Canadiens vs. Rangers Game 3 by the Numbers: A complete team effort

As the scene shifted to New York for Game 3, the question was how Montreal would recuperate after the emotional conclusion to Game 2, and how the Rangers would respond after being seconds from taking a 2-0 lead in the series. We got our answers in a dominating and comprehensive Montreal performance which was closer to what we expected coming into the series. Tonight was certainly less stressful than Game 2.

The Canadiens were in control from start to finish.

In Games 1 and 2, the Habs started strong but let the Rangers play their way back into the game in the second period. This time, there was none of that. In the first period of Game 3, Montreal outattempted the Rangers 20-11, outshot them 10-3, and outchanced them 12-6 (high danger 6-4). In the second period, the Canadiens outshot the Rangers 8-6, the chances were 7-7 (high danger 4-0), and the best the Rangers could muster was outattempting the Habs 15-13. Although the scoreboard only read 1-0 after 40 minutes, this was as definitive a one goal lead as one tends to see in the NHL.

Claude Julien has said that defense begets offense, and the Rangers had tremendous difficulty penetrating Montreal’s defensive structure. They were limited to four shots on goal (marked by the blue dots below) from inside the faceoff dots - with only a single one of those from the net-front region. The slot region was kept remarkably clean, with most Ranger attempts coming from distance or poor angles, as evidenced by the following graphic:

Trailing 1-0 and faced with a resolute Montreal defensive structure, the Rangers were forced to take more risks in search of an equalizer - even while shorthanded. This stretched their own defensive structure, and the Canadiens were able to capitalize twice more in the third period on plays featuring forced turnovers and coverage breakdowns.

Alain Vigneault didn’t make many adjustments - and those he made didn’t work.

After being shellacked in Game 2, eyes were on Alain Vigneault to see what adjustments he would make and how he would operate with last change. The Rangers bench boss responded by... swapping Nick Holden for Kevin Klein and moving Brendan Smith alongside Marc Staal. The change didn’t yield the desired effect, as the four posted an average CF% of 39.06 - marking the first game that Skjei and Smith were below 50%.

Despite having last change, Vigneault stuck to the same matchups that failed him in Game 2. The Chris Kreider-Derek Stepan-Mats Zuccarello line played mostly against Andrei Markov-Shea Weber (79-6) and Brendan Gallagher-Tomas Plekanec-Paul Byron (11-14-41) in Game 2, with Jordie Benn and Jeff Petry (8-26) taking the bulk of the remaining TOI. In Game 3, despite having last change, they were predominantly matched up against the exact same quintet.

Vigneault did manage to secure some more favourable matchups for his first line - their second-most common on-ice opponent (~20-25% of total 5v5 TOI) was the Artturi Lehkonen-Alex Galchenyuk-Andrew Shaw line in Game 2, whereas it was Dwight King-Steve Ott-Torrey Mitchell line in Game 3. However, the Montreal fourth line dominated the Rangers first line, posting an average Corsi line of +5/-1 and generating five scoring chances while allowing only one.

The Mika Zibanejad-led second line (left) and Kevin Hayes-led third line (right) were also unable to benefit from having last change. The Hayes line actually received an increased dose of Markov-Weber in Game 3 as Claude Julien shortened his bench to lock down the win, with Markov and Weber playing 25:48 and 29:07 respectively overall.

Looking at the forward matchups, Zibanejad’s trio (left), which had played against Max Pacioretty-Phillip Danault-Alexander Radulov (67-24-47) in Game 2, were instead equally deployed against the Danault and Plekanec lines. This decision may have been intended to draw the Plekanec trio away from the Stepan line, but as we discussed above, that intended outcome did not actually occur. Hayes’ group (right) saw no shifts in deployment, playing equally against the Danault and Galchenyuk lines in both games (27/21-65-62, with King for part of Game 2) and yielding the same result: a CF% in the 30s.

The Plekanec line continues to shut down the Rangers’ biggest guns.

A cursory glance at the Corsi sheet would indicate that the Plekanec line had a rough go of it. The trio averaged a CF% of 34.32%, worst on the team by quite some amount. Worse, they appeared to be outplayed by the Kreider-Stepan-Zuccarello line for the first time in the series, posting an average CF% of 32.64% against the Rangers’ primary weapons.

However, only two of the pucks fired at Carey Price by the Stepan line with the Plekanec line on the ice actually registered as a shot on goal: one midway through the first period, and one midway through the second. In contrast, the Plekanec line fired five shots at Henrik Lundqvist, with all five testing the Rangers netminder. Despite the CF% disparity, the final scoring chances were +3/-4 for this matchup. One problem that the Habs should be aware of prior to Game 4 is that three of the four Ranger scoring chances here were of the high danger variety. Still, this was a more than satisfactory performance by Gallagher, Plekanec, and Byron, and it makes one wonder why Vigneault continues to actively seek this matchup.

RADULOVE.

So, I need to write about numbers, right? How about 47!

Ok, I just needed an excuse to post the gif.

Hitting their stride.

After putting their fanbase through frustration and heart-pounding panic, the Montreal Canadiens are taking form and playing some excellent hockey. At the moment, the Rangers, even with last change, do not appear to have any answers for the Habs’ three well balanced lines and their defensive structure. The emergence of a viable fourth line combination in Game 3 further complicates Alain Vigneault’s strategizing, as he seeks to form a game plan to avoid heading back to Montreal down 3-1.

(Stick tap to Micah Blake McCurdy and Natural Stat Trick for raw data and graphics. For a primer of the metrics discussed in this article, please see Fancy Stat Summer School)