This is not a reaction to Oklahoma City’s recent loss to Dallas in the playoffs. If you read my posts over the course of the season (here, herehere, here, and here), I have constantly advocated the belief that Russell Westbrook is the best point guard in the NBA, and arguably a top-three player ahead of Kevin Durant. Let’s begin the analysis by looking at just how good Westbrook has been this season. Tom Haberstroh of ESPN recently wrote an article on this very topic. In one chart he compares Westbrook’s season to that of triple-double machine Oscar Robertson.

Westbrook and Oscar

Figure 5 via Nylon Calculus

What Figure 1 shows, consequently, is that Russell Westbrook’s season – when adjusted for possessions played per game – is greater than Oscar Robertson’s 1961-62 “triple-double” season. Thus, while Westbrook will never win this season’s MVP – something I am conflicted about – he is having a historically great season.

Though, for me to say the Thunder should “forget” Kevin Durant and build through Russell Westbrook I need to prove three things: first, that Westbrook is the same or better as a player; second, that building around Westbrook instead of Durant or both is a good decision; and third, that building around Westbrook is feasible.

Let’s begin by addressing the first issue: that Westbrook is a player of similar caliber or better than Durant. To this point, below is advanced data from basketball reference that begins to paint the picture.

Russ and KD

Figure 5 via Nylon Calculus

Consequently, Figure 2 shows that in many stats Durant is having a greater impact than Westbrook; yet, I want to point your attention to a statistic called “VORP.” What VORP does is isolate a players Box Plus/Minus (the points a team scores relative to their opponents when a player is on the floor. I.E., if a player has a +8 Box Plus/Minus, their team would have outscored their opponents by eight points when said player was on the floor), compared to the same score by an average player on an average team over an eighty-two game season, thus isolating a single player’s contribution.

VORP is amongst a group of stats that attempt to run a simple regression model on a player’s box score. In english, these stats work to isolate a player’s impact on their team’s success. Here, Russell Westbrook is noticeably better than Kevin Durant. Arguably, this means that Westbrook is harder to replace as an individual player than Durant. However, certain regression statistics dispute this narrative: Player Impact Estimate (PIE) and Regularized Adjusted Plus-Minus (RAPM). This is most likely due to Russell Westbrook’s significantly higher true usage percentage – total percent of plays a player is involved in – than Kevin Durant (eleven percent!). For more information, see Figure 3 below.

Dashboard 1-2

Figure 5 via Nylon Calculus has created their own contributory regression model known as PIE. What this attempts to do is find the percentage of all in-game events one player has contributed to. Here, Kevin Durant leads Westbrook 19.4% to 18.8%. Thus, even though Westbrook is effectively averaging the best stat-line in NBA history, PIE suggests Kevin Durant contributes to more in-game events. The issue with this, however, is that PIE does not differentiate the value of statistic-x over statistic-y. I.E., one point is viewed the same as one assist, and consequently volume scorers will be valued more heavily.

The statisticians at Nylon Calculus created RAPM to help alleviate this problem. Essentially, this statistic runs two regressions to isolate a player’s impact on the box score. Their results are in Figure 4 below:

Dashboard 1-3

Figure 5 via Nylon Calculus

Consequently this regression, like PIE, suggests Durant contributes slightly more to a game than Westbrook does. Yet, stats like PIE and RAPM suggest that the Thunder are better with Durant than with Westbrook, however they these models do not explain why a player’s score is worse than another, even though it demonstrates said score. For this situation we need to answer why Durant is more valuable to the Thunder than Westbrook.

The answer is fairly simple, and only slightly more complicated to explain statistically, essentially Kevin Durant is more superior at scoring than Russell Westbrook is at everything he does combined. Part of this is, due to Russell’s high usage, he is prone to miss shots and turnovers. Additionally, for players above 30% true usage percentage, Westbrook ranks a modest 12th at playmaking true shooting percentage (true shooting percentage on passes-to-assist). While this is impressive, it simply does not compare to Kevin Durant’s 63.4% true shooting percentage, especially considering he shoots on 62% of end-of-play opportunities.

Overall, Westbrook and Durant are fairly evenly matched players. Westbrook will fill up a stat-line, get his teammates involved, and play hard throughout the entire game. In fact, he is better at most aspects of basketball than Durant is; however, the latter is a vastly superior scorer, to a degree where he at least matches everything else positive Westbrook does. Nevertheless, neither are so significantly better than the other to make an argument based on talent.

Westbrook is a Better Piece to Build Around than Durant:

First, it’s important to look at playmaking…

Dashboard 1-5

Figure 5 via Nylon Calculus

This means that in plays which he contributed to the end possession, Russell Westbrook scores on 47% of these plays, passes on 43%, and turns the ball over on 9.6%. Moreover, as mentioned earlier, Westbrook ranks in the top-12 wing players in playmaking true-shooting percentage. This is in comparison to Durant, who predominantly benefits his team by scoring.

The dilemma with Durant’s play is it emphasizes the main criticism analysts have levied at the Thunder for the past four years: they are too isolation-heavy. When you have a player like Durant, who is such a prolific isolation scorer but struggles to get other teammates involved, it increases the necessity for an isolation-heavy offense.

Hammering this point home, the Thunder’s best five-man lineup in terms of pace adjusted box score is Westbrook-Morrow-Roberson-Ibaka-Adams. Coincidently, no Kevin Durant. This is because, when Westbrook is given ample time to develop with three-and-D players, the team can be very effective. Moreover, in the twelve lineups given in the link above, Kevin Durant only appears without Westbrook in one: the worst. Westbrook makes his teammates better, and perhaps even does so with Durant.

So the question is why should the Thunder opt to build around Westbrook instead of Durant, or even instead of both? On one end, and the side I affiliate with most, they probably shouldn’t. At the end of the day, Durant is a top-two scorer in the modern NBA, and has a skill-set that makes Westbrook a better player. However, due to Durant’s propensity to utilize isolation plays to score, the entire team gets frozen out. Thus, perhaps the best way to change Oklahoma City’s offense is to move on from Kevin Durant.

Finally, is a Thunder re-tool sans Kevin Durant feasible? If the goal is to build a team around Russell Westbrook’s talent then the answer is yes. A team built around Westbrook would have big men that can rebound at an elite level while not necessitating being under the basket to be effective. In Steven Adams, Serge Ibaka, and Enes Kanter the Thunder already have this.

Due to Westbrook’s elite drive-and-kick ability, wings that shoot three-point shots at a high clip and defend well are also necessary. Anthony Morrow and Andre Roberson do this fairly well, however, another “three-and-D” wing would be necessary. In this case, someone who could shoot three-pointers at a very high clip, such as a Kyle Korver, would be ideal.

Lastly, the Thunder would need a competent point guard who can run the floor when Westbrook is out. Perhaps Cameron Payne could be the solution here, as an insurance policy though, a player like Matthew Dellavedova would fill that void.

Ultimately, even a team with those players is most likely a first-round exit in the western conference. Because of this, it probably makes sense for the Thunder to attempt to re-sign Durant this summer, however they should not bend over backwards to do so. The Thunder can move-on from Durant, and while it will be hard in the short-term, it may behoove them in the future by increasing their upside. Nevertheless, removing a top-two scorer in the NBA is risky and could very well turn into a fool’s errand.

One thing I like and dislike about the past few playoff games:

-I love Dallas’ effort last night. They made the Thunder pay for lax defense and unimaginative offense. One thing is for sure, though, Kevin Durant will not play this badly again. Rick Carlisle will need to work more magic to give his team a chance to win a second game.

-I hate the Rockets. This team does not deserve to be in the playoffs, does not deserve to be on national television, and its fans deserve better. Not one player on their roster cares at all about this series against Golden State. They are the least entertaining team in the NBA because of attitude problems and I am praying the Warriors finish them off quickly so we can move on from bad, lazy basketball.



Please enter your comment!
Please enter your name here