A Comparison of Two Defensemen Against Star Forwards

March 5th, 2015 at 6:00 am by Derek

One of my favorite things that Tyler Dellow did on his site before he was hired away by the Oilers was compare players’ results while they were on the ice against the opposition’s star players. Possession statistics are riddled with usage effects, so it’s helpful to control for part of that usage to more effectively see how two or more players stack up against each other. In the case of top defensemen, we especially care about their play against elite competition. If a guy doesn’t generate positive results against the other team’s best forwards, then what’s the point of having him be your shutdown guy?

Today’s case study involves two defensemen with whom our readers should be pretty familiar. I first made a list of one star forward for just about every team in the NHL (the Pens get two for obvious reasons). My sample consisted of Pacioretty, Stamkos, Datsyuk, Tavares, Crosby, Malkin, Ovechkin, Bergeron, Spezza, Johansen, Eric Staal, Giroux, Ribeiro, Backes, Toews, Getzlaf, Daniel Sedin, Landeskog, Parise, Kopitar, Joe Thornton, Seguin, Hall, Little, Vermette, Kessel, Hudler, and Bjugstad. Those are the players that you’d want your best defenseman lining up against if you can effectively control your matchups.

My analysis was simple: see how much 5v5 time on ice my two defensemen spent against these players in the last three seasons and what their goals-for and Corsi-for percentages were during these matchups. All of my data came from the Hockey Analysis player pages. Let’s see how the two compare:

ABcompare

Player A has certainly been used in more of a shutdown role than player B, getting nearly twice as much ice time as his counterpart against star forwards over the last three seasons. Their goal ratios are an equal 50%, but these samples are still too small, especially Player B’s, to make any meaningful conclusions with goals%. The disparity in their possession numbers is quite noticeable, though: a nearly six-point difference in Corsi%! A 47.5% Corsi% for Player A is…not good, even considering the competition. Player B’s possession is simply spectacular; 53.1% is top-pairing material for overall Corsi%, and Player B got those results against the best forwards in the league.

If this hasn’t already become obvious, Player A is Dan Girardi, to whom the Rangers committed six years and $33M before last season’s trade deadline, and Player B is Anton Stralman, whom the Rangers let walk away for nothing on July 1st so that they could keep Girardi and sign Dan Boyle. To be fair to Girardi, there might be some score effects in play that hurt his numbers moreso than Stralman’s. Girardi gets a bit more time on ice when his team is holding a one-goal lead (37.35% of the total available TOI, per Puckalytics) than Stralman does (33.03%), so that would bring Girardi’s Corsi% down a bit. I’d also be willing to bet that Stralman’s zone starts are more favorable, though the exact data isn’t easily accessible.

However, these differences in usage wouldn’t nearly account for a 5.6% difference in Corsi%. Stralman has thrived against the league’s best forwards in the last three seasons while Girardi has struggled. And yet, which one is lauded as the prototypical shutdown defenseman? This is one of those cases where the numbers can show where our eyes might be deceiving us. Dan Girardi looks the part of your throwback, stay-at-home defenseman, but when he’s up against the best of the best, he gets dominated. Conversely, Anton Stralman doesn’t much resemble the traditional view of a shutdown defenseman, but his results show us that he excels in that role.

Stralman isn’t coming back, but it’s obvious that the Rangers need to replace Dan Girardi on the first pairing. Neither of their other right-handed defensemen, Dan Boyle and Kevin Klein, seem like very palatable options. Assistant coach Ulf Samuelsson has apparently tossed around the idea of putting Ryan McDonagh on the right side with Keith Yandle. The latter seems like he’s an adventure defensively, but at this point I’m willing to give anything a shot. This Rangers team has the potential to make another run at the Stanley Cup, but giving Dan Girardi 23+ minutes a game against top competition will only serve to hurt them. I hope that in the remaining weeks of the regular season, the coaching staff tinkers with the top pair in order to find a combination that actually works.

Note: if you’d like to see the full dataset with Girardi and Stralman’s numbers against each individual forward, I’ve put it on a Google spreadsheet.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>