Expect the tide to turn for the Colorado Avalanche

[Elite_video_player id=”8″]

Times are tough in Denver. The Colorado Avalanche are the proud owners of the Central Division’s basement apartment, dust, cobwebs and all, and have earned a measly 2 wins in their last 8 games. Their most recent effort against the Arizona Coyotes drew the ire of the hometown crowd, resulting in them being booed off the Pepsi Center ice.

Already, the usual topics of discussion during lean times have reared their ugly heads. Are the players competing hard enough? Do they even care about winning? Should Matt Duchene be traded? Should Gabriel Landeskog? Why did we pay Tyson Barrie $22 million dollars to turn the puck over all night?

Alas, these poor souls are misguided. As is often the case during a stretch of poor play, most of Colorado’s woes can be explained by one simple statistic: PDO. For those who are unfamiliar, PDO is the sum of a team’s even strength shooting percentage and save percentage and based on the theory that most teams will ultimately regress toward a sum of 100, PDO is often viewed as a proxy for how “lucky” a team is. In Colorado’s case, their league-worst 95.36 PDO is well below their expected PDO of 99.69, suggesting that they’ve been vastly unlucky so far this season.

Let’s have a look at PDO’s two constituents, 5v5 save percentage (Sv%) and 5v5 shooting percentage (Sh%), in an attempt to isolate what exactly has gone wrong for Colorado. 

Save Percentage

Colorado’s team goaltending has left much to be desired this season. Semyon Varlamov has been nothing short of a tire fire, having saved -4.11 goals above average (GSAA) while boasting a Sv% of .877 in just under 320 minutes of play at 5v5. Calvin Pickard has been much better comparatively, saving 1.28 goals above average, while recording a .937 SV% in 198.5 minutes of 5v5 play. Both Varlamov and Pickard have been below average at 4v5, registering -0.96 and -1.44 GSAA respectively. All told, Colorado’s goaltenders have stopped 89.9% of shots at 5v5, good for third worst in the league ahead of just Toronto and Philadelphia.

The good news? Varlamov’s poor play likely won’t persist, at least not to this extent. His career 5v5 Sv% is .925, and given that Varlamov is still just 28 years old, it’s unlikely that this is the beginning of a precipitous and permanent decline in performance. Varlamov’s 18.8 GSAA in all situations since arriving in Colorado ranks 18th in the league over that period of time; he’s not the cream of the netminding crop, but he’s comfortably a starting goaltender.

As an aside, the inevitability of Varlamov’s resurgence is definitely impacted by Pickard’s emergence as a candidate for the starting job. This conundrum in Colorado’s crease is further muddied by the looming threat of the expansion draft. Teams aren’t able to protect more than one goalie, and as a result, the market for goaltenders has cooled considerably, meaning a Varlamov trade would likely net a sub-par return.

Even if Colorado were inclined to reduce their asking price, and move on from Varlamov, Pickard has only made 41 NHL appearances in his career, and as such it’s difficult to accurately predict just how good he’ll be. Is Pickard the goaltender with the career .939 Sv% at 5v5, a mark that ranks 3rd in the entire NHL (min. 500 minutes) since these statistics have been tracked? Or is he something less?

Starting battles aside, Colorado’s goaltending should improve, and soon. Look for their 5v5 Sv% baseline to be somewhere in the neighborhood of .926 or .927, which are the marks they posted in 2014-2015 and 2015-2016 respectively.

Shooting Percentage

Colorado is shooting a league-worst 5.51% at 5v5 this season, which isn’t exactly conducive to putting the puck in the damn net. To put this in perspective, Arizona’s 5.71 5v5 Sh% in 2014-2015 is the worst full-season mark on record. Colorado’s average 5v5 Sh% over the previous 7 seasons, since Duchene entered the league, is 7.99%. Obviously there’s been some fluctuation in team talent over that period of time, but Colorado maintaining this level of scoring ineptitude is, fortunately, highly improbable.

Nathan MacKinnon in particular has come under fire for his lack of production, given that he has just 2 goals in 12 games. Mackinnon has recorded 36 shots on goal this season, totaling 3.98 individual expected goals (ixG) in all situations. His 0.98 ixG/60 at 5v5 ranks 17th in the entire NHL, and he’s been one of Colorado’s most impressive play drivers this season, posting a 8.45 Rel xGF%. He’s clearly been productive, he just hasn’t gotten the bounces yet, as indicated by a 4.76 on-ice Sh%, well below his career average of 8.31%. The points will certainly come. 

—

I’m not expecting Colorado to be an elite team this season, far from it; their ceiling is likely a Wildcard team, especially in the raging inferno that is the Central Division. Anything more would likely require an exceptional team goaltending performance, the addition of a couple productive wingers, or perhaps even both. That said, Colorado are much better than their record currently indicates, and fans should expect an uptick in the wins column in the not-so-distant future.

All data is from corsica.hockey, all visualizations are from corsica.hockey and hockeyviz.com

Follow Luke Steer on Twitter: @lukeasteer