Much has been made of the Rangers' recent power play struggles, with a parade of forwards moved in and out of the lineup in the hopes of re-invigorating what had been a stagnating group on the man advantage. Sometimes this has been bad (Pirri), and sometimes this has been good (Buchenvich), but assessing the impact of management's tinkering misses the point – was anything wrong to begin with?
The Rangers currently sit at 18th in the league in power play conversion percentage. Historically, the team is around their recent average, with success rates between 15% and 19% each of the last three seasons. This is league-average performance, and while we might want more from the power play, it isn't exactly objectionable. Much of a team's success at 5v4 comes down to luck in small samples, as the constant references to "hot" and "cold" power plays demonstrate. In the long run, most teams convert at a rate of just under 1 in 5 chances, and that is about where the Rangers sit.
However, subject as the power play is to luck and long-term regression towards the mean, there is a significant element of team control. Some teams sustainably outperform the rest of the league, converting at a higher percentage and more importantly generating more goal differential per 60 minutes played. The causal mechanisms underlying this success are numerous, and our understanding of what drives power play success is still limited, but two factors seem to take on precedence. The first is shot rates, where much as they do at even strength, a team's weighted and unweighted shots for are decent predictors of goals-for rates. Second, team structure is crucial, and though our measurements are crude, what we know suggests that good structure can drive power play success independent of good shot rates. In these respects, surprisingly, the Rangers outperform most of the league.
Going forward, the Rangers attempt just over 91 shots per sixty minutes, which places them fourteenth in the NHL. They are slightly better in Fenwick shot attempts (12th, 70.21/60) and 14th again in shots (48.58). They begin to separate themselves, however, in expected-goal terms. At 6.80 per 60, the team's expected goal rate is 7th best in the NHL. Problems with the stat aside, the Rangers also sit 5th overall in scoring-chances for per 60. In non-statistical terms, they generate an average number of shots, but the quality of those shots is above-average.
The team also separates itself from the competition defensively, and whereas the genuinely good underlying attacking numbers have not yet been fully realized in goal totals, the Rangers' defensive play on the man-advantage has been excellent in both expected and actual goals. The team's rate of goals-against per 60 is 6th best in the league at just 0.5, and the success continues across other stats, most importantly expected goals and corsi against, where the team in third overall in both categories. The expected-goals against rate of 0.55, coupled with an expected goals-for rate of 6.80, combines for an expected-goal difference of 6.25 per 60 minutes of power play, one of the best differentials in the league.
Structurally, Matt Cane has developed the Power Play Structure Index, a useful measurement of a team's shape at 5v4. While not without limitations, the measure is a step forward and does its job, predicting goals-for with reasonable effectiveness. For our purposes that, along with certainty that this structure measurement is largely independent of shot generation, is enough to draw some interesting conclusions about the Rangers' power play.
Though Scott Arniel has been a subject of much ire for the fanbase, Cane's measurement suggests the Rangers' structure is mildly above-average, at least this season. The Rangers' score is 16.173 (as of a few weeks ago) in the PPSI, a mark good enough for 12th in the NHL. The number is also a drastic improvement on previous seasons' results, which might tentatively be attributed to the introduction of new players like Mika Zibanejad, who has made a point of camping out above the right circle, and Pavel Buchnevich, who is perfect.
The two separate measurements, shot-rates and structure, give a picture of a Ranger power play quite different than the team's recent struggles may indicate. Before last night's wonderful effort, the power play had been running cold, and no amount of personnel shuffling on the part of Alain Vigneault seemed to make much of a difference. The power play is still, of course, subject to significant degree of luck, but the Rangers' non-goal statistics are indicative of better play than their conversion or goal rates suggest.
So long as the team continues to generate high quality within a passable structure, we should expect better results in the future, other things being equal. That's not to say there isn't room for improvement – structure in particular seems eminently coachable – but the truly dreadful results of late ought not be taken as reflection of the team's actual ability. With more luck, the goals will catch up with the process, and as the Rangers are shooting nearly 1.5% lower than their expectations, the power play could grow into a weapon as the season winds down and the playoffs open.