Why Score Effects are the key to your fancy stats (Puck Momalytics)
(Jen Lute Costella is our new analytics writer, breaking down the fanciest of stats for you each week. She's a mom. She's writing for Puck Daddy. Hence, Puck Momalytics)
On Monday night there were two games in the NHL, both of which afforded an opportunity to look at how teams try to get back in a game when they are trailing. The Montreal Canadiens were not successful in their bid to comeback against the Edmonton Oilers and lost 3-0. The New York Rangers scored five goals in the third period to hand the Minnesota Wild a tough 5-4 loss.
Over the years, hockey analysts have noticed a common trend in the way teams play with a lead and when they are trailing. It’s called Score Effects.
The idea behind Score Effects: When a team has a lead of two or more goals in the first or second period or even just one goal in the third period, it tends to protect that lead. Often, the leading team will go into a defensive shell and be much less aggressive offensively. The leading team takes fewer risks in order to avoid giving up a goal. The team trailing in these situations tends to be much more aggressive offensively, taking more risks to score and get back in the game.
The Canadiens put on a clinic in what we expect to see from Score Effects in their game against the Oilers. The game was tied at 5 on 5 (5v5) for 11.3 minutes of play. During that period where the score was tied, Edmonton posted 55.56% CF% (Corsi For Percentage) compared to Montreal’s 44.44%. Simply put, the Oilers attempted more shots than Montreal when the game was tied. Once Edmonton started scoring, Montreal threw their offense into overdrive to get back into the game. The Canadiens nearly doubled their shot attempt rate during the 29.9 minutes of play (5v5) that the team was trailing, moving their CF60 (Corsi For Per 60 or all shot attempts rate) from 42.60 at score tied to 78.33. At the same time, the Oilers CF60 dropped from 53.25 at score tied to 42.18 while leading.
A look at the 5v5 shot attempt chart from the analytics site war-on-ice.com puts this change into perspective.
Up until the first goal of the game, the play was back and forth with both teams having some significant power-play time as well. Once Edmonton scored, the team kept its foot on the gas in terms of offense and Montreal began to increase their attack.
After Edmonton’s second goal, Montreal clearly took over the offensive play in the game. The attack did not work to get them back in the game. Of their 39 shot attempts when trailing, 16 were blocked and 6 missed the net. They still managed 17 shots on goal, but Oilers’ goalie Ben Scrivens was on his game and sealed the win by shutting out the Habs. The Canadiens certainly had the right idea for trying to get back into the game, but this was one of those instances in which it just did not work.
In the only other game on Monday evening, the New York Rangers got back in the game and ultimately won, but not by following the usual Score Effects pattern. With the score tied for 16.9 minutes of play, the Minnesota Wild had 55 percent of the shot attempts (CF%) while the Rangers had 45 percent. While trailing in the game, the Rangers CF% actually dropped a bit to 44.44 percent. Their CF60 went from 31.97 at score tied to 39.36 when trailing. This is an increase in the Rangers shot attempt rate of course, but it is nowhere near what we would normally expect to see.
There are a few reasons for this. Last season, the Rangers had a very active offense.
The graph above shows the each team’s deviation from the average CF60 last season when the score was tied and when the team was trailing in a game. The Rangers were 7.61 shot attempts (CF) per 60 higher than the league average when the game was tied. In a situation where one would expect Score Effects to push the team’s offensive numbers skyward, the Rangers were 3.72 shot attempts (CF) per 60 higher than the league average. The actual change in the Rangers CF60 from tied to trailing was only 0.8 shot attempts per 60.
Obviously, teams make changes in how they play from year to year for many reasons, so the Rangers will not necessarily continue on this path; however, the rather low increase in the team’s CF60 during the game against Minnesota fits into the Rangers’ usual pattern. This is despite the fact that the CF60 in this particular game was much lower than the numbers the Rangers usually put up. The reason for that is simple; Minnesota is one of the best shot suppression teams in the league. The defensive systems used by the Wild greatly lower the shot attempt rates of even the best offenses in the league.
When Minnesota’s stifling defensive system is combined with the fact that, at least last season, the Rangers gave up shot attempts at a rate higher than the league average as shown in the graph above, it is not hard to understand why the team struggled to register a decent shot attempt rate when trying to come from behind. Again, last season’s numbers do not necessarily mean the Rangers will continue to play in this manner, but they are helpful to understanding how the team has played by providing a bit of context.
Despite the fact that the Rangers struggled to get much going in terms of shot rates against Minnesota, they still managed to come back and win the game. This is due in large part to the opportunistic and explosive offense they possess. New York’s on ice shooting percentage (on ice sh%) while trailing in the game was nearly 42%. They actually ended up with more shots on goal (12) than Minnesota (6) did during that period as well, so while they may have been out attempted in terms of shots, the Rangers managed to put more of them on net. Wild goalie, Darcy Kuemper, who had seen very few shots on goal in the first two periods, struggled mightily posting a save percentage (Sv%) of right around 58% in the third period.
Despite New York’s problems trying to get their shot rate up in the third period, the scoring chances given up by Minnesota, combined with Kuemper’s lack of sharpness and New York’s dangerously talented forwards, allowed the Rangers to hand the Wild an otherwise unlikely defeat.
Statistics used herein were gathered from war-on-ice.com and nhl.com
***
More From Yahoo Sports
