Mile High Sticking
Fansided

Colorado Avalanche Are In No-Win Situation With Varlamov As Their Goalie

By Anthony Noga

TORONTO, ON - DECEMBER 11: Colorado Avalanche Goalie Semyon Varlamov (1) takes a breather after fighting off two penalties during the regular season NHL game between the Colorado Avalanche and Toronto Maple Leafs on December 11, 2016 at Air Canada Centre in Toronto, ON. (Photo by Gerry Angus/Icon Sportswire via Getty Images)
TORONTO, ON - DECEMBER 11: Colorado Avalanche Goalie Semyon Varlamov (1) takes a breather after fighting off two penalties during the regular season NHL game between the Colorado Avalanche and Toronto Maple Leafs on December 11, 2016 at Air Canada Centre in Toronto, ON. (Photo by Gerry Angus/Icon Sportswire via Getty Images) /
facebooktwitterreddit
Prev
2 of 6
Next
colorado avalanche
colorado avalanche /

Methodology

The starting point is Hockey Abstract‘s concepts of quality starts (QS) and really bad starts (RBS). A QS is a start where a goalie’s save percentage (SV%) is at or above the league average (914 this year among goalies with 20+ starts) OR or 885 when facing 20 or fewer shots. An RBS is one where the goalie’s SV% is 850 or lower. Hockeyreference.com keeps tracks of these and related numbers.

Essentially, these stats use SV% to determine how often a goalie gives his team a chance to win. Save percentage isn’t necessarily the best statistic to measure a goalie’s worth, but it’s the best one I know of that is tracked game by game.

QS and RBS are insightful, but can be overly broad. Stopping 17 of 19 shots earns a QS, while stopping 17 of 20 earns an RBS. At the same time, there’s no distinction between a 17/19 night and a 40/40 night. Both are quality starts.

To better grasp the type of starts Varlamov has given the Colorado Avalanche, I modified these concepts into five categories based on SV%, which I labeled as letter grades: A) 930 and above; B) 915 – 929; C) 900 – 914; D) 851 – 899; and F) 850 and under.

It made sense to break up the categories at the current average 914, and 900 and 930 are nice round numbers roughly 2 standard deviations away from this year’s average. Any errors are my own.

I then broke down Varlamov’s tenure with the Colorado Avalanche into these five categories. The results, as expected, indicate that Varlamov will be very good or very bad very often. He won’t be in between for many games.

The only real question, based on the data, is which extreme he’ll be in more often this season.

Home/Editorials