Optimizing the Rangers LineupApril 16, 2017, by
It’s playoff time, and that means over-analyzing absolutely everything there is to analyze about hockey. Whether it’s a particular goal, a bad call, or the decision to play Nick Holden and Marc Staal in the twilight of a close game, it all deserves our intense scrutiny because well, we’re talking about a chance at a Stanley Cup here. One thing that’s especially crucial in the playoffs, and thus warranting our obsessive analysis, is chemistry.
A couple of weeks back the excellent Ryan Stimson put up a piece on just that. For those of you who don’t know, Stimson runs the Passing Project, which is an effort to track passes that take place during hockey games in order to better understand the little things that make a big difference over the course of a season. While the Passing Project doesn’t quite have every game tracked, they do have a substantial amount of work done already (almost 900 games) and Ryan’s work is worth your attention. The piece can be found over on hockey-graphs.com, and I highly recommend you read it.
The bottom line is this: using the massive amount of passing data that he and his cohorts have collected so far, Stimson was able to group different players by their playing style – balanced, dependent, playmaker, and shooter. From there, he ran some calculations in order to determine the best mix of player types, in order to figure out how a team should optimize their lineup, using xG% as the indicator of success. xG% is basically an automated/mathematical way to determine the kind of quality of chances a player/team has had (think Steve Valiquette’s “1 in 3 slot line passes go in” kind of thing – that would be .33 xG – except done with the kind of data found at corsica.hockey). You can take a loot at how Stimson determines player type, and what mix of player types leads to the highest xG, below, via screencaps I took (you should definitely check out Stimson’s article on chemistry/player types, in addition to his Tableau profile with all of the graphs and data as well).
So given all of this, I decided it would be fun to take a look at the Rangers’s current lines and whether or not they’re optimally configured. Obviously opinions may differ on what works best and how, but I think Stimson’s data and work can provide an interesting perspective on how exactly to compose an ideal lineup.
First we’ve got the top line of Chris Kreider, Derek Stepan, and Mats Zuccarello. It’s a touch hard to see, but more or less, we’ve got three well rounded players, with the exception of Zucc’s low shot volume. Chris Kreider doesn’t pass quite as much, but we can see he’s exceptional on transition play and build up play. This line is pretty excellently composed.
Next we’ve got Jimmy Vesey, Mika Zibanejad, and Rick Nash. Full disclosure, because Mika Zibanejad didn’t have as much data from his Rangers career so far, I used his data from his time in Ottawa, which should be more complete. Vesey here is the best forward in terms of dangerous shot contributions, with Nash having the best shot volume and total influence. Zibanejad, like Stepan, is a pretty balanced player, helping to anchor this line with his all-around ability.
As we move down the lineup things get a little bit more interesting, because the players aren’t all quite as well rounded, and so the prospect of moving guys around becomes slightly more appealing. Here Michael Grabner is not particularly balanced as a player, although his transition play is predictably exceptional. Hayes and Miller are both more or less well-rounded, except Hayes is better on shot volume, transition play, and dangerous shot contributions, while Miller is superior in terms of primary shot assists, build up play, and total passing.
Lastly we’ve got our fourth line. Here we can see that Oscar Lindberg is the most well rounded player on the line, and while Jesper Fast is pretty good in some areas, he’s lacking in terms of build up play, shot volume, and total influence. Tanner Glass is, well, not a lot of anything. For fun I’ve also included a graph of Tanner Glass compared to Pavel Buchnevich. We can see that Buch isn’t particularly great in terms of shot volume or build up play, but he excels on the transition and in terms of primary shot assists, and is particularly good at creating dangerous scoring chances.
Now I’m not exactly one to tell you what’s right or wrong, but I just think this all offers a different perspective on lineup composition and how to optimize your team’s chances, something that’s especially important in the playoffs when every second counts. Given all this data what changes, if any would you make? For me I think I’d put Buch higher up the lineup, drop Vesey to the third line, and put Grabner down on the fourth line. Maybe you’d do something different – that’s kind of the fun of all of this data, and all of this exciting hockey we have to watch."Optimizing the Rangers Lineup",