A first look at the somewhat controversial Real Plus-Minus

Nov 13, 2017; New Orleans, LA, USA; New Orleans Pelicans forward Anthony Davis (23) talks to center DeMarcus Cousins (0) in the second quarter against the Atlanta Hawks at the Smoothie King Center. Mandatory Credit: Chuck Cook-USA TODAY Sports
Chuck Cook-USA TODAY Sports

Real Plus-Minus is an interesting tool for NBA fans to use. It is somewhat controversial. There are very smart NBA writers I know who love it, some who hate it, and there are all kinds of in-between opinions on these things.

I consider it a tool in the box. It shouldn’t be used as a stand-alone player ranking (nor should any single-number metric), but it is helpful as a part of the conversation. The general concept of it is smart, though. The idea is that a host of things can never be accounted for in their impact, such as screens and defense or effectively challenging shots or preventing them… and so on down the line.

These things never show up in box scores, but they do, in a transitory way, show up on the scoreboard. If a center sets an effective screen and the ballhandler scores because of that screen, two points go on the board, but there’s nothing to reflect the center’s contribution. If he challenges a shot at the rim, and the shooter adjusts his release to avoid the block, missing because of it, those are two points that don’t go up.

The idea is that by looking at plus-minus stats (what happens when a player is on the court compared to when he’s off it) you can get the overall sense of his impact beyond the box scores. But other factors come into play too. Who is a player on the court with? Who is he on the court against?  Adjusted plus-minus stats attempt to adjust for these things. Real Plus-Minus does so by factoring in box score numbers and combing through play logs to make adjustments for lineups.

While the exact formula is a mystery, its results are published at ESPN.com, and for the most part, they are reasonably helpful, when taken as part of the picture.

So, let’s consider some of the early returns and how much they mean.

James Harden, +8.21

James Harden is the NBA leader at 8.21, which shouldn’t be much of a shock considering how well he’s been playing. Over the Rockets’ current six-game winning streak, he’s averaging 35.8 points, 11.5 assists, 5.0 rebounds, and 1.7 steals while shooting 49.2 percent from the field, 44.7 percent from 3 (on 5.7 makes per game), and 82.8 percent from the stripe.

What is not immediately obvious from those numbers is that he has demonstrably improved his defense this year. He’s more consistent and more effective. There aren’t a lot of defensive measurements, and with the NBA tracking having problems, this is the first number which quantifies that improvement. Harden’s Defensive Real Plus-Minus is plus-0.73 this year, which is up from  minus-1.57 last year, minus-0.98 the year before that, and minus-0.16 in 2014-15.

While it’s too soon to call Harden a “two-way player,” he is at least not a liability on that end of the court this year. That combined with his dramatic effect on the offensive end is why he leads the NBA in both RPM and RPM Wins (a figure derived from the combination of RPM and minutes played).

Stephen Curry, +6.56

Based on raw numbers, Curry is a plus-15.6, which would set a new NBA record (though, it should be pointed out this sort of math only goes back about 20 years). Here’s the rub: Kevin Durant’s plus-13.7 isn’t far behind. This brings up one reason why the adjustments to raw plus-minus are necessary. While plus-minus can reflect hidden help to the team, it can also mask hidden help from the team. Curry and Durant both benefit from one another, and from Klay Thompson and Draymond Green.

That’s why Curry at plus-6.56 comes in only at No. 3 in the early RPM returns.

No one doubts the Warriors (at least no one with working brain cells), but the very thing that makes them so great is what also makes analysts take their plus-minus numbers with a grain of salt. Their synergy makes them arguably the greatest team of all-time. It’s also what inflates their plus-minus numbers. RPM helps alleviate that.

DeMarcus Cousins (+7.56) over Anthony Davis (+2.90)

This is fascinating to me. Look at their box score stats comparison:

2017-18 Regular Season
Rk Player Season G MP FG% 3P% 2P% eFG% FT% TRB AST STL BLK TOV PF PTS
1 DeMarcus Cousins 2017-18 14 38.0 .472 .346 .548 .537 .769 14.0 5.8 1.6 1.6 5.4 3.7 28.2
2 Anthony Davis 2017-18 13 37.2 .565 .382 .600 .596 .787 11.9 3.1 1.4 2.2 2.3 1.9 26.2
Provided by Basketball-Reference.com: View Original Table
Generated 11/14/2017.


Cousins scores a couple more points and grabs a couple more rebounds, but Brow has more blocks, far fewer turnovers and scores more efficiently.

Why is Cousins’ impact so much greater? I honestly don’t have an answer to that one. Perhaps it will balance out. Perhaps it’s his better passing numbers. I really can’t tell you, but it is something I found interesting on the first look.

The Superteams

Three teams employ a total of eight of the top 13 players in RPM. See if you can guess who they are without peeking.

You probably guessed Golden State if you have any sense at all, and you would be right. Curry is third, as we’ve already discussed. Durant is No. 7 (plus-3.95), and Green is No. 9 (plus-3.81).

You might have guessed the Houston Rockets, but probably not. Along with Harden, there is Eric Gordon (plus-3.75) and Clint Capela (plus-3.63). Capela has really broken out this year, though it has gone beneath the radar and is manifested here.

The other team you probably won’t guess. The Philadelphia 76ers have Robert Covington (emerging as arguably the best 3-and-D wing in the NBA), who is sixth at plus 4.46, and Joel Embiid, who is eighth at plus-3.88. So yeah, that whole process thing seems to be working out in Philly.

Old Faces in New Places

Another interesting observation is that a lot of old faces aren’t doing so well in their new homes. Look at some of these names and overall rankings.

  • Carmelo Anthony: 197, minus-0.61
  • Jimmy Butler: 208, minus-0.77
  • Jae Crowder: 233, minus-1.02
  • Danilo Gallinari: 273, minus 1.44
  • Chris Paul: 367, minus-2.24
  • Dwyane Wade:  397, minus-2.81
  • Derrick Rose, 415 minus-3.37

Others have done better. Paul George is 17th at plus-3.23. Kyrie Irving is 31st at plus-2.19. But this demonstrates that especially in the early going, integrating a new high-usage player into a lineup isn’t always smooth and easy.

The Teams

All-NBA (based on RPM wins) 

PG: James Harden (3.59)
SG: Eric Gordon (1.87)
SF: LeBron James (2.12)
PF: Al Horford (2.02)
C: DeMarcus Cousins (3.43)

All-Offense (based on ORPM)

PG: James Harden (+7.48)
SG: Eric Gordon (+4.37)
SF: LeBron James (+4.30)
PF: Giannis Antetokounmpo (+2.95)
C:  DeMarcus Cousins (+4.43)

All-Defense (based on DRPM) 

PG: Ben Simmons (+1.41)
SG: Andre Roberson (+1.17)
SF: Joe Ingles (+1.84)
PF: Al Horford (+2.99)
C: DeMarcus Cousins (+3.13)

Not-All-NBA (worst possible starting five based on RPM Wins and 50 minutes) 

PG: De’Aaron Fox (-0.38)
SG: Lance Stephenson (-0.11)
SF: Paul Zipser (-0.06)
PF: Dragan Bender (plus-0.02)
C:  Tristan Thompson (plus-0.03)

Top Five Rookies

1. Jayson Tatum (+1.48)
2. Lauri Markkanen (+1.08)
3. Ben Simmons (+1.02)
4. OG Anunoby, (+0.86)
5. Caleb Swanigan, (+0.19)

That more or less covers the highlights. Just keep in mind that these are early returns and there is still a lot of noise in there.


Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

To Top