You would be hard-pressed to find a goaltender more deserving of the Vezina trophy than Henrik Lundqvist this season. He was in the MVP discussion early in the season. But pulling that off, based on the statistical standards required, would demand one of the greatest single seasons in NHL history. Even the Vezina may be out of touch because of the defensive structure of the current Rangers makeup.
Lundqvist has been incredible this season, but the standards for the Vezina trophy seem to have been set and it requires a goaltender to play 50 games minimum, and lead the league in save percentage. Over the last decade, only one goaltender who met this criterion failed to win the award.
Lundqvist has been great, but his baseline for average is below most of the league that a SV% in the high .920s would require him to dominate the field. He is still among the contenders, but he trails the main pack of Corey Crawford, Cory Schneider and Petr Mrazek, who are all around the .928 mark.
The problem as always remains the perception of what the best goaltender is. There is a major struggle in the hockey community when defining what statistical consistency is and it flows from people in the game to the fans. Alain Vigneault commented earlier in this season, “We need Hank to re-become the Hank that he’s used to, for us and from there, I think everything else should fall into place.”
The danger in this assumption is that it implies that Lundqvist is in control of his environment at all times. It is why the analysis we get when his numbers ebb and flow in small samples go straight to his playing style and his adjustments, as though these are responsible for his failures and his successes. Goaltending does require attention to detail and the work commitment to refining these details to perfection.
It is this repetition that leads to success, but the problem is in how we define this success. Early in the season, Lundqvist was destroying all his previous standards that I had tracked, but that was just because our sample was limited to 20 games. The biggest outlier in his numbers was his save percentage on clear-sighted pucks.
If his numbers had tracked anywhere near his 5,500-shot sample of .952, he would have been in the range where he currently resides. His .971 was inflating his save percentage and with Henrik’s unconventional inside/out style, regressing to his career .952 average, a slump was all but guaranteed.
It wasn’t Henrik going away from things that give him success, it was the law of averages catching up with him. No different than a player on an unsustainable shooting streak.
If we look at the 5,500 shots I tracked before the 2015-16 season, we see that Lundqvist settles into his true ability after about 2,000 shots. He fluctuates between .009 to .012 above the expected total of an average NHL goaltender under the exact same conditions. So when I see Lundqvist play and register a conventional .921 SV%, I make the automatic assumption that the league average for him isn’t .916, but likely somewhere in the .908 to .910 range.
When we start the sample from 0 at the beginning of each season instead of extending it from 5,500, we get erratic results with an extreme peak and a major descent.
This erratic small sample pushes our perceptions into searching for explainable reasons for his successes or failures. And discussions about his depth, structure and Benoit Allaire’s influence become the norm. Then the regression is complete and everything is fine when “Hank is Hank” again.
Goaltenders are slaves to the structure. They don’t just miss shots they can stop and stop shots they cannot in a random pattern. They continually place themselves in the same positions for success and failure. The great ones beat the play and solve the equation more often than their peers. The bad ones are behind the play regularly and sometimes manage to guess the equation, but these small sample runs always lose out to who the goaltender truly is.
The erratic nature of save percentage isn’t the goaltenders erratic behavior, it is the environments erratic nature. When we can separate that erratic nature from the sample, we can get a better read of the goaltender’s ability.
When we combine the large sample with the 2015-16 season sample, we see that Lundqvist hasn’t changed. He has remained consistently brilliant. After the 2,000-shot mark, he essentially maintains the same superstar goaltending in the exact same range for the next 4,000-plus shots. He wasn’t changing his depth or aggressiveness. He wasn’t struggling, the New York Rangers were.
Hank was always Hank.