clock menu more-arrow no yes mobile

Filed under:

Peeling away the layers of Jeff Petry’s plus/minus

The -30 attached to Jeff Petry makes him appear inept in his own end, but the Habs blue-liner was actually the victim of an extraordinary set of circumstances last season.

Montreal Canadiens v Columbus Blue Jackets Photo by Kirk Irwin/Getty Images


It’s a number that serves as a lightning rod for all of Jeff Petry’s detractors and an annoyance that won’t go away for the defenceman’s supporters. It’s a number that can’t be hand-waved with “oh, it’s just plus/minus” because there is so much separation between Petry’s -30 in 2017-18 and the -11 of the defenceman with the second-worst plus/minus value on the team, Joe Morrow. When one factors in defence partner Karl Alzner’s -7, Petry is shone in an even more negative light.

But there are reasons for Petry’s plus/minus, and these reasons have more to do with the defenceman’s circumstances than his ability.

When we break down plus/minus by game situation, we find that Petry’s outlandish -30 shrinks to a far more normal -14 at five-on-five, a value that is not too far away from the -10 and -11 of Alzner and Jordie Benn, respectively. Petry’s plus/minus is also negatively impacted by how he, especially following the injury to Shea Weber, assumed the vast majority of power-play minutes while Alzner and Benn were almost exclusively penalty-killers. This put Petry in a position where he could never receive a plus, and Alzner and Benn into situations where they could never receive a minus.

The major contributor to Petry’s -30 is the time he spent playing four-on-four and three-on-three. Despite playing only approximately 76 minutes in non-five-on-five even-strength situations (versus approximately 1429 minutes at full strength), Petry was on the ice for a whopping 17 goals against. This is not a phenomenon seen in any other Habs defenceman.

The dynamics of the game certainly change when players are removed from the ice surface, but logically an offensively-minded puck-mover like Petry should be better suited to more space and time, and certainly should be better in 4v4 and 3v3 situations than defensive specialists such as Alzner and Benn. Indeed, every other metric bears this out: Petry demolishes Alzner and Benn in terms of controlling shot attempts, shots and scoring chances.

Possession metrics in non-5v5 even strength situations. CF: shot attempts for; SF: shots for; GF: goals for; SCF: scoring chances for; HDCF: high-danger scoring chances for.

So why the -12 rating in non-5v5 even-strength situations then?

Well, the 2017-18 Canadiens were infamous for being unlucky, and Petry was no exception to this rule. When he was on the ice in 4v4 or 3v3 situations, while the Habs enjoyed an acceptable 8.62 shooting percentage, the Canadiens’ goalies combined for an astonishing 68.52 save percentage. To put that into perspective, there are likely netminders in the league with higher shootout save percentages.

On-ice shooting and save percentages in non-5v5 even strength situations.

To frame it another way, Alzner and Benn combined generated 11 high-danger scoring chances and allowed 15. Six of the 11 resulted in goals for; only one of the 15 resulted in a goal against. Petry put up a better ratio: 19 for and 16 against. Not a single high-danger chance generated with Petry on the ice wound up in the back of the opponent’s net, while four of the 16 opposition chances resulted in a goal against.

High-danger chance and goal generation in non-5v5 even strength situations.

(As an aside here: Victor Mete put up numbers of 15 for and two against, winding up with a 1:1 ratio in terms of goals for and against.)

Petry’s -30 rating was not a product of extraordinarily poor play, but rather a result of some astonishingly bad luck within a very small TOI sample. Before castigating Petry for his 2017-18 season based on this statistic while using the likes of Alzner and Benn as metersticks, keep in mind that Weber’s minus-eight in 26 games — accumulated playing the minutes and roles that Petry assumed with Weber out of the lineup — projected to a 82-game season would have resulted in a -25 value.

An imperfect science? Absolutely. But it puts Petry’s number into better context.

All numbers sourced from Natural Stat Trick