Welcome back to our series on Red Wings season grades for the 2019-20 season. Today we will be taking a look at the goalies.
My grades are going to be based on expectations coming into the season. Detroit looked to be expecting a scenario in which Jimmy Howard and Jonathan Bernier shared the net somewhat evenly. Neither goalie was expected to be among the top NHL goalies, but around league average was a reasonable expectation.
Bernier had played 35 games the season before, and it seemed likely he would play around the same number, if not a little more with Howard another year older. Nobody expected Detroit to be good this season, but nobody expected the team to be as bad as they were.
Some of you may have tried to push this out of your mind, but unfortunately my job here is to remind you like a tattoo on Leonard’s body in Memento. Jimmy Howard was bad.
It gives me no pleasure to write this because I’ve always liked Howard and thought he has been underrated for much of his career.
I’m not going to belabor the point, but I want to illustrate just how bad he was this year, as well as taking a quick look at the potential for him to bounce back next season (as some commenters here have been hoping for). I’ll also show you how Bernier, despite his poor start to the season, was one of the few Red Wings who exceeded my expectations this season.
The Charts (Only Two!)
Even in 2020, people still use stats like Goals Against Average (GAA) and unadjusted Save % (Sv%) to evaluate goalies. Just like with plus-minus in skaters, the problem with that is that there are much better ways to evaluate the same things.
I’m going to focus on one statistic today, although for people who are interested, I want to give you some other options as well as where to find them. At Natural Stat Trick, you can look at a stat called Goals Saved Above Average (GSAA). This is pretty self-explanatory; it measures how many goals the goalie prevented above what a league average goalie would be expected to save.
Evolving-Hockey has this stat too as well as one that I think is a little better called Goals Saved Above Expected (GSAx). You can also look at delta Save Percentage, which is the difference in actual save percentage and expected save percentage.
At the end of this article, I’ll put a brief appendix that has the formulas for these metrics. They are actually pretty simple and easy to understand.
Today I am using Standing Points Above Replacement (SPAR). It is a version of GAR/WAR, but it makes more sense for hockey than a WAR number because some games award 3 standing points instead of 2. As with any WAR based stat, it measure the amount of something (in this case standing points) a player provides their team with compared to a replacement level player (7th defenseman or 13th forward)
First, let’s take a look at SPAR per 60 for not only this season, but every season since 2007-08, the beginning of when this data is able to be collected. The 500 unblocked shots (FA) minimum gives you roughly 60 goalies per season, which will prevent outliers from goalies who only played a few games. This gives us 731 goalie seasons.
Jimmy Howard was above 0 every season except this year, when he had the 17th worst goalie season in SPAR.60 since 2007-08. Bernier was in the top half of the league, even playing behind Detroit’s defense. This also pokes a giant hole in the idea that Howard was so bad this season because he was playing behind a weak defense. (Also, metrics like SPAR take into account the quality of the shots against)
I used a quasirandom plot to spread the points out so you can more easily see the distribution. That’s why the dots are sometimes a little next to the line.
Jimmy Howard also had the 9th worst season-to-season drop-off since 2007-08 as you can see in the second chart (Ignore that first red dot in 10-11), while Bernier provided more value than 2018-19.
Goalies are volatile, but for Howard to have that extreme of a drop-off at age 35, while the other goalie improved playing behind the same bad defense, tells me that it’s likely his NHL career is over. It would take one of the 36 highest increases season to season for him to just get back to replacement level. Some team may take a flyer on him as a backup, but it shouldn’t be Detroit.
Jimmy Howard: F
This one is unfortunately easy. He had one of the worst seasons in recent NHL history, even accounting for Detroit’s poor defense.
Jonathan Bernier: A-
I really struggled between a high B and low A here. My expectations for Bernier were for him to be a decent backup goalie who could take over if needed as a starting goalie and not hurt the team too much. He obviously exceeded those expectations.
He had a horrendous start to the season, through December 8, he had the 5th worst GSAA (-7.91) from Natural Stat Trick. From December 9th on, he was the 6th best in the NHL with 9.91.
I won’t argue with anyone who grades him in the B range. I bumped him up to an A- because of how he played the second half of the season. As bad as the team was to watch, without Bernier, it would have been worse, as hard as that is to imagine.
Here are the formulas for the metrics I mentioned in the beginning:
GSAA = (League Average Sv % * Shots Against) - Goals Against.
GSAx = Expected Goals Against - Goals Against
dSV% = Save Percentage - Expected Save Percentage.