This Season, Low CA Hasn’t Meant Low GA

When looking at statistics, one reason that possession stats like Corsi are so popular is that they tend to denote trends about a team over time. The more Corsi attempts a team has, the more likely that they are controlling the puck, getting more pucks on net, and as a result, getting more goals. Teams that prevent Corsi attempts against are more likely to have better defenses that deny the opposition time and space in the defensive zone. This should in turn lead to less goals against, boosting the team’s defense and goaltending and giving them a better chance to win night in and night out.

The funny thing about possession statistics, however, is that sometimes teams may have numbers that seem to fly in the face of what those possession numbers may say. This season so far, the New Jersey Devils are one of those teams, and the disparity between Corsi attempts against and actual goals against is not working in the favor of John Hynes’ club.

First, let me clarify. The Devils are still a net negative team in possession. Through 23 games played, they have an overall 5 on 5 Corsi of 49%, good for 20^th in the NHL. Comparatively speaking over the last couple seasons that is quite exceptional for the team. Last season, the Devils were 29^th in the league with a 46.2% Corsi. So they have indeed improved significantly in overall possession from last season, but in the end are still negative and are still in the bottom half of the league when it comes to driving play forward.

Despite this, even through the bad times like last season, the Devils were always in the tops of the league in preventing shot attempts against. Last season the Devils allowed the second fewest Corsi attempts against at 3,175 attempts against. Only 3 teams allowed under 3,250. So even when possession was downright terrible in Newark, it was not because the defense was allowing boatloads of attempts against; in reality, it was the exact opposite. The Devils just could not get anything going on offense.

This season is much the same. The Devils are still an elite team at preventing attempts against. While not 2^nd in the league like last year, they are still a very quality 6^th in the league with only 954 Corsi attempts against. Only 12 of the 30 teams are currently under 1000. Preventing attempts is still the team’s strong point, while generating offensive Corsi attempts is another story. They are getting better at it, but still rank 28^th in the league at 916.

The interesting thing, however, is where these possession numbers actually lead to in terms of goals for and goals against. Possession is a wonderful stat to track, and is now arguably as much of a staple in hockey as OBP is in baseball. Corsi, however, may not always lead to the goal stats that you would expect, and it is not working well for NJ this season. The slight positive is that despite being 28^th in Corsi attempts For at 5 on 5, the Devils are actually 25^th in the league in terms of 5 on 5 Goals For with 36. That is a small but nice jump in the overall rankings. The big drop, however, is in the correlation between Corsi attempts against and Goals Against at 5 on 5. Despite being 6^th in CA, the Devils are only 16^th in GA, allowing 42 goals against at 5 on 5 so far this year. If they were 6^th in that category as well, which would match up with their Corsi against numbers, they would have only allowed 34 goals. That is an extra 8 even strength goals that the Devils have allowed over what would be expected given their CA ranking. And considering that the Devils have already lost 10 games this season by one goal, having those 8 goals against back could have meant a world of difference in terms of the Devils’ current record.

There are surely several reasons as to why the Devils are allowing more goals despite their still quality Corsi Against. I think John described very well on Thursday night why the Devils allowed 4 to Chicago. Many times the chances that they are allowing are way too good, and the opposition is simply taking advantage of having excellent looks at the net to drive home some pucks. The defense may be good at preventing a large volume of attempts, but many of the attempts that they do give up are quality ones.

Another reason could be a dip in goalie performance. John discussed in that same recap as to why really only one goal against could be put on Cory Schneider against Chicago, and you could go with the same refrain for the previous several games as well, at least in my opinion. However, there is no denying that his numbers are down this year. According to Hockey Reference, Cory boasts a very mediocre .912 save percentage on the season, despite posting a .924 percentage over his career. Even worse, over his last 5 starts, he did not once post a save percentage over .900. That is shocking to see, and frankly something I would not have thought would be possible. He has been the guy that has kept NJ in games despite poor play. But over the last several games, when the Devils are playing bad, he has not bailed them out. That has hurt.

In the end, things may normalize, and by the end of the season the team’s GA may end up lining up well with its CA. But as of now, New Jersey is giving up goals at a higher rate during 5 on 5 play than what you might expect given the way that they still manage to suppress shot attempts. And it has hurt the team, plain and simple. Thankfully the Devils have produced enough offense that the team has gotten three points in its last four games despite not winning any of them, but 3 points out of 8 is not good enough, especially considering the moral drain that a 4 game losing streak brings. Hopefully the excellent play at home continues long enough into the season that when this team finally gets some legitimate home stands, they can make up ground.

Your Thoughts

What are your thoughts on the disparity between the team’s Corsi attempts against and their goals against? What do you think is the major reason for this difference? What can the Devils do to rectify this issue? Or, do you think they just need to wait it out and let the numbers regress to the mean? Please leave your comments below, and thanks for reading.