Adjusting Khris Middleton’s Plus Minus


Mar 24, 2015; Milwaukee, WI, USA; Milwaukee Bucks guard Khris Middleton (22) shoots the game wining shot as time expires during the fourth quarter against the Miami Heat at BMO Harris Bradley Center. Milwaukee won 89-88. Mandatory Credit: Jeff Hanisch-USA TODAY Sports

In case you weren’t aware, Khris Middleton is in the company of MVP candidates when it comes to all-in-one stats. This was true even before his Wholly-Deserving-of-the-Game-Ball performance last night by launching the most impossible game-winning shot of the season.

By ESPN’s Real Plus-Minus metric (which I used to beat Vegas in predicting win totals last year, and by which I am currently beating Vegas this year), Khris Middleton is currently 8th, sandwiched between DeMarcus Cousins and Chris Paul. As of this writing, Real Plus-Minus estimates that by putting Middleton on the floor, Milwaukee gains 6.38 points per 100 possessions above an average player; for comparison, Russell Westbrook’s figure is +6.84, and Chris Paul’s +6.61.

Rather than get into a philosophical debate on what Real Plus-Minus can and can’t do, let’s dive into just *why* his RPM is so high.

1) He’s putting up numbers

Combining all his box-score stats, we can see that his Box Plus Minus is 1.6 (third on the team), and his PER 16.1 (fourth on the team).

2) When he is present, his teammates are putting up much better numbers.

The biggest reason for Middleton’s mind-destroyingly high RPM is how much better his team plays when he’s on the floor. The Bucks are outscoring opponents by 6.6 points per 100 possessions with Middleton on the floor, the equivalent of roughly the Dragic/Bledsoe Phoenix lineup and Atlanta’s starting five. When Middleton sits, however, the Bucks drop to below The Failed Rondo Experiment (which is the 2nd-worst five-man unit playing at least 300 minutes this season) at minus-7.9 points per 100. This difference of +14.4 points is higher than the +11.0 jump that occurs when Anthony Davis comes back to save the Pelicans from obliteration.

So Middleton is putting up some numbers by himself, but definitely not enough himself to turn Upsettingly-Bad-Dallas-2.0 into Best-In-The-East Atlanta. Is their much-better offense and greatly-better defense just the result of Kidd’s lineups? Let’s take a hard dive into the land of adjusted plus-minus by manually adjusting his plus-minus.

If you’re familiar with, you’ll note that you can quickly measure how many possessions teammates share with a player and how many they have without a player. By measuring the difference in what percent of the time two players share the floor, we can see how plus-minuses are biased. First, a listing of Middleton’s “With” minus his “Without” floor time (strikethroughs belong to those no longer suiting up for the Bucks).

You can see that Zaza and MCW’s on-court numbers are much more likely to be paired with Middleton. On the flip side, O.J. Majo/Parker/Sanders/Knight have all been more likely to be benched or circumstantially excluded (i.e. TRADED/WAIVED/INJURED) while Middleton is on the floor.

To manually adjust Middleton’s extreme +/-, we subtract his teammates’ total Estimated +/- inflation [1. (With-Without%) times (ON NetRtg – OFF NetRtg)]. That is, how much his teammates’ Net +/- should be scraped away from his own Net +/-.

We can see a couple factors driving his Net +/-, first the Jabari Parker’s lineups were pretty horrific, and that Middleton has gotten much more playing time in his absence. Secondly that Zaza shares a very strong Net +/- and tends to be paired with Middleton.

But then controlling for his teammates’ +/- we can see that Middleton’s Net +/- is still a lofty 8.08. Defensively we know that the Bucks are a force to be reckoned with, and Middleton is part of the reason why.