Plus-minus is the most polarizing statistic in hockey. There’s no quicker way to draw battle lines in the hockey community than by starting a dialogue about plus-minus. Some still look to it as a way to measure a player’s defensive value, while others see it as the most useless number in the box score.
And it’s not just fans and writers who are dismissive of it. Plenty of players have gone on record over the years to voice their disdain for it. When asked about it this past March, Boston Bruins defenceman Torey Krug said that he personally hates it and finds it to be misleading.
While plus-minus has certainly come under fire in recent years with the advent of more sophisticated metrics, it has a much longer, complicated history that includes criticisms about its usefulness dating back to its inception.
Today, many argue that plus-minus is meaningless because it depends on far too many variables. These include quality of linemates, team systems, deployment, personal shooting percentage, team shooting percentage, team save percentage and sheer luck.
“[It is] the most useless statistic ever devised,” says Brian Burke, President of Hockey Operations for the Calgary Flames. “It’s pretty simple. If your team stinks, so does your plus-minus. And it’s compounded if you play against the opposition’s top players.”
So, with that in mind, is it time to move on from the stat? Chances are your physician no longer treats your ailments with leeches—and that’s a good thing. In a similar vein, maybe we need to recognize plus-minus’s limitations once and for all, and abandon using it as a way to diagnose player deficiencies. However, that might be easier said than done.
Plus-minus’s history with the league dates back to the early 1960s, when some coaches and GMs were using it internally. Hockey Hall of Fame coach Scotty Bowman first started using it stat while he was behind the bench for the Montreal Junior Canadiens in the Ontario Hockey Association in the mid-1960s. Bowman believes the impetus for picking it up came from Canadiens coach Toe Blake, who was using it with the big club.
In those early years, plus-minus was used to evaluate a player’s offensive and defensive contributions to his team at even strength. According to the NHL, it is calculated as follows: “A player is awarded a ‘plus’ each time he is on the ice when his club scores an even-strength or shorthanded goal. He receives a ‘minus’ if he is on the ice for an even-strength or shorthanded goal scored by the opposing club.”
But, according to former league statistician Ron Andrews, the NHL did not start tracking the stat until 1963, with online data available only as far back as the 1967–68 season. And even when the league did start collecting the information, GMs still believed it should be confidential and neither distributed it to the media nor featured it in the league’s weekly Statistic Sheets. It’s unclear why it was finally released at the end of the regular season in 1975 along with the standard boxcar stats, but when it was the media was quick to point out its dubiousness.
In the August 1975 edition of the Hockey News, Charles Halpin wrote, “Now that the statistics are out in the open it seems they may not be as revealing as everyone had been led to believe over the years.”
Halpin was the first to argue that the numbers were probably not an accurate assessment of a team or a player, noting that “the top teams have the best plus players and the worst teams have the most minus performances.”
Moreover, given the simplicity of the numbers, he questioned why the league had been so hesitant to formally divulge the figures. In a parting shot, Halpin wrote, “Hockey fans have been wondering about the game’s plus and minus figures for a long time. After reading them for the first time in the Hockey News they may go on wondering for many more years.”
Although the limitations of plus-minus were evident early on, players still derived some benefit from it. Many had bonus clauses structured into their contracts for a high plus rating, and similar clauses still exist today. As part of the league’s collective bargaining agreement, players on entry-level contracts are eligible to receive Schedule A bonuses, which include ice time, goals, assists, points, points per game and, of course, plus-minus. This past season, according to General Fanager, one of the players who was eligible for the plus-minus performance bonus was Artemi Panarin. While this stat might seem odd for a club as seemingly forward thinking as Chicago, an executive within the organization said it came down to practicality. In Panarin’s case, since he was such a high-profile signing, they opted to include all of the Schedule A bonuses rather than trying to eliminate certain categories.
Beyond the financial incentives, there was hardware to be won for having a good plus-minus. In 1983, the NHL began handing out the Emery Edge Award to the league leader in the stat (min. 60 games played). Again, hockey writers were dubious. In 1986, Globe and Mail columnist Neil Campbell wrote, “Only a marketing expert could come up with a trophy like the Emery Edge Award. A real hockey man would recognize it as what it is—an individual award that arises from a team accomplishment.”
While later iterations of the award included sponsorships from the likes of Alka-Seltzer and Budweiser, the NHL has since discontinued the trophy. Pavel Datsyuk earned the final Bud Light Plus-Minus Award for his +41 in the 2007–08 season.
Meanwhile, players on the other end of the spectrum were adversely affected by a poor plus-minus. Playing for the cellar-dwelling Edmonton Oilers in 2009–10, Patrick O’Sullivan had the unfortunate distinction of the league’s worst plus-minus. When he got married that off-season, he recalled how his groomsmen presented him with a green jacket on his wedding day to commemorate it. While his -35 on the ice left much to be desired, it certainly would have been an incredible golf score.
He’s able to laugh about it now, but he still believes the distinction hurt his standing in the league.
“It was one of the things that really hurt me for the rest of my career,” O’Sullivan says.
It didn’t matter that amongst players who played a minimum of 500 minutes that season, O’Sullivan produced 1.38 points per 60 minutes (even-strength), which ranked him 237 out of 552. It was the perception that he was a defensive liability that stuck. When he was traded that following off-season, one member of the Edmonton media, still fixated on O’Sullivan’s plus-minus, “seriously wonder[ed] how much of an NHL game he ever had.”
Other more high-profile players have also seen how their plus-minus can be used as an entry point to poke holes in their game. In the 2013–14 season Alex Ovechkin scored 51 goals, but the focus was on his team worst -35, despite the fact that the Capitals finished the year with a -4 goal differential. This wasn’t the first time that criticisms of Ovechkin’s overall game surfaced, but his glaringly bad plus-minus rating provided the fodder for the misguided narratives that were weaved about him that season.
George McPhee, the Capitals’ GM at the time, believed that the criticism was unfair.
“[It’s] a great example of how misleading [plus-minus] could be,” said McPhee, who has told reporters he never used the stat in his roster decisions. “The very next season when the team was coached much better, his numbers were a whole lot better.”
McPhee’s assessment holds water—in 2014–15, Ovechkin put up nearly identical goal and point totals, but his plus-minus returned to the right side of zero at +10. Meanwhile, the Capitals were giving up 4.6 fewer shots per game and had a much-improved goal differential at +38. While some may still fault Ovechkin’s play in 2013–14, it’s hard to overlook the obvious connection between his negative plus-minus and his team’s overall poor play that year.
While GMs like McPhee may not have brought the statistic into their own decision-making process, plus-minus does creep into arbitration hearings. Though it’s becoming less prevalent all the time. Although the former Capitals GM didn’t handle a lot of arbitrations during his tenure in Washington, the team’s current director of legal affairs, Don Fishman, has. As someone who usually takes the lead for these types of negotiations, Fishman has noted how advanced statistics—notably offensive and defensive zone starts—are working their way into arbitration briefs as parties put less and less weight on plus-minus.
According to Fishman, this is largely because it’s an “easily rebutted stat.” As far back as 1998 and in more recent rulings as well, arbitrators have recognized that plus-minus must be used “cautiously” and that comparisons are “tricky” because it is so dependent upon the strength and style of the particular team.
“[Plus-minus] really was the early indicator of Corsi and shot attempts per 60. The problem with it is that it’s just too narrow,” Fishman said.
Even Bowman, an early adopter of the stat and the current VP of hockey operations for the Blackhawks, recognizes its limitations.
“You cannot look at plus-minus as black and white,” said Bowman, who with his nine Stanley Cups as coach and five as GM over five decades has as good a vantage point on hockey history as anyone alive. “It’s a tool, just like any other… there’s a few inaccuracies about it and I think analytics is much more sophisticated now.”
It’s clear that players, general managers, arbitrators and coaches don’t think much of the stat. But how about those on the forefront of “advanced stats”? Jennifer Lute Costella, co-founder of LCG Analytics, says, “It has the sophistication of a kindergartner self-portrait.”
For her, there is simply too much variability in the stat, and while it takes into account who is on the ice when a goal is scored, it doesn’t indicate whether a player contributed in any manner at all.
“[It] just totally misses the mark,” says Lute Costella. “Because we have so many statistics, particularly shot metrics, that we can put a lot of context with, there are just so many better options for it that are easy to use and easily accessible that I don’t really see why it’s being used anymore.”
Is plus-minus something that we can see stricken from the records in our lifetime? It’s tough to say, but amongst executives and players it already seems to be happening organically. The minutes from a Competition Committee meeting in June 2009 reveal that plus-minus was a topic of conversation. After Tampa Bay Lightning player Jeff Halpern criticized the stat as arbitrary, Nashville Predators GM David Poile agreed and added that it had a negative impact on player psyche. Lou Lamoriello, then New Jersey Devils GM, took it a step further and asked if it would be possible to remove it from the league’s statistic package. In response, Commissioner Gary Bettman noted that the issues surrounding plus-minus were on the league’s radar, and that revising the minus on empty-net goals would be considered, short of removing it altogether.
Given the stat’s longevity, one would have expected to find glowing endorsements about this new method of evaluating individual performance when digging through the early records on its origins—but that has not been the case. It is just too one dimensional and unreliable. As most critics point out, plus-minus usually reveals the state of the team linked to that player. It’s no coincidence that 70 percent of the players who have had the league’s worst plus-minus since 1967–68 belonged to one of the two worst teams in that year.
And so it appears we’ve reached an impasse with the statistic. No longer relevant in an era where more sophisticated metrics and evaluations are readily available, plus-minus continues to die a slow death. And George McPhee, as one of many, hopes to “see it go the way of sword fights and dragons—if it hasn’t already.”