I thought I’d made this point in the summer, but it turns out I hadn’t. Jonathan Bernier, despite having the same overall save percentage than James Reimer coming into this season, had one of the highest quality start rates between 2007 and 2013, per Rob Vollman in the Hockey Abstract:
Reimer had an identical .9150 save percentage but just a 48% quality start rate. That’s brought down a little by all that time he played in the 2011-2012 season with a concussion, and also playing behind a terrible penalty kill between 2011 and 2012.
A quality start is defined as a game where a goalie stopped 91.7% of shots, or stopped 88.5% of shots while allowing two goals or fewer. Before the 2010 season, the threshold for a quality start was 91.3%. It was designed to reliably measure consistency between two goaltenders:
Take two goalies with identical .920 save percentages, for instance, but one of them reliably and steadily stops close to 92% of the shots every single game, giving their team a chance to win every one, while the other goalie frequently throws up either shut-outs or blow-outs. Clearly, the first goalie is much better than the second, despite identical save percentages.
It’s interesting when applying this to Reimer and Bernier. Bernier has come out of the gate waving a hot glove hand and has looked unbeatable through three starts and a relief appearance. His save percentage is .974 through four games. Like Randy Carlyle said, it’s too early to anoint him as the No. 1 at this point and it goes without saying that a hot start is not indicative of a hot finish—check Nikolai Khabibulin’s 2012 season for proof of that—but I found it real interesting that Reimer has never had a hot streak like this one out of Bernier.
Actually, neither has Bernier. Prior to this mini-run, his best four-game streak was in 2012 between February 4 and 28. He actually went just 1-1 in that span, covering two Jonathan Quick relief appearances, but stopped 72 of 75 shots and had a .960 save percentage. Reimer’s best, between April 13 and 20 of 2013 saw him go 3-1, stopping 150 of 157 for a .955 save percentage. If I’m not mistaken, the final game of that run was the game that saw the Leafs clinch a playoff spot.
That aside, just by looking at the game data for the two goalies, I think I may have determined that Bernier is just a bit more likely to go on these streaks. Like the hypothetical two goalies in the problem posed above by Rob, Bernier represents the second goaltender. Identical save percentages, but let’s look at four-game runs throughout their careers:
Those are effectively divvied up into nine “buckets”. If we look at the top three, middle three and bottom three and qualify them as “high quality” “mid quality” and “low quality” runs, we can see how often you can expect either goaltender to go through a run:
Reimer, to me, seems like the more consistent of the two goaltenders, but because Bernier is more volatile, that means that he gets more of the higher quality runs along with the lower quality runs. Reimer is boring. Stick him between the pipes, and 80% of the time he’ll go on a run between .880 and .940.
There has to be something to explain why the two goaltenders had an equal career save percentage coming into the season, yet Bernier got more quality starts. The answer is that there are more “blow ups” on the other end.
So it comes down to which type of goalie you prefer. Part of the reason I imagine Carlyle is hesitant to call Bernier his No. 1 already is that he’s been around the game long enough to know that a hot streak isn’t at all indicative of a goaltender’s overall play. You get the sense that some media really, really want to call Bernier the starter, but this run isn’t going to last.
I didn’t have to tell you that, but man, he’s been utterly unflappable so far. Other than the one goal against Cory Sarich against Colorado, he’s stopped everything he should. Against Ottawa he stopped a lot of pucks he shouldn’t have. He has an even strength save rate of .978 and a penalty kill save rate of .962. Despite the Leafs lagging a bit in even strength offence (2.37 goals per 60 minutes this season compared to 2.72 goals per 60 minutes a year ago) they’re still winning games.
It’s easy to say during a hot streak that you’d rather have the goalie that has the chance of playing more games with a .960 or above, so I’m going to reserve judgment on which goalie I *prefer* going forward. I think it is worth pointing out that the more inconsistent goalie is more prone to the type of streak that Bernier is currently on. Despite how much commentators and broadcasters love “consistency”, inconsistency is pretty awesome, too. You can look at the goaltender below Bernier on the list at the top of the page, Sergey Bobrovsky, who won a Vezina last year after a ridiculous short season, despite not having a notably high overall save percentage.