Friday, May 20, 2011

Can the Thunder Win without Durant?

Durant didn't score 40 in Game
Two, but OKC got the win
In dramatic, bounce-back, season-saving fashion, the Oklahoma City Thunder pulled out a win last night over the red-hot Dallas Mavericks at the American Airlines Center in Dallas. It was the first time all postseason that the Mavericks lost a game at home, and it tied the Western Conference Finals up at one game apiece. A huge reason that the Thunder were able to win Game Two was that they limited Dirk Nowitzki to 19 fewer points and 14 fewer free throw attempts. By keeping the Mavs' best player from dominating the game, Scott Brooks' crew forced the rest of the Dallas team to hit shots--and they didn't. However, the Thunder did not just perform differently on the defensive end on Game Two--their offense was completely changed as well. Despite scoring only six fewer points in Game Two than in Game One, the Thunder got 28 more points (50 instead of 22) from their bench. They also got 16 fewer points from NBA scoring champ Kevin Durant (see above) and 11 more points from the team's third-highest scorer, James Harden. This got me the Thunder really play better when Durant scores fewer points? Any team obviously wants its star players to score as many points as possible, but do the Thunder really succeed when Durant is such a huge part of the offense? Or do they need contributions from Harden and others to succeed? Well, we ran the stats, so hit the jump to check out the results.

To figure out whether it was Durant or Harden or somebody else's scoring performance that caused the Thunder to be successful, I ran a logistic regression of the regular season scoring data for every OKC player who met the following four criteria:
  1. Played in at least 42 (half plus one) regular season games for the Thunder
  2. Averaged at least 10.0 points per game in those contests
  3. Are still on the Thunder's active roster
  4. Has played in at least seven games thus far this postseason
The postseason sample size is too small to allow for statistically significant results, but a sample size of at least 42 games is sufficient. These criteria require that a player average double-figures scoring while playing a majority of the Thunder's regular season games (meaning no trade-deadline acquisitions). They also require them to still be a part of the Thunder's offensive gameplan in the postseason.  Therefore, we are looking only at key offensive contributors who, broadly speaking, have been a major component of the Thunder offense all season. That leaves three players for Oklahoma City: Kevin Durant, James Harden, and Russell Westbrook. Westbrook scored almost the same number of points in Games One and Two, but Harden's and Durant's varied across the two games. The results thus far from the Western Conference Finals suggest that the Thunder play better when Hardens scores a lot and worse when Durant scores a lot, but Westbrook's scoring has no real impact. Let's see if that's really true.

Logistic Regression of the Oklahoma City Thunder's Success
Based on Performance of Top Players
Kevin Durant
Russell Westbrook
James Harden
* Significant at .1; ** significant at .05; N=82
Standard Errors appear in parentheses

The regression reveals that the scoring pattern established in the Western Conference Finals is actually a bit of an anomaly. Despite the Thunder's success when James Harden (see left) picked up the scoring slack from Kevin Durant, that was not the case in the regular season. During those 82 games, a strong performance from Kevin Durant was highly predictive (Probability of 0.03) of an Oklahoma City victory. Harden, on the other hand, did not have a statistically significant impact on the Thunder's success. Though it appears he had a slight positive impact, that could likely be the result of random variation. This means that, over the entire 2010-2011 regular season, James Harden's scoring output did not significant help or hurt the Thunder. Sorry, bloggers and reporters who think that Harden is the most important piece of the Thunder's team. His impact might seem great to the naked eye, but the statistics do not bear that out. The interesting result, to me, is that of Russell Westbrook. Although he is a point guard and averaged nearly six fewer points per game than Durant during the regular season, Westbrook's scoring output was just as important to the Thunder's success as Durant's was. Over the course of the regular season, a poor scoring night for Oklahoma City's point guard spelled doom for the Thunder, while a good scoring night generally brought success. I guess it's not surprising that the top two scorers on the team were the most important during the regular season, while the third-best scorer wasn't that crucial. This does, however, show that the results of the Western Conference Finals thus far--at least as far as OKC's scoring and their game success go--have been a fluke. That's not a good sign for the Thunder.

No comments:

Post a Comment