My first encounter with Jaroslav Halak was during the 2010-2011 playoffs as he shut down the President Trophy winning Washington Capitals in 3 straight games and eliminated the Caps from the 1st round. As a part-time Caps fan (I live in Maryland) Halak became a dirty word. In fact, the name of Halak still causes Caps fans to twitch.
Normally I chortle at the Flyers' endless pursuit of every goalie on the planet. But mention the name "Halak" and I pucker up...— JapersRink (@JapersRink) June 28, 2013
However, when the St. Louis Blues acquired him as part of a deal with the Montreal Canadiens I couldn't have been more thrilled. But, after the shortened season this year, I have been both surprised and dismayed by how some Blues fans have turned on Halak and are eager to have him traded. And according to some anonymous sources (and we all know how reliable those are) apparently the Blues are interested in finding a deal involving Halak. So I wanted to take a look at the three Blues' goaltenders in an effort to find some justification for keeping Halak in the Bluenote. And here are my results.
Goaltending analysis is not easy. There are some fancystats out there such as Rob Vollman's Quality Starts. But with the shortened season and Halak being out for almost a third of the games with an injured groin I felt like that would not be the best measure for comparing the three goalies. I recently read @garrethohl Arctic Ice Hockey's Goaltender Review for the Jets and that gave me some ideas such as taking a look at the distance of shots faced.
This data comes from Greg Sinclair's Super Shot Search. The most goals scored on the Blues in 5v5 play were scored on Jake Allen and those shots came from 10 feet or closer. In fact, Allen had the worst save percentage in the 10 and closer category out of all three goaltenders in all 3 distance categories. That makes me a bit nervous considering that is the distance from which 33% of goals are scored against the Blues. Trading Halak means we are relying upon a backup that is not very reliable from the up close range.
However, this analysis does not (and should not) stand on its own. We need to put it in context of how the team played when each of the goaltenders are on the ice. For that I used the team's Fenwick % for when that goalie is on the ice (Fenwick is a measure of possession by counting all shots for and against a team but excludes blocked shots). I have also had this idea that if a team has a high Fenwick % (meaning they control the puck more than their opponents do during a game) then that puts more weight on a goalie to perform better. The team is limiting the number of shots on net which means those shots are more important to stop; especially if a team has a low Goals For per 60 minutes of 5v5 ice time. The Blues fall just under the median of all teams in the NHL for 5v5 GF/60. What does this mean? It means that with the Blues scoring so few goals per 60 minutes of 5v5 ice time, and the fact that they are a "puck possession" team that limits shots on goals, our goalies need to have a pretty good 5v5 save percentage.
Here is a chart I put together that combines these three concepts. First, we see that Allen and Halak both have above average 5v5 save percentage for NHL goalies this past season. Unfortunately, Brian Elliott is below average. However, if you take a look at the size of the bubble, you will see that the team's Fenwick percentage is the lowest for when Elliott is on the ice. Blues just didn't dominate possession while in front of Elliott. For the third component, I created a +/- type stat by subtracting the Goalie Goals Against per 60 minutes of 5v5 ice time from the Team Goals For per 60 minutes of 5v5 ice time. I am trying to capture this idea of can a goalie keep their team "in the game" based on the number of goals their team scores. The higher the number the better, because it means that the team is scoring goals while the goalie limiting goals against (all 5v5). Or if a team, like the Blues, does not score as often, the goalie is making those few goals count. So the color of the bubbles is based on this Team GF/60 minus Goalie GA/60 stat. This provides additional perspective for our goalies' performance. As you can see, Halak has the best TGF-GGA result. Jake Allen has the worst. But I don't think you can blame Allen too much. He has the best 5v5 save percentage out of the three goalies, and the team's fenwick is right at 50% when he was on the ice. In contrast, Halak has the best TGF-GGA with Elliott close behind. What does this mean? Well, my interpretation is that Halak is a better goaltender for the style of hockey that the Blues play when compared to Elliott or Allen. With the limited number of shots faced, Halak has an above average 5v5 save percentage and has the best TGF-GGA number out of all three goalies (which means with the Blues limited scoring Halak also makes the saves to keep the Blues "in the game").
So this is why I like Halak. Halak is the best Blues' goalie up close to the net. He has above average 5v5 save percentage while facing fewer shots than the other two Blues' goalies, and his TGF/GGA number is the best out of all three goalies. Halak is the right goalie for the Blues style of hockey. Not to mention, he is under 30 years of age with only one year left on his contract. Let's give him one more season. I don't think the Blues will regret it.