Last summer I wrote about why I like Halak and why I thought he needed to stay with the St. Louis Blues and why I thought he was going to be the goaltender for the team this season. Since January is apparently goalie month here at St. Louis Game Time, and because CanesAndBluesFan beat me to writing about why the Blues should NOT trade for Ryan Miller, here is my take on Jaroslav Halak versus Brian Elliott.
The view from the NHL data
Halak's save percentage as of today is 91.2% which has him tied for 22nd place (amongst goalies who have played at least 10 games) with Corey Crawford and Mike Smith. Meanwhile Elliott is tied for 12th place with Eddie Lack and Robert Luongo. Case closed right? Elliott with the higher save percentage is obviously better than Halak and should get the starting role and the Blues should trade Halak. Right? Welllllllll not so fast there my friend. If you take a look at the Shots Against and the Goals Against (is the actual number of goals scored on goalie, it is not their goals against average) from the NHL site you will see that both Elliott's Shots Against and Goals Against is below average. Halak's SA and GA are both slightly above average and yet he still has an above average save percentage. Even though Elliott has a better save percentage than Halak, he also hasn't been seeing as many shots as Halak has.
The view from Behind the Net
Behind the Net's goalie report gives us a better comparative view of goalies since the number you pull from that site can be 5v5 only. So we exclude any power play or penalty killing situations from the numbers (unlike nhl.com's data). Behind the Net also provides the numbers in the context of per 60 minutes of ice time. This also evens out any disparity in the number of games played. In my post last summer, I used a stat that subtracts the goalie's Goals Against/60 from the team's Goals For/60. If we know a team scores X many goals per 60 minutes of ice time, then can we measure how well a goalie keeps his team in the game by subtracting the number of goals per 60 minutes of ice time the goalie gives up? Maybe. At least that is why I attempted to do with the Behind the Net data. As you can see, Elliott still comes out on top of Halak but just barely. So from this view I don't think we can draw any hard conclusions.
The view from Extra Skater
Extra Skater has taken the hockey stats world by storm this year and has been a very valuable resource for all sorts of #fancystats and traditional stats this season. One of the more valuable pieces of data I found are the offensive stats that are included on the player pages for goalies. Unlike other sites, Extra Skater provides the same stats for goalies as they do for any other player. While most of the stats provided are not all that useful for goalie analysis, a few of them do come in handy for our look at Halak versus Elliott.
So far we have seen that even though Elliott has a better save percentage than Halak, Elliott has also faced less shots, and at even strength there is virtually no difference between their ability to help support the Blues goal scoring by keeping the team in the game (so to speak). But the question arose on Twitter the other day about whether or not the Blues score more when Elliott is in net than when Halak. Or in other words, do the Blues support Elliott more than they do Halak. I'm not sure if I like that question the way it is framed. It kind of alludes to some sort of psychological mumbo-jumbo about teams feeling more confident in or willing to play harder for one goalie over another. So I rephrased the question as, "does Elliott play against weaker competition than Halak?"
Each goalies' stats page on Extra Skater has the team's shooting percentage at 5v5. The Blues are shooting 11.9% at 5v5 when Elliott is on the ice where the Blues are only shooting 8.9% when Halak is on the ice. The Blues must hate Halak so much that they refuse to score for him. Right? Right. Fuck that psycho-babble mumbo-jumbo. I took Halak's and Elliott's game logs from Extra Skater and then looked up the point percentage on NHL.com for each of the teams that they played against. Since there were some games in which Halak or Elliott played only a partial game, I only awarded them the point percentage for the number of periods in which they faced the team (using TOI per game I rounded up or down to nearest period). I used this adjusted point percentage as a measure of competition for each goalie.
|Avg Strength of Competition
|Median Strength of Competition
As you can see, according to my quick and dirty strength of competition stat based on the adjusted points percentage of the teams that each goalie has faced on ice, Halak has definitely faced stronger competition than Elliott. This probably explains why the Blues have a better shooting percentage when Elliott is on the ice and why Elliott has a better overall save percentage than Halak.
Elliott might have a better overall save percentage than Halak, but Halak is facing tougher teams than Elliott has so far this season. For all the talk about Halak and Elliott sharing goaltender duties, it is clear that Hitchcock views one as the starter and the other as backup. So why is everyone freaking the fuck out about Halak and trying to suggest that Elliott might be the better starter and Halak should be trade bait? Elliott's success has come in part due to facing weaker teams than Halak. Bottomline: The Blues have two goaltenders with above league average save percentages. Any differences in stats is due to the strength of the teams that they face on ice. So can everyone just sit back and watch the games for a while instead of whining about a goalie controversy that just doesn't exist?
PS: Stick-tap to CrossCheckRaise for pointing me to this tweet that provided me with the direction for this article.
Is Halak getting the same support from the skaters as Elliott does? It doesn't always look that way to me.— Paperwork Ninja (@PaperworkNinja) January 17, 2014