New York Islanders Passing Stats through 9 Games of 2014-15

"I pass to you, you pass to me. Is great relationship." - Anthony Gruppuso-USA TODAY Sports

Note: None of this could have been done without aid from Ryan Stimson. His template and guidance were crucial in my effort, and he is the only reason you are reading this. He is finally getting some of the recognition he deserves at the War On Ice Conference in Pittsburgh next weekend, so look out for news coming from the conference.

Overview :

If you haven't had a chance to read Ryan Stimson's work over at In Lou We Trust, you may be unfamiliar with passing stats. Essentially, these stats are concerned with 5v5 passes that generated Corsi events, e.g. missed shots, shots on goal, and deflected shots. These were dubbed by Ryan as "Shot Attempt Generated" or SAG. If they make the net, they are "Shots Generated" or SG. Note that these are included in SAG.

Much of Ryan's work surrounds the metric called SAGE, or "Shot Attempt Generation Efficiency." Essentially, this is calculated by taking a player's (or team's) SG and dividing by SAG. This is used to determine efficiency in generating shots on goal from passes that resulted in a Corsi event. For example, if Travis Hamonic were to make two passes in a game, one that generated a shot on net and one that generated a missed shot, his SAGE would be 50%. Over a larger sample size this can be used to determine how hard a team has to work to generate shots. This metric was highly predictive last year at predicting the outcome of particular games, with a correlation of 0.817. (See: Ryan's article in the Washington Post.)

In future posts, I will go more into SAGE in relation to Islander players and the team as a whole. With a population of only nine games, however, this wouldn't be very predictive for the rest of the season. On with the charts!

Isles Passing Summary

Okay so there's only one chart. However, I think it already tells an interesting story after only nine games. To keep things simple, you may want to look at only A2 SG/G and SG/G. These are potential assists: passes or secondary passes that led to a shot on goal. Keep in mind these numbers were weighted by average TOI, so forwards can be compared to one another, and same for defensemen. At first glance, these numbers may be lower than you expected, but our definition of a pass is stricter than the NHL's. Redirected passes and rebounds are not considered passes in this project, so the numbers take a bit of a hit. Still it's helpful to look at them in relation to one another. So what is this telling us?

The Gang Gets Defensive
It's telling us that Lubomir Visnovsky is far and away the best Islanders d-man at generating offense. But you already knew this, didn't you? Lubo is over 20% more involved in transition than any other d-man by any metric, and I'd say based on his long history in the NHL, this is no noise. The order of the rest of the defensemen by involvement in offense should be unsurprising. Leddy is next, followed by Boychuk, Hickey, and de Haan. There is a large drop off then to Strait and Hamonic, the latter of which may be surprising. Hamonic has been hot and cold in his career for EV assists (his high water-mark being 16, and low point being 4 in the shortened 2012-13 season), but in any event I doubt he continues his lowly pace right now.

The Gang Gets Offensive
Through nine games, Kyle Okposo (1.9) and Mikhail Grabovski (1.8) are the best passers on the team by a great deal, only serving as further evidence that John Tavares has lost a step (1.4). Relax, I joke because nine games is a small sample size and I don't expect JT to lag behind Okposo and Grabovski in potential assists by more than 30%. He should expect to see a boost in his numbers, while the others stay around the same pace. Putting these three together on one line may have explosive results, but I personally believe in maintaining more of a balance in all four lines.

Now onto the hot topic of Cory Conacher. The data shows Conacher to be the fourth worst forward at generating shots on goal per game (0.9), so the eye test here did not lie. This directly conflicts with his elite A/60 rate the past two seasons you've been hearing so much about. When I found out who first started using that in Cory's defense, I'll make sure to give him/her a piece of my mind. In reality, if we had this kind of information over a longer period of time it would be easier to determine which was the real fluke. In any case, the Cory as LW1 experiment should probably be ended, as Tavares's numbers are being adversely affected.

Other findings of note include Ryan Strome's potential assist rate (1.4) being comparable to JT's, which is a promising sign. Right behind him is Nelson, followed by Bailey and Kulemin. In last place is, of course, Matt Martin (0.69), who I was surprised to see that low. From the eye test, I thought Martin took a step forward from last year but I guess it was just wishful thinking. Still, if Nielsen (0.74) is that low as well, I guess anyone has a chance to rebound.

Closing thoughts
I thank you for reading my first piece in what I hope to be a recurring series throughout the year. As my understanding of passing stats grows, so will the length and complexity of these posts, so hopefully we can all learn together. Again, none of this would be possible without Ryan Stimson, so be sure to send him a shoutout in the comments or on Twitter if you enjoyed this.

<em>Submitted FanPosts do not necessarily reflect the views of this blog or SB Nation. If you're reading this statement, you pass the fine print legalese test. Four stars for you.</em>