looking at the statistics from the 8 players who played significant minutes both seasons, it seems that three players shoulder most of the blame. Durant, Harden, and Kristic are shooting much worse than they were last season, and given Durant’s huge percentage of the team’s field goal attempts, he is responsible for almost half of the team’s eFG% decline this season. In previous seasons Durant has shown himself to be a poor starter, and of everyone on this team I think KD has earned the benefit of the doubt. I’d be surprised if he didn’t turn it around fairly quickly, and if he does then practically the whole problem is solved. Harden is also having a poor showing, as many fans already know. His assist and rebounding rate declines are particularly troubling, suggesting that not only is his shot off, but that his effectiveness and perhaps effort have been reduced in other areas, too. Based on my pseudo-psychotherapeutic analysis of his tweets, I’m worried about his confidence more than his skills and would like him to get some extended minutes regardless of his poor play. (Unfortunately I don’t think there is much data about melancholy-tweeting basketball players and their historical trends.)
On the plus side, there is Ibaka and Westbrook. Westbrook’s numbers are improved across the board, and it seems like both he and Green learned how to improve their foul drawing this summer. The team’s free throw rate increase is predominately due to their contributions, and those 8+ extra attempts a game have really bailed out the Thunder offense at times this season. While it is certainly a dramatic increase, I think it can definitely be maintained, because it isn’t a matter of luck, but of decision-making on the part of Russ and Jeff. They are both athletic and skilled enough to attack the basket constantly, and if they settle for jumpers less than last season, their free throw numbers should remain high.
Ibaka was actually leading the NBA in individual offensive rating for a while earlier this season, and has been shooting much, much better from the free throw line and on the court. This is one of the trends to be most skeptical about, because Ibaka doesn’t take a ton of shots, but I believe in its reality. Ibaka could make those short jumpers all day, and it doesn’t seem like many defenders have realized it yet.
So, that was a ton of data, let’s wrap this up. Basically, based on what I’ve seen so far, I’m pretty optimistic about this team’s offense. With Harden and Durant playing far below what is reasonable to expect from them, and with Westbrook and Ibaka playing quite well, I think OKC’s offense could be significantly improved from last year when all is said and done. The three-point shooting can’t possibly continue to be so bad, and while I think the free throw rate will drop a little, it will hopefully remain high as Westbrook continues to harness his explosiveness.
However, the defense remains a big concern. In a lot of ways, the Thunder surprised everyone last year with their great defense. Was it a fluke? Has it been figured out by the rest of the league? I don’t know. The defense also declined over the second half of last season, so there is always the possibility that this is just a continuation of that skid. I do expect the opponent three-point shooting to decline, but expecting it to end up close to where it was last year is asking for too much. The weird thing about all of this is that both on offense and defense, it is our perimeter game that is struggling. As much as everyone has been hoping for a dominant big, the Thunder are getting killed on the perimeter on defense, and only outside shooting is preventing OKC from being great on offense. The Thunder have a lot of talent in the backcourt and on the wings, so hopefully this is a solvable problem. But of course, it’s early and we can still hope this all is just a problem of sample sizing…
