Of all the things hockey statistics sites can provide us, finding who played with who on any given power play is actually quite difficult. Most of the teammate effect analysis (sometimes called With or Without You analysis) focuses on even strength play.
Fortunately, there are wonderful people on the internet with the access to such data. Micah McCurdy (known as @IneffectiveMath on Twitter or the person behind the WOWY spider charts) dug through his numbers to compile some raw data about how the Dallas Stars, with a particular focus on John Klingberg and Alex Goligoski, fared in 5 v 4 situations. I did a tiny bit of additional math to break those down into comparable units.
First, using those numbers, let's look at how the team fared with each player on the ice compared to the average as a whole:
That's a lot of data. Let's break it down.
I chose to go with a per two minutes ratio rather than the classic per 60 ratio because those numbers get a little absurd at /60 when you're talking special teams, and /2 is a better way to think about the value of each individual to a special teams unit.
The set of team data encompasses all minutes, and the Klingberg and Goligoski totals are not exclusive (i.e., they account for some portions of time when the other man was on the ice as well - we'll get to how much they overlapped in a bit).
What's a bit surprising here is that Goligoski finishes below the team average in terms of generated shots in all situations, though he closes the gap when it comes to goals for. He also is a low-shots-against player, suggesting perhaps that when he was on the ice, the power play unit with him as more methodical in looking for open shots, preferring the sure thing rather than creating chaos around the net.
That's not necessarily a bad strategy, which can be seen in the fact that his goals for to total shot attempts ratio is very close to the Klingberg and overall team ratio (0.65 to 0.68 to 0.69 respectively), as is the shots on goal to total shot attempts ratio. It's simply a different manner of approach, one that may or may not be best suited for the top unit on this team. And the lower event totals with similar conversion rates does result in a lower overall goals ratio per minute.
It is also very different from the high-volume style that Klingberg seems to favor. Units featuring him clearly had shot production well above the team average of every type and brought the team average up in terms of efficiency (with goals scored at the rate he was on the ice, the Stars would have had 4.7 more power play goals at the end of the season).
The real question I set out to answer, though, is about deployment of Klingberg and Goligoski together versus separately. The Stars spent a significant portion of last season with a four forward look on defense - how effective was that in comparison to combining the forces of their two best offensive defensemen?
McCurdy was awesome enough to provide those numbers as well.
Klingberg and Goligoski only played about 25 percent of their power play minutes together, and their combined forces were significantly less effective than when they were deployed separately. It's not a case of Goligoski bringing Klingberg down either, and Goligoski's production jumps significantly when he's away from Klingberg on the power play as well.
With a relatively limited number of minutes to look at, it's hard to take much from the huge drop in total goals for with them combined, but as the drop is shown in every iteration of shots taken as well, there's likely something to the idea that, for whatever reason, Goligoski and Klingberg don't complement each other with the man advantage.
Why might that be? That's a really tough question to answer. The two player very well together at even strength, after all.
One possibility is that what makes them good at even strength - the ability to maintain strong defensive posture and not let the play turn around on them while creating solid opportunities for Dallas - is not as important in a power play situation. Penalty killing teams are much more likely to clear the puck down the length of the ice than try to create offense in transition, after all. And teams on the PK don't send nearly the numbers forward on the forecheck, taking away some of the advantage of great stretch passing.
Another possibility is that the styles of power play each player wants to run work in opposition to each other. Goligoski seems to slow the power play down looking for a slam-dunk pass while Klingberg's general reads seem to lead to the puck on the net more. The different reads may have led to a less structured power play unit as a whole that was between ideas most of the time. I suspect this is what's responsible for Goligoski's greater success away from Klingberg.
Yet another possibility, and I think the most likely to explain Klingberg's increased production, is that he really benefits from the uber-talented four-forward set the Stars put out there toward the end of the season on the top unit. The superfriends trio of Jamie Benn, Tyler Seguin and Jason Spezza combined with whoever was the flavor of the time at the net front (filled at various times by Ales Hemsky, Cody Eakin, Brett Ritchie and Patrick Eaves, among others) gives a rotating cast of high-talent options that can all operate on Klingberg's shoot-often (or slap pass to an open forward at the side of the net) mindset.
Given all of the above data, it seems pretty clear that a four-forward unit with Klingberg running the point is the most effective option the Stars have for the first power play, at least until something is proven otherwise this season. Goligoski could also theoretically direct a four-forward second unit or be a part of a more traditional 2-3 setup alongside someone like Jason Demers or Jordie Benn.
Of course, that was all last season. Chemistry could be different this year with new key parts like Patrick Sharp and a healed Valeri Nichushkin pushing for power play time on the first or second unit. Hemsky could rediscover his old form. Ritchie is obviously out for a decent period of time.
But based on the data from last season, the Stars at least have a good working knowledge of how to most effectively deploy their two best offensive defensemen on the power play.