Skip to main content
Menu
The Case For Russell Westbrook, MVP

Russell Westbrook is the MVP. You are likely already familiar with Westbrook’s claim to the award because every conversation that suggests someone else is the MVP must do the work of explaining why it is not, obviously, Westbrook.

Westbrook’s case for MVP is self-evident. His season-long triple-double is a historic accomplishment, and its grandeur only grows when adjusted to account for the way the game is played today. Here are the top seasons for the triple-double stats sorted by John Hollinger’s Versatility Index, which shows how good players are at those three metrics, combined1, which adjusts for pace:

PER 100 POSSESSIONS
SEASON PLAYER POSS. PER GAME PTS REBOUNDS ASSISTS VERSATILITY INDEX
1 ’16-’17 Westbrook 97.8 44.8 15.1 14.7 21.5
2 ’16-’17 Harden 100.0 38.4 10.7 14.8 18.3
3 ’15-’16 Westbrook 96.7 33.9 11.3 15.1 18.0
4 ’14-’15 Westbrook 95.7 41.1 10.6 12.5 17.6
5 ’04-’05 Garnett 89.1 31.4 19.1 8.0 16.9
34 ’61-’62 Robertson 124.9 26.7 10.8 9.9 14.2
The best pace-adjusted triple-double seasons

Versatility index is the geometric mean of points, rebounds and assists (per 100 possessions).

Source: basketball-reference.com

Nevertheless, clucking about the righteousness of one MVP candidate over another inevitably returns to an epistemological debate about “value.” And there are a variety of cases to be made for players who had less outstanding, but perhaps more “valuable,” seasons than Westbrook did. James Harden moved to point guard and turned in a season that was two parts Steve Nash, one part Corey Maggette, and his Houston Rockets have faint yet plausible finals hopes. LeBron James had the best statistical season of his career at age 32, in his 14th year in the league. And Kawhi Leonard squeezed 61 wins out of a depleted San Antonio Spurs roster on which Dewayne Dedmon has a reasonable claim to being the second-best player. Each of those players’ teams has a far greater chance than Westbrook’s Oklahoma City Thunder of making the finals and winning a championship.

But what if a player is uniquely valuable when the stakes are highest? We’ve seen an example of this before: LeBron James during the 2015 Finals. During that series, James took two games against the ascendent Warriors basically all by himself. James led all players in points, rebounds and assists, and did so while carrying a true shooting percentage of 47.7 and a usage percentage of 39.3. It was a marvelous series for James despite his poor efficiency, in part because his efficiency remained basically in proportion to what’s expected of the most efficient stars despite an altogether absurd workload.

Russell Westbrook has done over 82 games what James did for six. He has scrambled our sense of what game-altering dominance looks like in the age of advanced stats, and he’s done it largely without the benefit of the most important tool of the modern game: reliable 3-point shooting, from himself or his team. Westbrook’s success this season has argued convincingly that top-end efficiency isn’t an absolute requirement for success in today’s NBA, so long as you can make up for a dip in quality with sheer force of quantity.

Efficiency is not a vacuum

Among a certain part of the advanced stats-minded crowd, Westbrook is easy to dismiss as an outdated, high-usage, low-efficiency volume shooter in an NBA that has moved beyond ball hogs. This makes sense if you view the split between Westbrook and high-efficiency players like Harden as a proxy battle in some broader war for the aesthetics of basketball. On one side you have the game played as a brutal, Pat Riley-style combat sport, and on the other the spread-’em-out game of the Rockets or Warriors. But Westbrook is defying the norms of efficiency, too. He’s just doing it in his own way.

Take a look at this chart showing usage percentage and true shooting percentage, which originally ran in an article by my colleague Ben Morris, in which he made the case for Steph Curry as the MVP:

In general, the more possessions a player uses,2 the less efficient his personal offense becomes. You can see the frontier of exceptional player seasons forming a rough diagonal, sloping down from Kevin Durant’s 2016-17 in the upper left to Westbrook’s in the lower right. Generally, that’s the frontier of achievement for maximizing efficiency and usage, and anything that breaks past the outer rim is in the running for the best season in NBA history. Curry’s 2015-16 was more or less unprecedented, but was followed up quickly by Isaiah Thomas and Harden this season, each putting up absurd efficiency numbers with what have traditionally been extremely high usage rates. Then there’s Russell Westbrook.

While a glance at the advanced stats (55.4 true shooting; 41.7 usage percentage) will give you the gist of the relationship — less efficient, more usage — they mask just how far out of the norm Westbrook has been. He has bucked the trend that’s afflicted super-high-usage NBA players for as long as the league has existed: Westbrook’s usage has exploded … and his efficiency hasn’t really changed. As a challenge to the basic makeup of NBA efficiency trends, Westbrook’s season is just as much of an aberration as Curry’s 2015-16.

This is the final entry in our series making the case for five NBA MVP candidates. We’ve also made the cases for James Harden, Kawhi Leonard, Lebron James and Stephen Curry. Also, check out our NBA predictions.

That said, just because Westbrook’s season has been impressive does not necessarily make it useful or valuable. And while it’s undeniably hard to do what Westbrook has done, it’s been an open question how much value there is in his tirelessly filling up the box score while also eating up possessions. Players such as Curry and Harden, who shake up the ratio by adding efficiency to a standard-issue star player workload, are far easier to evaluate. But a player who can take on limitless responsibility with seeming immunity to defensive attention is a dangerous tool in situations where good possessions are hard to come by, even if he isn’t the most efficient guy on the floor. We’ve seen Westbrook deliver in those situations this season.

An unstoppable, moderately efficient force

Westbrook has been laboring under LeBron-esque playoff demands all season long as the late-game anchor for a severely offensively handicapped team.

Westbrook’s crunch-time numbers this season are comical. We’ll define “crunch time” here as the last five minutes of a game (or overtime) in which neither team has a lead greater than five points. In those situations, Westbrook has been unstoppable. His already absurd usage percentage jumps from “just” 41.7 to 62.3. (Sixty-two point three!!!)

Something else interesting happens to Westbrook during crunch time: As his usage goes up, so does his efficiency. His true shooting percentage creeps up to 56.9, and his assist percentage goes to 58.3. While he’s on the court in crunch time, the Thunder is outscoring opponents by 21.7 points per 100 possessions (up from +12.5 overall). In fact, Westbrook’s work rate late in games is so prolific that he produces as much value as entire teams. No, really.

We know this thanks to a little-noticed stat on the NBA’s stats site called Player Impact Estimate, or PIE. PIE is the share of all box score activity in a game (so points, rebounds, assists, steals, blocks) with deductions for negative stats (turnovers, missed shots, personal fouls). The average for a player should be about 103, and the average for a team about 50.4 It’s an especially useful stat when used in tandem with net rating, because you can then see both how well a player is doing individually (the PIE rating) as well as how well the team is performing overall (the net rating).5

Anyway, Westbrook’s PIE in crunch time is 40.3, meaning he accounts for about 40 percent of both teams’ combined activity all on his own — a greater share of game stats in his clutch minutes than five teams6 collect as a whole. The Thunder as a team has a 61.4 PIE in crunch time, fourth in the league, which tracks more or less with its 19.9 net rating, which ranks second overall.

STATS DURING CRUNCH TIME
PLAYER CRUNCH TIME NET RATING ASSIST % TS % USAGE % PIE
Russell Westbrook 148 min +21.7 58.3% 56.9% 62.3% 40.3
Jimmy Butler 141 +12.8 31.5 63.0 41.3 29.4
Kawhi Leonard 137 +24.0 23.0 55.4 40.4 28.5
Isaiah Thomas 163 +18.2 25.0 65.4 46.0 26.4
DeMar DeRozan 139 +16.1 28.8 54.8 42.9 26.1
Anthony Davis 163 -2.7 8.3 51.7 36.1 24.8
LeBron James 126 +15.1 40.6 67.3 32.1 24.6
DeAndre Jordan 105 +3.2 1.9 68.8 18.7 23.4
Stephen Curry 90 +10.8 27.3 61.8 36.1 23.3
James Harden 133 -3.0 40.0 55.6 51.0 21.5
Top NBA players by involvement during crunch time

Source: NBA.com

A player totally unfettered from the effect of a defense is dangerous all game long, but a particular nightmare late in games.

Team composition matters

OK, so Westbrook can get his whenever he wants to get his. No one really doubts this. But Westbrook’s ability to get his teammates quality shots is a lingering question because Westbrook is not Curry, who distorts the parameters of the game without even touching the ball. Curry’s teammates find better shots and make more of them without Curry ever having to generate a traditional assist. But Westbrook’s teammates … let’s just say not even Curry could charm Andre Roberson into hitting his wide-open 3s.

The Thunder roster is not quite as bereft of talent as it’s sometimes made out to be — Steven Adams is a very good center, and Enes Kanter, Victor Oladipo and a few young players like Domas Sabonis all have their uses. But the team’s players are the worst long-range shooters in the league.

This is made clear when we separate out the team’s doomed long range shots.

2-PT SHOT PERCENTAGE 3-PT SHOT PERCENTAGE
PLAYER OFF PASSES FROM PLAYER OTHER SHOTS DIFF OFF PASSES FROM PLAYER OTHER SHOTS DIFF
Westbrook 58.2% 48.6% +9.6 33.1% 31.4% +1.7
Harden 62.6 52.5 +10.1 34.0 37.6 -3.6
James 55.1 52.3 +2.8 40.9 37.2 +3.7
Leonard 47.2 49.6 -2.4 43.8 38.3 +5.5
Curry 58.5 55.5 +3.0 41.8 35.6 +6.2
How teammates shoot after passes from their stars

Source: NBA.com

When a Westbrook pass leads to a 2-point shot, his teammates are shooting 58.2 percent; when they take 2-pointers not directly following a pass from Westbrook, they shoot 48.6 percent. This is a massive difference, but also a logical one: Players shoot better when their point guard sets them up for shots.

Things go downhill quickly once OKC ventures out beyond the arc. There, Westbrook passes lead to makes on only 33 percent of attempts; without Westbrook passing to them, his teammates make 31.4 percent. Both numbers are staggeringly bad. The Thunder simply don’t have players who can shoot NBA 3-pointers.

PLAYER TEAMMATES’ WIDE-OPEN 3-PT %
Westbrook 30.9%
Harden 38.3
James 42.4
Leonard 41.4
Curry 40.0
Which stars have teammates who can shoot?

Source: NBA.com

As a team, the Thunder were a bit above average at creating wide-open 3s (meaning the nearest defender was six or more feet away). Getting open 3s is good! Except, they shot 32.4 percent on those wide-open looks, good for dead last in the league. Westbrook himself shot 40 percent, which means the rest of the team shot 30.9 percent. Again, on wide-open 3s. No defender within six feet. Thirty point nine percent. A tabletop cactus could shoot 30 percent with the defense out to lunch.

So while Westbrook does not have as profound an effect on his teammates’ shooting as his peers, this is hard to pin on Westbrook himself when he’s holding the bag for an Oklahoma City front office running back the ’93 Knicks.

“Stat hogging” is not a phenomenon

One final line of suspicion about Westbrook’s stat line revolves around the notion that the numbers are inflated by methods unnatural to the game. One of those allegations: that Westbrook’s teammates let him collect rebounds to help stuff the stat sheets.

But it’s not that simple. For one thing, stars have always received preferential treatment on cheap rebounds. There’s an old story about Rockets players getting gassed up when Yao Ming finally began to yap at teammates who tried to scoot in on missed free throw rebounds — generally the easiest to collect — because those are reserved for the star big man, and the NBA runs on hierarchy. And Kevin Love made a habit of grabbing the ball at the ends of quarters, just after the buzzer sounded, and doing a quick turn, point and grin in the direction of the scorer’s table trying to get credit for the board.7

The Thunder also aren’t as blatant about giving Westbook rebounds as they’re made out to be. Yes, it’s conspicuous that Westbrook is pulling in 8.5 uncontested rebounds per game, up from 5.9 a season ago. But we can track how often teammates give up a rebound so that a nearby teammate can pick up the ball: It’s a stat called deferred rebound chances. This season, the number of the Thunder’s deferred chances has decreased to 16.8 per game from a league-high 17.7 a season ago. What’s changed? Well, the 6.6 uncontested rebounds per game Kevin Durant collected in 2015-16 needed to be redistributed somewhere.

Unlike shooting or passing, rebounding suffers from severe diminishing returns. There are only so many rebound opportunities, and only so many bodies needed to corral them. Oklahoma City finished first in overall rebound rate, and third in defensive rebound rate. The Thunder have decided to use this surplus of rebounding to leak extra bodies out into the break, knowing their point guard can collect the rebound and start the break. In other words, the Thunder have made a conscious effort to let Westbrook get the rebounds because they think it helps them win, not just because they wanted Westbrook to hoard triple-doubles. The Thunder are fifth in percent of points scored via transition, so it’s working out for them.

But there’s a downside: This strategy often leads to Westbrook playing abysmal defense as he hunts for the rebound — the number of shots he contests is dismal, and by far the lowest among league leaders in defensive rebounds, though they are more or less in the same range as those contested by Harden and LeBron. But then, Westbrook has never been a great defender, and it makes a certain amount of sense to have him sacrifice already questionable defensive attention in service of the offense, especially since the defense manages just fine (10th in efficiency) without him. That’s not an ideal outcome, of course. In a perfect world, Westbrook would be more engaged defensively, and have teammates with shooting range that extends beyond the college 3-point line. But the Thunder’s willingness to let Russell be Russell is its own sort of progress.


For years now, we’ve been hearing about how evolutionary players such as Kevin Durant or Anthony Davis or even Steph Curry were set to move the NBA game forward. That has largely turned out to be true: Offense in the league has improved at record rates, primarily because players and front offices have maximized efficiency at every opportunity. The league has found a way to squeeze more production out of more specialized players. And that works just fine as a general rule. But Russell Westbrook’s season proves that’s not the only way to remake the NBA in your image. Shaving away minor imperfections in pursuit of the ideal less-for-more ratio isn’t necessary if you come equipped with a never-ending supply of more.

Footnotes

  1. Versatility Index is the geometric mean of points, rebounds and assists per 100 possessions.
  2. Plays on which a player takes a shot, draws a shooting foul or commits a turnover.
  3. Since there are 10 players on the court.
  4. Since each team makes up half the players.
  5. Net rating is just a team’s scoring differential per 100 possessions.
  6. The Pistons, Heat, Lakers, Nets and Suns.
  7. These anecdotes came from Bill Simmons interviews that are lost to history after Grantland was shut down.

Kyle Wagner is a senior editor at FiveThirtyEight.

Comments