Saturday, February 12, 2011

Quality Starts

Evaluating goaltending is difficult. Wins are a BS stat, GAA relies heavily on how many shots your defense actually gives up, save percentage is strong affected by special teams (especially the essentially-random PK sv%), and even strength save percentage requires a big sample to sort through the goalies. And so on.

A better metric may be quality starts. To get a "quality start," a goalie has to either:

a) Stop an above average number of shots (.912 this season)
b) Allow 2 goals or less with a save percentage greater than .880

Teams on average win about 75% of their goalies' quality starts, and win about 25% without a quality start. If the goalie stops less than 85% of shots against, we'll call it a really bad start.

So how are the Caps and their opponents doing in terms of quality starts?

If more than one goalie plays for a team in a given game, I counted the one that gave up the most goals as the "starter." Any start not "quality" or "really bad" I defined as "normal." Here are the Caps and opponents' goalies to date in quality starts:

"Wins" are in regulation or OT; Shootout wins are listed as "SO" and were not counted in win totals.

We can see that since the win over Ottawa that ended this year's "streak," the Caps' goaltending has been extremely solid and consistent. Unfortunately, opponents' goaltending has improved as well.

What about the Caps' goalies individually?

I defined a "wasted" start as a quality start that ends without a win. A "saved" start is a non-quality start that ends with a win. Again, shootout wins were not included.

There is a clear 1-2-3 here. Semyon Varlamov should be getting some Vezina votes in a couple of months, having posted almost as many quality starts as Michal Neuvirth, but in six fewer games. Neuvirth seems to be on the cusp--turn a few of those "normal" starts into quality starts and he'll be a legitimate #1. Meanwhile, Holtby is looking like Steve Mason--either quality or not, with little in between.

Frankly, I'll have to say I'll be surprised if this lasts. Before the Caps' goaltending regresses to the mean, BB better have his boys playing better hockey. We cannot count on such reliable netminding come April (although we probably can count on winning a higher percentage of those quality starts).

Caps MVP: Varlavirthby.

Update: With a 4-1 loss to LA now in the books, Varlamov's quality start% will have dipped a little. He had a "normal" start, 4 GA and a .857 save percentage.

No comments:

Post a Comment