More plus than minus in plus-minus
It’s not perfect, but stat is useful in identifying and explaining value
Whether he realized it, U.S. basketball coach Gregg Popovich dumped kerosene on a longrunning Twitter debate with an offhanded comment Friday night. In the moments after the Americans beat Spain in a FIBA World Cup tune-up, Popovich was asked for his favourite basketball statistic. In typical form, he responded, gruffly, with his least favourite.
“Do they have plus-minus on these like they do in the NBA?” he said while glancing down at a box score. “I don’t even look at that. Never looked at it. Just because you don’t know who was in the game at certain times, or you have a weak lineup, or someone played badly who helped someone else have a negative game. It’s the last thing I look at, if I do.”
For many basketball players and fans, that assessment was a music-to-the-ears rebuke of an analytics staple. “Thank you,” Washington Wizards guard Isaiah Thomas tweeted, appending three exclamation points for emphasis. “That s--- don’t mean anything.”
There are plenty of reasons to skeptically view a player’s plusminus — a measure of the game’s point differential while he is on the court — over a onegame sample size. But to poohpooh plus-minus entirely, especially after it has been carefully fine-tuned by basketball-savvy statisticians, would be a mistake.
Popovich listed a few common causes of noisy results within a single game: A player’s plus-minus can be greatly affected by the play of one of his teammates or one of his opponents, or by the relative strength of a lineup. Garbage time can skew results, as can a brief run of exceptionally strong or poor play. When plus-minus critics scream about the stat’s uselessness, they often turn to memorable outliers. It’s possible for a player to have a sensational game and still finish as a minus: When LeBron James scored 51 points in Game 1 of the 2018 NBA Finals against the Golden State Warriors, he posted a minus-13. Labelling James as a “minus” during such a classic performance feels unrepresentative and unfair.
Once the sample size increases, though, plus-minus can become incredibly useful in identifying and explaining value. Take Stephen Curry’s absurd run from 2014 to 2019. During that time, as the Warriors dominated the NBA, Curry led the league in scoring just once and posted point, rebound and assist numbers that were excellent but not as impressive as those of Russell Westbrook or James Harden.
Meanwhile, Curry ranked first in plus-minus four times in six years, placing second and sixth leaguewide in the other two seasons. By traditional numbers, he was clearly an all-star. By plus-minus standards, he was a basketball god. Turning a blind eye to Curry’s remarkable plus-minus run could lead to all sorts of errors in analysis. If one didn’t truly account for Curry’s impact on winning, it would be easier to paint him as only a shooter, to misapply credit for his individual work to the Warriors as a whole or even to sell short Golden State’s dominance throughout his prime.
Even big fans of plus-minus — including ESPN’s popular Real Plus-Minus, which seeks to adjust for the quality of a player’s teammates and opponents — will admit that it’s not a perfect measure. Some players, such as Klay Thompson, consistently perform worse by the measure than their reputation would suggest. Others, such as Danny Green last season, appear to have a starlike impact despite being merely effective role players.
Yet the five best Real Plus-Minus performers last season finished in the top five in MVP voting: Giannis Antetokounmpo, Curry, Paul George, Harden and Nikola Jokic.
It’s worth noting that Thomas, a one-time MVP candidate who has fallen on hard times because of injury, ranked 453rd of 514 players by Real Plus-Minus last season. One can hardly blame any former star for disliking a metric that views him as a fringe NBA player in large part because of his poor defence.
The hard truth, though, is that the NBA free agency market treated Thomas as harshly as Real Plus-Minus, given that he settled for a one-year minimum contract with a lottery team this summer.
This back-and-forth about plus-minus is healthy, as long as Popovich’s words aren’t misconstrued as an invitation to discount a large, and growing, field of work.
Impartial and curious basketball observers should welcome good information in all of its forms.