A very lengthy discussion on the Devils occurred in the morning Fly by yesterday. I was going to respond to it in the comments, but it would have turned into a "too long, didn't' read" type comment. So instead, I wrote a too long didn't read fan post (though I hope some of you are able to take the time to read and respond to what I have written here).
But in all seriousness, there was a lot of good discussion on the Devils, their possession metrics, on Brodeur, Schneider, and why the Devils record doesn't match up with their possession metrics.
First let's talk Brodeur. The discussion got into Brodeur's ability to handle the puck. People started estimating how valuable Brodeur' puck handling is, and I think the guesses got way out of hand. Someone said that Brodeur's puck handling might be worth .5 goals per game!!!!!!
If we go by the standard 6 goals equals one win, that means if Brodeur's puck handling is worth .5 goals per game, if he plays 60 games he adds 5 wins to his team just by being a good puck handler!!! I think we can all agree that .5 goals per game is way too much right? Well let's not just guess, let's actually find out, using real data.
So using Brodeur's data from the 2007-2008 season to the current season, we can calculate how many corsi events for and against the Devils let up per 20 minutes, and do the same thing for the backups. Below is what we see.
|Corsi For per 20||Corsi Against per 20|
So when Brodeur is in net, the Devils let up nearly 1 fewer shot attempt (i.e. corsi event) per 20 minutes of even strength play. However, with Brodeur in net, the Devils also take slightly more than .5 shot attempts against per 20 minutes of even strength ice time.
My initial response to that data is that Broder seems to be showing an ability to prevent shots against. However, I think that part of Brodeur's ability as a puck handler is causing total shot attempts to be lower, both offensively and defensively. Perhaps Brodeur is slowing the game down bit, I'm not sure. However, taking into account shot attempts for and against, we are seeing that Brodeur is giving his team an edge of roughly .42 shot attempts per 20 minutes.
Let's get out of shot attempts and figure out how many goals that equates to.
Using four years of league data, the average Corsi shooting percentage for teams is about 4.23%. In other words, average teams score on 4.23 percent of all their corsi attempts for.
From 2007 to 2013, the average NHL team plays roughly 42 minutes of even strength play in a game. If Brodeur plays 70 games, his puck handling saves the Devils roughly 2.6 goals over the puck handling of his back-up goalies.
That is not a great amount of added value that we can attribute to Brodeur's puck handling. On a contextual note, the strength of opponents that Brodeur played vs. the strength of opponents that the backups played were basically identical. Brodeur's opponents had a OppCF%(opponents Corsi For percentage) of 49.93%, while Brodeur's backups faced opponents that had a 49.77 OppCF%.
So I understand that people see Brodeur play the puck and it looks like he's making a huge impact, but from the data that I've looked at it doesn't seem to be the case.
This now brings us to the goalie controversy between Schneider and Brodeur. When Schneider was traded to the Devils, I thought it finally meant the end of the Brodeur era in New Jersey. The facts of the matter are that Brodeur has really fallen off in the last 4 years or so. From 2010-2013, Brodeur has the 38th lowest even strength save percentage among the 42 goaltenders with 3000 minutes or more. Looking at the top of the list, we see Schneider with the third highest even strength save percentage in the NHL.
Brodeur has been one of the league's worst starters in the past 4 seasons, while Schneider has been one of the best, albeit in a smaller sample size. In 117 games, Schnieder has a fantastic overall save percentage of .925.
So I (wrongly) assumed that the Devils would ride Schneider, who so far in his career has showed that he potentially could be one of the elite goaltenders in the NHL, over Brodeur, who through the harsh work of father time has become in reality a very poor NHL goaltender.
I however was quite wrong. Brodeur has played 24 games to Schneider's 19, even though Brodeur has continued his struggles with a .903 save percentage. Schneider has not been as good so far in the small 19 game sample, but he still has been much better than Brodeur at a .914 save percentage this season.
So in truth, I find it pretty remarkable that the Devils continue to play Brodeur so much. Schneider should be getting a starter's share of the starts, as he is quite possibly an elite NHL goaltender and at worse is far better than Brodeur at this point in Brodeur's career. And as we have just recently discovered, Brodeur's ability as a puck handler is nowhere near large enough to close the gap between him and Schneider.
Some people on the Fly-by pointed out that the Devils are scoring more goals at even strength with Brodeur is in net than when Schneider is in net. But when we look closer at the data, this seems to be completely the result of SH% differences. With Brodeur in net, the Devils are shooting 9.35% at even strength, but with Schneider they are only shooting 5.79%. When we look at shot attempts, the Devils are actually generating more shot attempts for per minute with Schneider in net than with Brodeur.
So now that we have the Brodeur and Schneider talk finished with, we can discuss the Devils. When I pointed out the Devils elite possession metrics, most people responded that the Devils still weren't good. Their record doesn't match up with their Corsi %, but that argument isn't good enough, so I thought I'd take a closer look at what is caused the Devils poor record.
The first and most obvious problem has already been discussed. The Devils are handicapping themselves by playing Brodeur so often. Schneider should be getting as many starts as possible if the Devils want to maximize their point totals.
But there is more to the Devils struggles than just playing Brodeur. One thing that I have been thinking a lot about recently is the percentage of shot attempts that result in shots. I'm not sure if teams can control their ability to get shots through consistently or if the differences we see are more the result of random chance, but some teams have a better or worse Shots for% than their FF% or CF%.
The Devils are one such team. Although the Devils have a 5v5 close Corsi For % of 54.2%, their 5v5 close Shots for % is only 51.3%. Their possession dominance (as measured by Corsi) is right now being overstated a bit because they are missing the net or getting their shots blocked too often.
Again, I'm not sure whether this is a real skill that the Devils are failing at or if it's just random and we should expect the Devils to start getting a greater percentage of their shots on net over the rest of the season.
While the Devils have been a poor shooting team in recent years, this year their even strength shooting percentage is 10th in the NHL at 8.1%, so it's not that they are shooting poorly at even strength. That doesn't explain their struggles at all.
Let's take a look at special teams. The Devils neither take nor draw a ton of penalties, which I bet is partially due to their low-event strategy. Overall, the Devils have scored 23 power play goals and let up 19 short handed goals. However, when taking into account short handed goals, the Devils overall on special teams are a +2 goal differential. That's nothing special, but it also doesn't seem like something that would make the Devils such a poor team.
So right now, I see two big reasons why the Devils stellar Corsi For % hasn't led to a better win-loss record. The first is that playing Brodeur so much has lowered their goaltending performance. This of course could be easily changed going forward. The second is that the Devils shots for % is not nearly as high as their CF%, as a result of the Devils missing the net or getting their shots blocked too much.
However, there is two things we have not yet talked about.
The Devils have not won a single game in the shootout. The Devils are currently 0-6 in shootouts, a skills competition that is basically a coin-flip. We should expect that record to improve over the rest of this season.
Finally, the Devils have played 24 of their 43 games on the road. Because home teams have a distinct advantage (winning about 56% of their games), the Devils have been at a disadvantage playing so much on the road. This however will correct itself as the Devils will play 22 games at home versus 17 games on the road for the remainder of the season.
So combining the impact of Brodeur playing too much, the bad shootout record, the high number of road games played, and the lower shots for %, we can start to see why the Devils win-loss record is only 17-18-8 with such dominant possession metrics.