THIS WAS THE MOST INDIVIDUALISTIC TEAM IN NBA HISTORY. No team apparently in history (Wilt Chamberlain's 50 ppg season is a contender we can't evaluate) had one player use as high a percentage of its offensive possessions as these Thunder. (That doesn't even capture all the free throw rebounds he got.)
That list is below and you can see that Westbrook and Kobe Bryant's 2006 season are pretty far ahead of any other players in history. The top 20 here consists of guys you know with one name: Lebron, Iverson, Jordan, Kobe, Wade, Harden, Stackhouse, and Arenas.
A few notes on this list:
With the acquisition of Paul George, Westbrook shouldn't repeat on this list. Instead of using 40+% of team possessions when on the court, Westbrook should drop to the low 30% range. Per this skill curve, his efficiency should go up a few points, but not a ton. In fact, his 2016 efficiencies weren't all that different than in 2017. That is remarkable in two ways, one good and one bad. The good way is that players usually start taking worse shots or committing more turnovers as they try to do more; Westbrook seems to overcome that. The bad way is essentially the flip side of the same coin - he doesn't make himself a threat as much when he's off the ball.
I had to change the axis range in order to display the OKC players. By Method 3, Westbrook was worth over 600 points over an average player (roughly 20 wins). By Method 4, he was worth almost 900 points over an average players (roughly 30 wins). In order to see some of the differences across other players, you gotta cut the maximum on that axis.
Is it possible that Westbrook was worth 900 points over average? Based on how much he played, that would be worth roughly 35 wins over a replacement player. That would mean that the rest of the team is worth about 12 wins. That would mean that the trade that "fleeced" the Bulls was useless. That would mean that Steven Adams ain't worth a thing. Same thing with Andre Roberson and Enes Kanter. That would mean that Westbrook's season was the best in history by a gap so big that it is as big as the gap between #2 and #17 on the list. I really don't think that the high end numbers are possible for Westbrook. I won't go into details, but the statistical tricks that go into a couple of these methods can blow up with extreme cases like Westbrook even when they are fine most of the time. I think that is happening here. What surprises me is that the wins about add up over the Thunder even with this error.
I am surprised at how many methods think that Kanter is a bad player. Not that he's great, especially defensively, but he contributes offensively very well around the rim, whether in the half court or in transition. He is one of the more uncertain players based on the variation in metrics' opinions of him. (The most uncertain is Westbrook because of Method 4 ranking him so far off the charts.)
Roberson is a good wing defender (by any metric that splits it out), but I can see why his overall value is debatable. His offense almost never comes out as something you have to deal with. He isn't a great transition player either.