There's a neat narrative in the NBA today (which we help perpetuate) that LeBron James and Kevin Durant are the best players in the league by a mile, each season a clash between two great legacies in the making. But how true is this, really?
Below are the league's top players, based on three popular, comprehensive stats—the closely related wins shares and wins produced per 48, plus PER—and two metrics—Real Plus/Minus and Expected Point Value Added—that ESPN introduced this year to much fanfare.
- Win shares per 48: 1. Kevin Durant (.295), 2. Chris Paul (.270), 3. LeBron James (.264)
- Wins produced per 48: 1. Paul (.335), 2. Andre Drummond (.331), 3. James (.327), 4. DeAndre Jordan (.323), 5. Durant (.308)
- PER (Hollinger): 1. Durant (29.9), 2. James (29.4), 3. Kevin Love (27.0), 4. Anthony Davis (26.5), 5. Boogie Cousins (26.2), 6. Paul (26.0)
- Real Plus/Minus: 1. James (8.56), 2. Paul (7.28), 3. Andre Iguodala (6.95), 4. Durant (6.53)
- Expected Point Value Added (2012-13): 1. Paul (3.48), 2. Dirk Nowitzki (2.60) 3. Deron Williams (2.52)
These stats are not without their weaknesses, and I wouldn't point to any of them to individually argue for any player. Taken together, though, they tell you two things: 1) If you make a new stat—and that stat involves even a slightly non-traditional definition of what it means to be "good"—there's a decent chance you'll find Chris Paul's name somewhere near the top of the list; 2) Paul deserves to be at least somewhere in the best-player conversation, if only because he makes you wonder what the use is of having the conversation in the first place.
You can pretty readily see why he isn't there. The last decade-and-change of pro basketball has unfolded in the shadow of Michael Jordan and what's come to be known as "hero ball." The best player in basketball, according to the precepts of the hero-ball era, is the player who takes and makes the biggest shots: Jordan dropping 49 and 63 on the Celtics in his first two playoff games, in 1986; LeBron scoring 48 points—including 29 of the Cavs' final 30—in a double-overtime victory over the Pistons in 2007. The best player is surely versatile and unselfish, but what separates him from the rest is his ability to score, to turn his team's offense into an expression of his will. Close your eyes and think of this kind of player. Chances are you're imagining someone who looks a lot more like Michael Jordan than he does Chris Paul.
That's because Paul, a point guard, doesn't really play hero ball. (He can when needed: With the Clippers down 97-86 with seven minutes to play in their playoff opener against the Warriors, Paul scored 10 of Los Angeles's next 16 points to tie the game at 102.) He's never scored more than 43 points in a game—LeBron and Durant have pulled off the feat a combined 46 times—or more than 35 points in a playoff game, which Durant and LeBron have managed 31 times. Paul can't easily be extricated from a team context, since what he does best of all—and perhaps as well as anyone ever to play the position—is initiate an effective interaction with a teammate. Reconciling the value of such broad effectiveness with individual brilliance is something even the most forward-minded basketball heads still struggle with, especially for a player like Paul, who alternates between the two at will.
As our own Tom Scocca wrote a couple years ago, "Basketball wants heroes, not effective interactions." The sport has yet to coalesce around a single, kitchen-sink stat the way baseball has coalesced around WAR, in part because there are as many different ideas about how to weight those interactions as there are analysts to run the regressions. Those five stats I mention above reflect five different attempts, and Paul's presence at the top of four of them suggests that they're telling a story about "best"ness that's at odds with the one basketball has been telling for more than 20 years.